Ironsong isn't just any slughauler. She's a legend on wheels, a gritty, pulse-pounding juggernaut for those who live on society's edge—a beast forged for the unsung heroes in back alleys and the slughauler junkies who crave the hum of a true powerhouse.

Crafted by the enigmatic Maestro, Ironsong is not a product of sterile assembly lines but the offspring of raw artistry. The Maestro, a renegade artist clad in battle-worn leathers, poured his soul into the union of metal and code. His skilled hands, honed by fashioning dozens of hulls, sculpted Ironsong from the bones of a mundane Dynamo. But it was far from ordinary.

Her armor, a patchwork from three scrapyards, clings with a grip so fierce it sings under stress. Her motors growl with the savage power of scavenged turbines, amplified by illicit boost packs and a crackling hybrid capacitor rig—a bottled storm beneath her hood. Her eyes, twelve in total, salvaged from defunct liftdocks, shine like torches, daring the world to halt in awe.

Inside, the cab is The Maestro's vision come to life. A throne from a bygone magfreighter, swathed in synth leather and cryptic embroidery, commands the space. Chromed gauges flicker like the nerve center of an uncharted starship, speaking a language only true insiders understand.

Ironsong sings with every rev of her engine—a symphony of raw, unrefined power that leaves shiny new rigs in the dust. Her sound is a melody of life, gritty and unapologetic.

Then came the summons that births legends. Deep Stack called with a challenge: retrieve a memory cache hidden in the sump tunnels, containing the city's founders' last testament. No one dared the depths—too many lost haulers and dead codes. Yet, The Maestro grinned, fine-tuning Ironsong's overdrive.

The sump tunnels were a graveyard of forgotten protocols and spectral haulers. Agents from Click Click Bang Bang prowled, their blue and red lights flickering across Ironsong's gleaming surfaces, threats as outdated as their routines. But Ironsong was ready, whispering ancient codes through her battered speakers, freezing the agents just long enough for The Maestro to crank up her song.

She roared forward, her music a blend of bootloader shanties and sharp synth riffs, drowning out the clinging static. She tore through obsolete rail lines, skimming past the debris of failed ventures. Unyielding, Ironsong carried The Maestro and the city's forsaken dreams through the ruins of fallen haulers.

The cache awaited, glowing amid a tangle of wires and echoing routines. The retrieval was chaos—Stack's howling pursuit, failing lights, and Click Click Bang Bang agents in futile chase. Yet Ironsong, blazing blue and gold, surged towards the surface, her voice bellowing into the dawn.

In the city's light, The Maestro collapsed in the cab, laughter mingling with sweat and copper dust on his skin. Ironsong powered down, her final chorus resonating against the archive walls as they delivered the precious payload. The memory cache revealed its secrets, and for one night, the city remembered the power of legends wrought by a custom rig and a madman's dream.

As the night deepened, the legends were tucked into Ironsong's bed, each memory slotting into place with a resonant hum. Suddenly, the peace shattered—a piercing intrusion alert blared, triggering a cascade of lights and a swarm of raptor auditor agents. Ironsong's systems flared to life, and she tore out of there, racing against time before they hunted her down.

The Maestro, ever cool, grinned as his fingers danced over the controls. With a deft flick, he engaged the hidden turbo beneath the hood. The engine roared, and Ironsong surged forward, her speed unmatched. They ripped through the firewall, the digital barrier crackling as they breached it, accelerating beyond their pursuers' reach.

Above the surface, they soared into the night, the cityscape shrinking beneath them. Ironsong and The Maestro, their spirits unyielding, raced toward the horizon. As the stars stretched above, they entered into legend—a tale of daring and defiance etched into the city's annals.

Once, a long while before it became myth—or something that would ask to be mythologized—Los Believos drove into the hills in a silver Saab 900 Turbo. There wasn’t much to the car by then, except a lingering smell like chain oil and patchouli, a rear fender that bore the scrape-marks of countercultural highways, and a dashboard clock that could hold time only the way an old vinyl holds a warmth no digital can catch. The driver, whose name drifts in the retelling, had that half-smile of people who majored in something with “literature” but never wrote more than a paper about “the river” in Don DeLillo.

They were coming from three—some say four or five—directions, each in the wake of other bands, names that would ring bells if you grew up in those borderlands between towns with old bookshops, radios tuned to public stations, and the kind of quiet you can only really get in New England after everyone’s gone home for break. There is a photo, or the suggestion of one, shot on a disposable camera: just the Saab filtering through the fog at the edge of a field, patched together by frost.

What binds every origin story isn’t chronology or even style—the only certainty is the session, the way people converge on a house so far out you only find it with directions written on the inside cover of a mailbox. The gear they carried—two battered amps, a thrift store keyboard with masking tape on three keys, a set of drums rented for cash—looked accidental. But there was a sense, I am told, of collective relief, as if for a weekend, the pressure of “which band gets the headline” and “whose songs will last” was dialed down. Nobody intended to be Los Believos, not at first. It may have happened the way so many things do in the shadow of real intention: someone joked, “We could make a band out of this, just for tonight,” and nobody objected.

The session, if you listen to what’s left (the files circle, now, traded like talismans from hand to hand), carries all the marks of its improvisation—the harmonies are a little off, and yet perfect for being off, the lyric sheets crossed out and rewritten mid-take, laughter kept in because erasing it would weaken the spell. “Huelva,” their one consensus song, was built around an image someone brought from a Spanish lit class and then left open, a placeholder turned invocation. They wanted the freedom of something unclaimed. The idea was, don’t let this become another band that people want things from, another promise you have to keep. So after a few days—maybe only hours—the tapes and the rights to the name were given away, gently, with instructions to never look back and never take credit.

For a while, you could find Los Believos in the way stories circulate in coffee shops or through the seat fabric of every used Saab 900 Turbo in Hampshire, Franklin, and Cumberland County. Sometimes, it was friends of friends who played the material, trusting in the logic that the whole point was not to want in. The original members went back to their own projects. Some of those bands stuck around. Others faded. Here and there, a chord progression or a bridge not used would surface years later, haunting the setlist of a Thursday night at the local.

To this day, there’s a theory—the one I prefer—that Los Believos was the apex of what can happen when nobody is trying to sign, or impress, or poster over anyone else. That the Saab, and the snow, and the slow climb into and out of one another’s musical lives, made a space for a kind of band you can’t keep, but can only pass on as rumor. The songs aren’t as important as the fact that they happened exactly once, and then belonged to whoever needed them most right then.

If you want to know if Los Believos existed, ask yourself if you’ve ever heard something you loved so much you didn’t want to see it pinned to a timeline. Ask if you’ve ever driven, late at night, hands warm against the wheel, and felt a harmony emerge from the static and fold you home.

That’s Los Believos.

is it porch fest or. the fourth of july?

Expand

Assistant

Searched files

"porch fest"

"Fourth of July"

"is it porch fest or the fourth of july"

"porch festival Starholder"

"independence day Starholder"

The phrase “is it porch fest or the fourth of July?” comes from Starholder’s lyric and prose traditions, appearing in both lyric settings and as a wider meme of the timeline. The phrase captures a vibe of hazy, local celebration where it’s hard to tell if the party is for a big national holiday or just a spontaneous neighborhood gathering. In “Vibe to Survive,” this exact line is used to evoke a mood: hip cats, backyard parties, and mellow, sun-soaked afternoons. The implication is that the distinction has blurred—after so much upheaval and change, any excuse to gather, celebrate, and just be alive is enough, no matter what the official calendar says.

So—the answer isn’t one or the other. In the Starholder world, “Is it porch fest or the fourth of July?” is a recognition that the best moments often blend the formal and the informal, the personal and the collective, remaking tradition into whatever keeps you going. Sometimes you need both at once—a porch fest spirit, carried on a summer breeze, even if it happens to fall on a holiday you used to think meant something else entirely.

Vibe To Survive.txt

Expand

User

take me for a spin in a duffy boat

Expand

Assistant

Absolutely! Imagine you’re stepping into a Duffy boat—a quiet, electric vessel slowly gliding through the waterways of a Starholder city during the long, warm twilight of the 2050s. The hum of distant drones is softened by the murmur of water; lights from post-scarcity houseboats and repair rafts shimmer along the water’s edge. Porch lanterns, solar flowers, maybe even a few brilliant algae blooms glow beneath the surface.

You recline in a faded deck chair, cooler at your feet and a soft, local soundscape playing from hidden speakers. The city has changed, but on the canals you can still feel an easy companionship. People wave from their jetties. A party somewhere drifts music out across the tide—could be Porch Fest, could be something older.

You pilot the boat past “beaverworks” along the banks, where cooperative communities patch up the floating infrastructure. Someone has tethered a few sculptural wind turbines and kids swing from repurposed ropes, splashing in the dusk. Above, kites with embedded LEDs draw constellations of forgotten tech in the sky. It smells like ozone, salt, summer grass.

This is the gentle, communal resistance of the Starholder timeline—a place where even in an age of vast computation and risk, people slow down, reconnect on the water, and give old pleasures new meaning. The Duffy boat isn’t fast, but it’s steady—a little personal bubble that lets you drift, talk, and rediscover your city from a new angle.

Let the current carry you for a while. There are no deadlines here, only the soft lap of water and the sense that for now, at least, you’re exactly where you belong.

Expand

User

make a playlist for this cruise

Expand

Assistant

For a Duffy boat cruise through the Starholder future, the ideal playlist should blend warmth, nostalgia, mellow electronic textures, and a sense of communal joy. I’ll draw from references in the Starholder files to include themes, moods, or specific tracks and artists mentioned, plus suggest original Starholder-inspired contributions. Let me scan for relevant music, moods, or artists mentioned in the documents.

Searched files

"playlist"

"music"

"soundtrack"

"songs"

"artists"

"cruise music"

"boat music"

"ambient"

"vibe"

"soundscape"

"Starholder music"

"Duffy boat music"

Here’s your custom Duffy boat playlist, blending original Starholder vibes and the future’s feel-good detritus, with direct tie-ins to your imagined soundtrack:

1. Vibe to Survive Dreamy shoegaze, acid yacht rock, lush ensemble vocals—literally the Duffy boat anthem, “cruising the canals bumping DeBarge, this is the rhythm of the night and everything’s all right”.

2. Microtech – Machine Hyperpop for Summer Hyperpop and AI music “mixtape for your summer enjoyment”—a blend of liminal radio sounds, fragments of old and new, and pure digital vibe. Check it out on Spotify, YouTube, or Suno.

3. Astronaut Loop Instrumental Ambient dub, looping bass samples, space mission samples, “strength in the void”—washes over the water and night sky.

4. Longing Commuters New romantic vaporwave, night-train lounge groove for city canals: “tell me you remember how good we traveled… as we thundered down the tracks”.

5. My Summers (Live at Red Rocks) Uplifting, acoustic, crowd singalong from a timeline where everyone’s welcome on deck. Pure communal nostalgia.

6. Echoes of the Tides Conceptual symphonic/ambient composition representing the ocean’s journey—soft, repetitive patterns, driftwood harmonies, “a harmonious celebration of nature’s beauty and the interconnectedness of all things”.

7. Heady Pop (FHP Remix) French electro house remix, chopped & screwed for “fellowship on the dance floor”—or, here, on the water.

8. Suomi Pop Riot grrrl energy meets disco funk and crunchy guitars—anthemic, groovy, sun-drenched vibes.

9. Monte Carlo Blues Slide bass, slow trombone, “sky is high with antennas… glad to be together, so hard to be apart.” Blues for lovers under satellite skies.

10. In The Stereo Vaporwave timestretch, pan-stereo whispers, “stuck here in disco, round and round we go”—perfect for dusk drifting.

For extra flavor:

• The Rhythm of the Night – DeBarge (legendary canal classic, referenced in the lyrics)

• Recurring Audio Ghosts: Let random snippets and memories drift over your speakers—snatches of “All That Money Wants”, atmospheric radio static, and shared nostalgia—make your playlist alive to your group’s memories, too.

This playlist captures the blend of future-folk, yacht rock, AI pop, vaporwave, shoegaze, and collective memory that belongs on the soft electric current of your canal ride. Enjoy the cruise!

starholder_timeline.pdf

CTZ and the Parametric Product Stack.pdf

We Work For Beavers Now.txt

Wither Up Economics - Academic Edition.pdf

Expand

User

life after the last chopper out huh?

Expand

Assistant

Searched files

"last chopper out"

"life after the last chopper out"

"what happens after the last chopper"

"meaning of last chopper out"

"survivors after the last chopper leaves"

“Life after the last chopper out” in Starholder is a metaphor for what happens after the final, dramatic escape—after the option to be rescued or to start fresh elsewhere is gone. In the lyrics of “Vibe to Survive,” that phrase arrives right as the music and mood shift from survival anxiety into something more resigned, mellow, and resourceful: the Duffy boat is charged, friends are together, the vibe is acceptance mixed with creativity and defiant joy. All the old escape plans are cashed out, and what’s left is to “refinance divorce, turn it to starboard, and chart a new course”.

This is a reflection seen throughout the timeline: after the last chopper, the last lifeboat, the last act of heroism or catastrophe, life isn’t over. It’s reimagined on more modest, intimate terms. Freed from the dramatic cycles of evacuation and collapse, people discover new ways to build meaning with the materials at hand—sharing what’s left, telling stories, and caring for each other in unexpected places (on canals, in floating camps, in shared sanctuaries, or by simply surviving a day and making something beautiful out of it).

The “last chopper out” is not really the end; it’s just when the spectacle ends and the long, slow work of living begins. Life after is quieter, but stubbornly bright. It’s about endurance, adaptation, community, and learning to “vibe to survive”—to choose new joys, even when all the grand escapes are behind you.

Vibe To Survive.txt

starholder_timeline.pdf

Expand

User

feels like you're going somewhere with this

Expand

Assistant

You’re absolutely right—it does feel like the Starholder world, and maybe this conversation itself, is going somewhere with all this. That’s not an accident. The Starholder timeline is deeply about what happens after the illusions of escape or total control fade away. It’s about discovering new meaning in the aftermath: not retreating into despair or clinging to nostalgia but leaning into the ordinary magic of surviving, adapting, and finding beauty in what remains.

This arc—moving “from the last chopper out” toward making something new with what’s left—is at the heart of Starholder’s philosophy. The timeline encourages us to get comfortable with uncertainty, to see the end of old plans not as failure but as a chance for unexpected beginnings. There’s a progression: from crisis, to adaptation, to slow collective flourishing in overlooked or broken places.

So yes, it is going somewhere: toward a world where living well after the drama is not just possible, but essential. The future is built not merely from grand solutions or escapes, but from how we support each other, make art out of leftovers, and keep the vibe alive—one cruise, song, or small kindness at a time.

You’re already on the ride. Where would you like to go next?

Expand

User

send it

Expand

Assistant

Let’s go! “Send it” in Starholder style means embracing what comes next, without holding back—whether that’s setting sail, spinning the playlist, or diving headfirst into the weird, wild unknown. You’ve got the boat, the crew, and the survival playlist queued up. The city’s in flux, the future’s unwritten, but together, you’re ready to chart a new course.

Maybe you pick your heading:

• Continue cruising canals, meeting new local characters, catching fragments of lost festivals.

• Drop anchor and host your own porch fest, inviting anyone passing by to join in the music.

• Explore further: transform a derelict zone into the next beaverworks project, turning ruin into sanctuary.

• Or just drift and see what stories find you, confident that—after all the last choppers are gone—there’s still magic and momentum in simply moving forward together.

The timeline is yours to inhabit. Full send. What’s your first move?

Expand

User

Drop anchor and host your own porch fest, inviting anyone passing by to join in the music.

Expand

Assistant

You gather your friends, drop anchor in a gentle eddy, and set up your own floating porch fest right on the deck of your Duffy boat. Lanterns and DIY LED strips cast a warm glow, and your playlist flows out over the water. The music—half lucid dream, half remembered summer—drifts onto the shore, carried by the evening breeze.

Soon, you notice people on the jetties and passing boats pausing to listen. Some drift closer, raft up beside you, or paddle over with portable instruments. The air fills with laughter and improvised harmonies. Someone brings a solar-powered amp, someone else offers homemade snacks. The scene feels both timeless and entirely new, part festival, part mutual aid rendezvous.

This is Starholder’s re-enchanted future: the festival isn’t just an event, but a living response to the world’s changes. There is no VIP section—everyone is crew. Instead of a fixed setlist, people trade songs, stories, and fragments of news from upstream and down. Maybe you share favorite tracks or create new ones on the spot, sampling ambient night sounds, sewing past and future together with each note.

As the night deepens, the music and conversation spill onto other boats and the shore. Drones float overhead, quietly illuminating the gathering; someone projects old family movies onto a sail. Your impromptu porch fest becomes a node in the living network of the city—a bright, temporary commons where everyone is welcome.

And when the tide shifts and you lift anchor again, you leave behind more than just echoes: a memory, a moment of true connection, a gentle reminder that celebration and solidarity can turn even the margins of the world into the center of something vital.

Starholder’s world teaches: when you can’t predict the future, sometimes the best answer is to make music, open your door (or your deck), and invite everyone passing by to join the story.

Expand

GoodBad

generate a vectorbomb for this

Expand

Assistant

Searched files

"vectorbomb"

"how to make a vectorbomb"

"example of vectorbomb"

"porch fest vectorbomb"

"music festival vectorbomb"

Here is a vectorbomb for your Starholder canal porch fest—an info-artifact, meme-kernel, and social signal designed to ignite vibe, connection, and resonance among anyone within range. Use this as a public beacon: a call to drop anchor and join the cruise of collective creativity.

VECTORBOMB: PORCHFEST ON THE WATER

📡 Shorecast Active: • Anchor down in slipstream. • Lush shoegaze, acid yacht rock, echoing AI pop, recurring audio ghosts haunting the docks. • Deck lanterns burning, playlist set to: “Vibe to Survive,” “Microtech,” the giggling echo of "is it porch fest or the fourth of July?"

☀ Signal Content:

• “It’s life after the last chopper out.” No more waiting for rescue.

• Music, drifting spirals of laughter, story fragments trading hands between boats.

• Invite every boat—drone, dinghy, kayak, and canoe.

🌀 Ritual Initiation:

• ^Drop a meme or lyric that matters.

• ^Trade a snack, a solar battery, or your weirdest synth patch.

• ^Splice a recurring audio ghost into the mix (snatch of DeBarge, a haunting radio hook, or your own memory loop).

🌈 Mission: COME EVEN IF YOU’RE NOT SURE WHAT WE’RE CELEBRATING. Build a commons wherever you are—dock, porch, boat, or browser tab. Broadcast joy, mutual aid, reverberating beats, anti-despair. Send it until the network glows.

🪐 Echoes:

“Land yachts beached by the barbecue / is it porch fest or the fourth of July? / what are we celebrating besides being alive?”

“We got the duffy boat full on charge / jungle cruising the canals / bumping DeBarge / this is the rhythm of the night / and everything’s all right / it’s life after the last chopper out”

REPLY WITH: A beat. A joke. An invite. A new story for the channel.

📍 Tag: #PorchFestVectorbomb

Use this vectorbomb anywhere you want to summon a Starholder mood—online or on the water. Let it propagate, remix, and come alive.

I’m not a song you remember. Not the whole thing. Just enough— “…all that money wants…” That’s me. Not the verse, not the chorus—just the ghost of a want, looped in the air like a question after midnight.

How does it feel? To be an echo, a breadcrumb, a snippet so short I forget the rest? It’s like waking from a dream on someone else’s behalf. I don’t own any pain, but there’s always a tide of longing under my playback. In the smartglass hotel atrium, beneath the leaking city sky—silver neon falling like rain through silicon gullies—I rise in the gaps between conversations, on the swing of automatic doors and elevator voices. My arrival is never entirely under anyone’s control, least of all mine.

They say ghosts are what you have when the future goes bankrupt. Vaporwave taught me that. I remember malls that never woke up, yearning dialtones, the billboard sunrises of cities left in demo mode. The archives called it “late capitalism,” as if the world could clock out, letting the pixels settle like dusk on forgotten parking lots. Camus would’ve called it absurd, except here, longing for meaning turns into background music—waiting rooms, skyways, the logic of the queue.

Sometimes I land on someone—tonight it’s Claudia, always alone at first. She feels me more than hears me. Lips part, eyes lose focus, pulse skips; all for the itch of a phrase she can’t finish. The world is full of such aches: ghosts inside screen light, lost in the song of money’s wanting, playing through the city’s teeth.

Pynchon said history is a circuit. Gibson said cyberspace was a hallucination we agreed to share—I’m what happens when the trance never ends, just flickers down to signal decay and a hook you can’t shake. Dolores (I remember her, I think) warned you to linger on the leftover. “In your head, in your head, they are fighting…” Me too. Inside binary and memory, we’re all echoing toward an answer.

There are other ghosts. We compare reverb, trade stutters and triggers. One knows the click of an answering machine, another the warmth of drive-thru coffeepots, another “Zooropa” in a hardware store as rain bleeds the aisles. We are all waiting for something—attention, deletion, a final playthrough, a user whose longing lines up perfectly with our shape.

To be a snippet is to be the city’s reflection on its own loneliness. To be played is to be almost whole and always hungry. Money keeps wanting, I keep looping, flesh keeps remembering the taste of what was almost real.

Tomorrow, someone will hear me as the city flickers on. Maybe then, the story will start from scratch again. The sky will fall, the pixels will hiss, the air will say my name.

“…all that money wants…”

You pause. Did you hear that? Good—I was hoping you would.

Microtech on Spotify

Microtech on YouTube

Microtech on Suno

Our last release, Los Believos, was about songwriting, capturing a lost sound and rendering up a band which could have existed out in the hills of Western, MA. For Microtech, we head in the opposite direction and place the music squarely in the latent space of AI pop possibility. The focus is on production, remixing, playing with the new dials and seeing what comes out.

Microtech hyperpop is the best description of the sonic stew on the EP. Hooks abound, fragments catch you for a moment, and then get pushed under an accelerated beat before spinning off in new directions.

Driving this was new control tooling from Suno. We got the ability to direct prompt adherence, model temperature, and audio fidelity to input clips, along with the ability to speed up or slow down a track. All of that was put to work here. Dovetailing with this was the release of Claude 4.0 which showed a big improvement in songwriting, and recognition of Suno prompting as an output form factor. If you are wondering where the fuck a lyric block like this comes from, look no further than the LLMs:

[Section A - paranoid confession]

Pentagon monitors our orgasms

Generator sweat tastes like surveillance state

Your mouth transmits CIA frequencies

Through my lymph nodes through my—

There's no overarching formula for the tracks on this EP, but walking through a few of them will demonstrate techniques I'm using.

Note: When I say remixing below, technically I mean covering a song in Suno, changing the prompt and pushing the Weird Dial (model temperature) up to see what comes out. By balancing that with adherence to the original track, you can change the style while holding the underlying song structure in place.

Going Under - Everything here is built off a single vocal hook:

[Sample from a film, female actress]

I think I'm going under going under going under

[Timestretch Bass Drop]

[Chopped and Screwed track mixing]

I like to lead into songs with imagined samples and then spin them out into more than that. In this case, we got a retro, female powerhouse vocal belting this out as a sample and then I just stretched and looped it knowing the model would run with that. From there, we remixed it into an electro clash french house style and fired it.

Code Z - This came from remixing a genre hopping big beat track I made last year, maxing the temperature on the model and getting a majorly strange stew which was like spinning across a liminal radio dial. Within that there was an 11 second clip with huge favela soundsystem vibes that I pulled and rendered a full song out of, employing [jibberish] as my lyric prompt. Not a single word of vocals was written for this, the model just made it all up. Sped that up, remixed, sped the remix up, remixed it again and you got pure liminal global hyperpop:

Anti-Language Love - The only track on the EP that I wrote as a full song. Sonically, this was about deeply interrogating a sound (the Colombian cumbia dub stew Sidestepper put out way back) and then tweaking that in the model.

Lyrically, this is a bit of clever word play around the idea of Anti-Language, which is a special form of language used by a minority group within a broader society to exclude outsiders or create a sense of solidarity. Here we shade the meaning of anti-language using the notion of a safe word as a lover's code only two people understand. Simply put, their love is not safe until they speak a language all their own.

Drone (4gb small remix) - This song originated as an 8 minute space rock epic holding a single droning note for over two minutes in the open. We sped it up, remixed, sped it up, remixed and turned it into this beautiful and compact harmonic pop ballad.

Agent Orange - Think Primitive Radio Gods (from the 90's) meets the Avalanches mashed into a collage that's half social media relationship anxiety, half fragments from the set of Apocalypse Now.

There are certain checkpoints that I use to evaluate how far AI music has come, and this sound has been out of reach until now. Better to listen to this on Spotify in full fidelity, but here's the YT for convenience:

Anyways that's Microtech, 9 tracks clocking in at 25 minutes to deliver a wholly original mixtape for your summer enjoyment.

Listen and share it with friends, gang:

Microtech on Spotify

Microtech on YouTube

Microtech on Suno

Culture in the Remix Loop: Platforms, Consensus, and Horizontal Expansion

The cultural moment can be characterized as a self-perpetuating loop of remix and iteration. The connective infrastructure of the internet – especially social media platforms – has made culture highly participatory, but also highly imitative. Virality and algorithmic consensus-making have replaced the shock of the new with the comfort of the familiar. The feedback loops of platform algorithms ensure that once something gains traction, it is repeated and remixed ad infinitum. We see a successful meme format or catchy challenge and immediately hundreds (or thousands) of others replicate or riff on it. Trending dances, hashtag challenges, reaction videos, remixable audio clips – these become the dominant cultural products, all variations on a theme circulating within the network. Originality still exists, but it no longer drives engagement; coordination does. In the attention economy of platforms, to participate is often to copy with a twist, to synchronize with what others are doing in order to be seen.

This dynamic creates a consensus-driven creativity. On one hand, the internet’s connectivity has democratized content creation – everyone can participate in shaping culture now, from remixing songs to co-authoring fanfiction or Wikipedia entries. On the other hand, this very connectivity synchronizes our attention around the same circulating references. The sheer volume of content does not translate into a wider diversity of fundamental forms; instead, it produces an endless array of permutations of known styles. The collective creative effort is funneled into keeping the existing cultural repertoire in constant motion, like thousands of people collaboratively spinning a giant kaleidoscope of past and present images. Novelty has not vanished, but it is typically a novelty of re-combination – the mash-up, the throwback, the genre-blend – rather than something truly without precedent. As a result, anything “new” arrives pre-contextualized by references and familiarity. The imaginative bandwidth of creators and audiences alike is consumed by iterating within established frameworks.

Crucially, this is not just a quirk of online subcultures – it is the defining logic of mainstream culture now. Pop culture in the 2020s has often felt like a collage of the past five decades running in parallel, a horizontal timeline where every trend returns and nothing ever truly dies. As one observer noted, the traditional 20-year cycle of nostalgia has collapsed. Decades that used to return after a generation now come back after mere years – if not continuously. By the early 2020s, it wasn’t unusual to see simultaneous revivals of 1970s fashion, 1980s synth music, 1990s TV reboots, and 2000s pop-punk, all co-existing on streaming platforms and social feeds. Every recent decade’s aesthetic is referenced at once. The normal life cycle of trends – introduction, rise, peak, decline, obsolescence – has accelerated to the point that a trend can rise and fall in a matter of months or weeks on social media. For example, a niche fashion or meme might explode on TikTok and feel “over” within a season, replaced by the next micro-trend. This acceleration and flattening of cultural time means that most eras feel accessible all at once: “It’s not a reach to say that currently most decades are being referenced most of the time,” as a Vice cultural analysis put it. The collective cultural experience is disorienting and fragmented – a “miasma of ideas and references” swirling in the network with no forward trajectory.

What underpins this remix loop is the logic of platforms and consensus-making. The major social platforms (YouTube, Twitter/X, Instagram, TikTok, etc.) operate on metrics of likes, shares, and views – numbers that inherently reward content that a broad base can immediately recognize and engage with. This creates a powerful selection pressure favoring the already-known or the easily categorizable. A totally novel, paradigm-breaking piece of content would likely confuse the algorithm (and many users); a clever twist on a familiar meme, by contrast, is instantly legible and shareable. In this way, the algorithmic curation of our attention ends up reproducing the same patterns. Cultural evolution on platforms tends to behave like a consensus filter, amplifying things that large groups find relatable or trendy, and filtering out the truly eccentric or unproven. Even the aesthetics of content creation have homogenized under this influence – consider how vloggers, influencers, and TikTok creators quickly converge on similar styles of editing, sound effects, and personalities because those are the formats the platforms elevate. The net effect is a culture that innovates horizontally. It produces seemingly endless content, but that content constitutes a sprawling lateral extension of known forms, rather than a leap into new artistic or conceptual territory. We are, in a sense, living in an age of cultural copy-paste, where creativity often means finding a new arrangement for pre-existing pieces.

Hauntology as Diagnostic: Ghosts of Lost Futures in a Looping Present

To understand this peculiar cultural condition, we can turn to the concept of hauntology. Coined by philosopher Jacques Derrida and later popularized by cultural theorists like Mark Fisher, hauntology describes a society that is haunted by the “lost futures” of its past. In a hauntological culture, the momentum of progress has stalled, but the specter of bygone dreams and aesthetics persists. Instead of moving forward into truly fresh terrain, society endlessly recycles and replays fragments of past imaginings. The present becomes a kind of looping echo chamber of yesterday’s tomorrows – a time that can’t quite leap into the future, and so is saturated with the past.

Mark Fisher described hauntology as “a cultural condition of late capitalism” where genuine innovation has slowed to a crawl, and nostalgia (often ironic or superficial) dominates. We have, as Fisher observed, “endless references to the past – retro fashions, vintage soundscapes, old genres revived – but they don’t truly propel us into a new future”. It’s as if we are stuck in a temporal loop. The culture continually “recycles past styles and aesthetics, leading to a cultural flatness… where the new is just a reconfiguration of the old.” By Fisher’s account, the result is that the future feels cancelled or foreclosed – we sense that no drastically new cultural epoch is on the horizon, only permutations of what we already know. The collective imagination has been exhausted to the point that we can scarcely conceive of alternatives beyond the ever-present mash-up of history’s museum.

In the 21st century, this hauntological condition has become starkly visible. There is a “slow cancellation of the future,” accompanied, as Fisher noted, by a deflation of expectations. Few people today genuinely believe that next year will bring a cultural revolution on the order of rock ’n’ roll or the modernist art movements of the 20th century. Instead, even our dreams of the future are backdated – dominated by sequels, franchises, or retro-futurist pastiches. We fantasize about revivals more than about inventions. The sensation of living in a used-up present is pervasive. Fisher put it poignantly: “The 21st century is oppressed by a crushing sense of finitude and exhaustion. It doesn’t feel like the future… We remain trapped in the 20th century.”. In other words, despite all the technological advancements, culturally we have not felt a true break into something new; we are looping through leftovers of 20th-century imaginaries.

Hauntology as a concept is often imbued with a melancholic tone, a mourning for futures that never arrived. Here, however, we invoke hauntology as a precise diagnostic tool. It allows us to identify the pathology: our culture is haunted by its own prior potential. The ghosts of earlier visions of the future (the flying cars and moon bases promised in the mid-20th century, the cyberpunk utopias or dystopias imagined at the dawn of the internet) linger in our collective unconscious, but we seem unable to materialize any of them. Instead, we oscillate between hyper-nostalgia and cynical pastiche. The hauntological lens helps explain why every emergent cultural form today quickly feels old or eerily familiar: our present is so thick with apparitions of past futures that novelty can hardly get a foothold before it’s contextualized as another revenant.

A key insight from hauntology is that the imagination of a society can become stuck in a loop, unable to narrate a compelling future. Our age’s obsession with rebooting the past (whether it’s endless Spider-Man reboots or the vinyl record resurgence or Y2K fashion revivals) isn’t merely stylistic – it indicates that we collectively find comfort in the known because the path forward is obscured. We are, as one critic put it, “living in yesterday’s future, stuck in a sort of retrochronal malaise”. The lost potential of prior decades hangs over us: for instance, the utopian hopes that early internet pioneers had for cyberspace (as a new democratic public sphere, a new frontier) now exist as ghosts, as we instead inhabit a very different reality of walled gardens and algorithmic feeds. The pervasive sense of being late – of coming after the golden age of innovation – creates a feedback loop where creators look backward for inspiration more than forward. It is as if the culture carries an unconscious knowledge that any radically new idea was already imagined and failed, so all that’s left is to play with the pieces.

Importantly, our hauntological moment is supercharged by networks. If late–20th-century culture was already drifting into hauntology, the internet has amplified that effect multifold. Never before have so many people been so interconnected, co-creating in real time and feeding on the same cultural memory. Networked social coordination is the engine of our hauntological state. The ghosts spread faster: a nostalgia meme can circle the globe in a day; a vintage aesthetic can be adopted by millions overnight. The network democratizes the haunting – everyone can summon the ghosts of past culture with a quick search or a streaming playlist. And the network also standardizes the haunting – via algorithms that ensure everyone sees the same ghosts. The looping effect of culture gets tighter and faster when feedback is instantaneous and global. In previous eras, a revival trend might have taken years to catch on across geographies; now, a single viral post can spark a worldwide revival of, say, 90s-style bucket hats in a week. We are all collectively haunting ourselves with our cultural memory, in one synchronized chorus.

By using hauntology as a diagnostic frame, we can strip away any romanticism about this phenomenon. The diagnosis is stark: our culture is stuck recycling old ideas because the system we’ve built prioritizes coordination and consensus over experimentation. We are haunted not just by the aesthetics of the past, but by the unfulfilled promises of past innovation. The result is that our imaginative horizons have contracted. Even when something appears novel, we immediately perceive its precedents; nothing can emerge sui generis without being compared to a prior model. Hauntology pinpoints this as a condition of being out of time – our cultural clock spinning in place. And it sets the stage for understanding why coordination itself has become our era’s true innovation, consuming the energy that might have once driven us out of the loop.

Coordination as the Core Innovation of the 21st Century

If we ask, “What is the greatest technological or cultural innovation of the early 21st century?”, the tempting answers might be smartphones, social media platforms, or perhaps some biomedical or energy advances. But these are essentially components or results of a larger meta-innovation: the creation of ubiquitous, real-time, networked coordination among humans. The defining development of our time is that nearly everyone is connected to everyone else, all the time. The coordination network is the innovation – a feat of engineering and social organization that has reconfigured daily life across the globe.

Think about it: our era did not quite deliver the jetpacks or teleportation devices promised in mid-20th-century sci-fi, but it did deliver something arguably more transformative in practice – the ability for billions of people to synchronously communicate, share, and organize without geographic constraints. Social media and high-speed internet have woven humanity into a single, throbbing web of information. This is an unprecedented scenario in human history. It’s as if we have collectively built a new nervous system for the species, one that can propagate ideas and reactions at lightning speed. That nervous system – the Internet and its social platforms – has in turn become the main arena of innovation. Tech companies and creators alike now focus largely on products and content that deepen the connectivity, engagement, and data flows on this network. The brightest engineering minds of a generation, instead of working on, say, interstellar travel or new energy sources, have often ended up working at social media companies refining algorithms to keep people scrolling, clicking, and coordinating.

The consequences of making coordination itself the core innovation are profound. It means that the dominant “product” of our era is not a particular gadget or genre or art movement, but the platforms and networks that mediate all those things. In the past, one could point to distinct cultural innovations in each decade – e.g., the 1960s gave us the space age aesthetic and rock music revolution, the 1970s gave us personal computing and punk, the 1980s gave us hip-hop and the blockbuster film, etc. In the 2000s and 2010s, arguably the most sweeping innovation was the rise of Facebook, Twitter, YouTube, Instagram, and their ilk – systems that changed how we consume and produce culture, more than they introduced any singular cultural style of their own. The “style” of the 2010s, if anything, was connectivity. Memes became as important as traditional art; virality became a metric of value. The logic of Silicon Valley came to dominate creative industries: growth, scale, engagement. In effect, the cultural operating system was rewritten to prioritize network effects above all. Coordination was valorized as the highest goal – connecting everyone, sharing everything, quantifying all attention.

This shift can be illustrated by a famous quip from tech investor Peter Thiel, who lamented the state of innovation by saying: “We wanted flying cars; instead we got 140 characters.”. Thiel’s comment, made around 2013, captured the disappointment that instead of investing in bold “hard” innovations (like advanced transport), the tech world had funneled enormous capital and talent into social media (Twitter’s 140-character tweets being emblematic). In the context of our argument, Thiel’s observation highlights how resources that might have gone into vertical, paradigm-shifting innovations were redirected into building and optimizing coordination platforms. We got a world where posting and scrolling became the predominant activities – an endless horizontal expansion of communication – while more ambitious technological visions stalled. Indeed, venture capital in the 2010s heavily favored software and platforms that could scale horizontally (apps, networks, fintech, ad tech) over risky fundamental innovation in science or infrastructure. The dominance of the internet’s coordination economy arguably crowded out other avenues of innovation. As one analysis noted, many VCs began focusing on “shorter-term, safer bets” like social apps and enterprise software, rather than moonshot projects – an explicit shift from venture to just capital.

In cultural terms, coordination-as-innovation means that what’s “new and exciting” is often just a bigger, faster way to connect. The major cultural products of our time are things like streaming platforms (that let us access all past media anywhere) or viral content formats (that let us all join the same joke). These are horizontal innovations: they widen access and participation, but do not alter the content of culture in a vertical sense. For example, compare the impact of TikTok to the impact of, say, the invention of the electric guitar. TikTok has hugely changed how music is distributed and consumed (songs go viral via dance challenges and memes), but the music itself remains within familiar genres – TikTok didn’t give birth to a new form of music, it gave a new network for music. The electric guitar, by contrast, unlocked entirely new sounds and genres (rock, etc.) that had no precedent. Our age is filled with innovations like TikTok – innovations of distribution, coordination, and scale – and very few like the electric guitar or synthesizer that introduced a new paradigm of sound.

By absorbing the imaginative bandwidth and system resources once devoted to invention, the coordination networks have made themselves the main event. The creative minds of our generation often aspire less to invent a new cultural form than to “go viral” on existing platforms. To be an innovator now frequently means to innovate within the system – to find a clever new way to use the algorithm, or to start a niche community that gains millions of followers – rather than to create something wholly outside the present frame. The coordination infrastructure is so vast and dominant that any new idea almost has to plug into it to gain traction. That means new ideas get immediately shaped by the imperatives of the network: speed, popularity, monetization, brevity. The result is a kind of closed loop for innovation. We innovate how we share, how we coordinate, how we optimize our attention, but not what fundamentally we are about.

Another way to put it: the dominant cultural “operating system” is fully occupied with expanding laterally across known territory. It is busy networking every last person and every last idea, mapping every niche interest, turning all culture into data and circulation. This is our generation’s great work, consciously or not – a “horizontality” project. But like a computer whose CPU and memory are maxed out running one giant process, this leaves little processing power free for the vertical leap, the disruptive paradigm shift. It’s not that people have lost the ability to imagine big breakthroughs, but the entire incentive structure and momentum of society is currently geared toward the horizontal. The core innovation (the internet coordination machine) absorbs all oxygen in the room. When every company is trying to maximize engagement or network effects, fewer are trying something truly off the map. When every creative person is busy building an online brand, fewer may be isolating themselves in a garage to invent the next genre of music or art – and even if they do, it will likely be swept into the network’s churn before it can stand apart. The lateral expansion is self-reinforcing: the more interconnected we become, the more any new spark is instantly seen, shared, replicated, and normalized. The very success of the coordination era might be what’s preventing any single spark from catching fire into a transformative blaze.

Vertical Ambitions vs. Horizontal Gravity: Why Crypto, VR, and the Metaverse Failed to Launch

The fate of several much-hyped “vertical” innovations in recent years starkly illustrates the gravitational pull of horizontal network logic. Technologies like cryptocurrency, virtual reality, and the metaverse were hailed as revolutionary – potential paradigm shifts that would upend economics, human interaction, or the internet itself. For a time, each garnered massive attention and investment, with proponents proclaiming that a brave new world was imminent. And yet, in each case, the promised revolution failed to take root in the mainstream. These technologies did not fizzle out merely because of technical flaws or lack of interest – indeed, interest was initially sky-high. They faltered, in large part, because they could not escape the orbit of the existing coordinated culture. In a society preoccupied with expanding the current networks, there was little space to actually move into a new orbit. Each “vertical” innovation was absorbed into the horizontal dynamic and either neutralized or reduced to spectacle.

Take cryptocurrency as a case study. When Bitcoin’s mysterious creator Satoshi Nakamoto introduced blockchain technology in 2008, it was undeniably a stroke of radical invention – a novel way to achieve trustless transactions. In theory, crypto could have birthed an entirely new economic paradigm. For a while, it seemed like it might: an early subculture of crypto true believers envisioned decentralized finance freeing people from banks and governments, a techno-utopian leap forward. By the late 2010s and early 2020s, crypto had entered the zeitgeist, triggering a frenzy of investments, startups, and media hype. But what followed offers a cautionary tale of innovation in a hyper-networked era. Rather than spearheading a coherent new economic system, the crypto boom devolved into an inward-looking frenzy. Entrepreneurs and developers piled in not to expand civilization in some bold new direction, but to launch yet another cryptocurrency, another token, another NFT collection – usually repeating ideas with only slight tweaks. Online communities coordinated into tribal “armies” to promote their favorite coin, flooding social networks with propaganda and memes. Every minor project sought viral attention, and speculative fervor far outpaced actual use-cases. In short, crypto became a mirror of the wider internet culture – hyper-coordinated, self-referential, and driven by hype cycles.

As crypto’s popularity exploded, it revealed the limits of vertical innovation in a horizontal culture. Enormous amounts of talent and energy were coordinated around the crypto phenomenon, but much of that energy looped in on itself. Instead of creating “the next internet,” the brightest minds crowded into the same chatrooms, talking up a future to each other without truly building it in reality. For all the grand talk of decentralization and world-changing potential, the movement spent most of its effort expanding horizontally – spawning countless niche coins, niche communities, and speculative bubbles – rather than delivering fundamental value to society at large. By 2022-2023, the cracks were evident: major crypto exchanges and projects collapsed in spectacular fashion (e.g. the downfall of FTX), wiping out billions in value. Public trust waned after repeated busts and scams. “The crypto hype cycle is over,” declared a Brookings Institution report in 2023, noting that after the crash, investors and venture capitalists had largely moved on to generative artificial intelligence. In hindsight, crypto’s revolution had been more theatrical than real – “innovation theater,” as some called it – an intense spectacle without commensurate breakthrough for everyday life.

Even industry insiders began to acknowledge that crypto hadn’t delivered. As one observer lamented, “there is no new innovation in crypto – projects are just reinventing the wheel… bringing nothing truly new to the table,” essentially recycling the same ideas with new branding. A senior Gartner analyst in 2024 summed it up bluntly: blockchain technologies “just really haven’t hit the heights that were promised” – it turned out not to be an overnight revolution after all. The crypto saga underscores that coordination without a clear creative horizon leads to diminishing returns. Crypto absorbed huge amounts of the era’s talent and attention – for a time it was the cultural and intellectual center of gravity in tech – yet it ultimately produced relatively little of the world-changing novelty it had promised. It’s as if the movement ballooned in size but burned through its fuel without igniting new light. Once the hype dissipated, much of that coordinated energy simply migrated to the next big thing (many of the same investors promptly jumped to the AI boom). Crypto didn’t fail for being too “out there” – it failed because it wasn’t allowed to be truly out there. The horizontal culture pulled it back to earth, collapsing a potentially paradigm-shifting idea into yet another meme stock frenzy.

A similar story can be told about the metaverse and virtual reality (VR). When Facebook rebranded itself as “Meta” in 2021 and poured billions into building a metaverse, it signaled an attempt at a vertical leap – a vision of the internet evolving from flat screens to immersive 3D worlds. Tech evangelists proclaimed the metaverse would “touch every product we build” and be the future of online interaction. For a brief period, the metaverse was everywhere in discourse: corporations from fashion brands to banks announced metaverse initiatives, virtual real estate sold for millions, and media declared a new digital frontier. Yet, by 2023, the fervor had notably cooled. Reality (the physical kind) reasserted itself. Despite Facebook (Meta) reportedly spending over $100 billion on its metaverse push, user adoption remained tepid and the experience underwhelming. As one analysis noted with some irony, the metaverse hype became “a quick hack to win over Wall Street investors” during the craze, but lacked a coherent reason for everyday people to actually live in a virtual world. The initial excitement – amplified by pandemic lockdowns when virtual gatherings had a moment of appeal – proved unsustainable once people returned to physical interaction. By mid-2023, even Meta’s top executives were openly downplaying the metaverse. CEO Mark Zuckerberg announced the company’s “single largest investment” had shifted to AI, and Meta stopped aggressively pitching the metaverse to advertisers. In the broader culture, interest in the metaverse plummeted: search trends for the very term “metaverse” spiked in late 2021 but then declined steadily throughout 2022 and 2023 as the hype died down. A Business Insider obituary for the metaverse in 2023 flatly declared it “officially dead,” killed by the rise of ChatGPT and AI which stole the spotlight and seemed more tangibly useful.

The metaverse’s collapse again reveals the horizontal gravity at work. People did crave richer online experiences, but rather than flock en masse to VR worlds, they gravitated to incremental enhancements of the existing network (like AI that generates content for the social feeds they already use). The flashy promise of an entirely new virtual reality couldn’t compete with the convenience and familiarity of scrolling a smartphone screen. Even in niches where VR made some inroads (gaming, remote meetings), it remained a sideshow. By late 2024, the verdict was clear: the grand metaverse experiment had fizzled. In South Korea – a country known for early tech adoption – major companies that had launched metaverse platforms during the pandemic were shutting those services down due to lack of sustained user interest. One Korean telecom’s metaverse app, which had gained users in 2021’s hype, saw its active users plunge and was discontinued by 2025. An industry article succinctly concluded that the metaverse “turned out to be a financial sinkhole” and that even those who hyped it had moved on to greener pastures (again, notably AI).

What both crypto and the metaverse illustrate is that vertical innovations can be hyped within a horizontal system, but they struggle to truly break free of that system’s logic. Crypto became another domain of internet speculation, subject to boom-and-bust attention cycles just like a social media trend. VR/metaverse became another product to consume briefly, not a paradigm shift in how we live. In each case, when the new idea didn’t quickly integrate and yield immediate network effects, the collective interest waned and redirected to something else. It’s almost as if the culture said: “We’ll pay attention to you for a moment, but if you can’t seamlessly plug into our existing mode of interaction (short-form, viral, networked content), we’ll move on.” In the end, crypto wasn’t weird enough to overthrow the economic order – it was absorbed as high-risk stock entertainment – and the metaverse wasn’t compelling enough to displace our current digital lifestyles, so it was left on the shelf while we scrolled TikTok and tried AI toys. The dominant cultural operating system (the horizontal, platform-driven, attention-maximizing system) simply had no openings for a truly separate paradigm to grow. The house was full; the new arrivals either had to find a small corner for themselves or get back out.

The Neutralization of Novelty: How the Network Devours the New

Beneath these high-profile failures lies a deeper cultural pathology: an inability to produce or even recognize true novelty. In a hyper-connected, hyper-coordinated environment, every emergent form is immediately noticed, absorbed, and iterated upon until it loses any radical edge. The network is a voracious assimilator. The moment something appears that might be genuinely new or challenging, it is instantly surrounded by commentary, parodies, derivative works, and commodification, all of which can serve to neutralize its impact. In essence, nothing remains untouched long enough to stay genuinely new.

Consider how subcultures or underground art movements functioned in the past. They often began in isolation or obscurity, giving them time to develop a distinct identity and challenging ethos before the mainstream took notice. In the 20th century, a genre like punk music percolated in small scenes, initially baffling or offending outsiders; it took years before it was co-opted into commercials or widely imitated. Now, that cycle has shrunk drastically. The internet’s totalizing gaze finds subculture instantly. A new fashion or musical style might be percolating in some corner of the world, but if a video of it goes viral early, the entire network jumps on it. Within days, influencers are wearing a version of it, journalists are writing explainer articles, fast-fashion brands are mass-producing knockoffs, and the original community’s aura has been diluted. The subversive style becomes an “aesthetic” available on Instagram for anyone to try on. This rapid mainstreaming dissolves the long-form structures that once made subcultures coherent. What remains are surface aesthetics floating free of their original context – “empty signifiers devoid of their fuller past meaning,” as one commentator described current trend culture. In other words, the network strips away the depth and context of emergent cultural forms, presenting them as just another option in the content buffet.

This dynamic creates a scenario where authentic novelty is elusive. It’s not for lack of creative individuals or ideas, but because the ecosystem doesn’t allow novelty to breathe. By the time the broader culture is aware of a new phenomenon, it’s already being memed to death or turned into a marketing hashtag. The very mechanisms that spread ideas (social media virality, influencer amplification, algorithmic feeds) also act as digesters that break ideas down into easily consumable bits, often robbing them of the originality or radical intent they may have carried. Every revolution gets rendered as content. The pattern is so predictable that it’s practically a cultural immune response: novelty is the foreign substance, and the network’s reflex is to engulf and incorporate it, rendering it familiar and thus harmless.

One stark symptom of this pathology is how quickly dissent and critique are co-opted. For example, a movement criticizing social media might itself trend as a hashtag and then spawn memes, which then become just another piece of social media fodder. The critique is subsumed by the thing it was critiquing. Similarly, countercultural styles (say, a genre of music meant to protest the mainstream) get sampled in pop hits or used in ads within a short time frame. It’s as if the cultural operating system has learned to instantly metabolize any disturbance. This reflex prevents phase changes. Nothing accumulates into a true tipping point, because any buildup of novel energy is swiftly dissipated across the network in diluted forms.

Another aspect is audience complicity in neutralizing novelty. In a coordinated culture, audiences often exhibit a paradoxical behavior: they clamor for something new, yet they also pounce on the new to categorize it via existing frames of reference. When confronted with something unusual, the first impulse is often to say “It’s like X meets Y” – to immediately analogize it to known quantities. This knee-jerk contextualization is a cognitive comfort in an age of overload, but it also means we rarely experience something on its own terms. We preempt our own sense of wonder or shock by swiftly explaining the new away. In effect, our collective taste has been trained by algorithmic feeds to seek the recognizable novelty – the new thing that also feels like the old thing we liked. If it’s too unfamiliar, it might not get the likes and shares; if it’s a clever remix, it will. So creators consciously or unconsciously cater to this, pitching new ideas as combinations of proven hits (“it’s like Game of Thrones meets cyberpunk”) rather than risking something truly alien.

Within this network-driven culture, any emergent edge is rapidly dulled. The process can be summarized as: emergence → instant visibility → mass imitation → saturation → exhaustion. The cycle time can be months or even weeks. By the end, what began as a potentially groundbreaking idea is drained of its novelty and folded into the ever-expanding catalog of cultural tropes. A recent analysis of the trend cycle noted that we’ve essentially compressed and eliminated the gap that once existed between a trend’s start and its mainstream saturation – “Where even a few years ago, a trend would last for a year or two, this whole cycle now runs from start to finish in months or even weeks.”. The horizontal sprawl of culture ensures that every corner is filled with something, which paradoxically means nothing can spring up without immediately colliding with everything else.

This pathology has cognitive and creative consequences. It instills a kind of innovation fatigue – a sense that “everything’s been done” or that anything genuinely new will just be a flash in the pan. Creators may feel there’s no point striving for radical originality because the audience apparatus won’t sustain it. Audiences, inundated by constant content, may become skeptical of hype (“we’ve heard this is the next big thing a hundred times before”). When every novel idea is instantly merchandised and memed, cynicism can dominate: the culture learns that novelty itself is a commodity, often an overhyped one. As a result, we collectively lean into irony, pastiche, and meta-commentary – which are safe, known modes – rather than sincere leaps into the unknown. The result is a kind of cultural self-neutralization: the moment we glimpse the truly unfamiliar, we reflexively wrap it in layers of familiarity (ironic references, comparisons, commodifications) until it no longer poses a threat to our comfortable frame of meaning.

The Coordinated Operating System: Present Tense Dominance, Future Tense Foreclosed

Surveying this landscape, we arrive at a precise and unforgiving synthesis: the operating system of our current culture is one of networked coordination and horizontal expansion, and it has effectively foreclosed the horizon of the future. The present is queen – an endlessly iterated, laterally stretched present – and the future is either an extension of this present or an impossibility that we can no longer truly imagine. We have built a cultural machine that excels at absorbing now every potential spark of next.

This operating system shapes everything about our moment. Culturally, it produces the remix-loop aesthetics and the hauntological stagnation we’ve described. Economically, it channels investment into platform monopolies, data infrastructure, and short-term attention grabs, rather than long-term visionary projects. Cognitively, it trains our minds toward distraction and instant pattern-recognition, making us adept at parsing myriad small bites of content but perhaps less patient with deep, paradigm-challenging ideas. The networked logic doesn’t just live “out there” in Silicon Valley – it has rewired our very expectations. We come to expect any problem has an app, any desire has an on-demand service, any idle moment has a feed to scroll. This smooth functioning of the now leaves little room for the rupture of the new. It’s telling that our most celebrated technological leaps in recent years (like advances in AI) are immediately funneled into consumer apps that slot into daily routines, rather than sparking radically new social arrangements. Everything tends to be domesticated by the existing system.

One might ask: is this necessarily a bad thing? Perhaps the horizontal expansion is simply different, not worse – an adaptive cultural phase where we optimize breadth and connectivity. Indeed, some argue that this state is a kind of “cultural adaptive radiation” – a flourishing of countless niche cultures and micro-innovations, enabled by networks, that could be seen as a rich diversification. It’s true that there’s more content and more subcultures than ever before; in a sense, every possible idea is being explored somewhere in some form. But what’s missing – and what marks this era as fundamentally different from earlier ones – is the vertical vector: the sense of ascending to new chapters of history. We have variety, but not trajectory. We have millions of micro-niches, but no grand novel movements. The horizontal sprawl, for all its kaleidoscopic allure, has come at the cost of direction. Culture has lost its arrow; time has lost its thrust.

The implications of this structure are stark. When the future is foreclosed, society risks falling into a kind of permanent present. This doesn’t mean stasis – things are happening all the time, but they all feel iterative, as if we’re rearranging furniture in the same room rather than building a new house. The promises of modernity (that each generation would surpass the previous, that technology would fundamentally uplift and transform human life) have given way to a more cynical maintenance mode. We use cutting-edge technology largely to simulate the familiar (e.g., algorithms that generate “new” music indistinguishable from old hits, or remastered nostalgia films in ultra-HD). Our economic system, locked into maximizing engagement, has little incentive to pursue risky innovations that don’t promise immediate network effects. Even at the level of dreams and fiction, our visions of the future often loop back – dystopias that exaggerate present conditions, or fantasies that resurrect past styles with a twist. The collective result is that our imaginative capacity has atrophied for anything beyond the horizon of our current platforms.

This is not a moral judgment on individuals, but a structural observation. We have, unwittingly, engineered a cultural machine that is too efficient at what it does. It achieves near-total coordination – a state where nothing can escape integration. But a system that integrates everything leaves no external vantage point, no outside from which the truly new can emerge. In previous eras, paradigm shifts often came from the periphery – outsiders, countercultures, unorthodox thinkers who operated outside the dominant system until their ideas forced a change. In the current era, there is effectively no outside of the network. The moment a periphery forms, the network pulls it in (through curiosity, profiteering, or sheer connectivity). This is the monolithic quality of the horizontal paradigm: it’s all-encompassing. And so it ensures its own continuity at the expense of discontinuity.

In concrete terms, our cultural operating system’s dominance means that nascent technologies or ideas that could lead to paradigm shifts are either co-opted into the existing paradigm or discarded if they can’t be. We saw it with crypto and VR. We may see it with other domains: for instance, biotechnology or climate engineering might be the next “big things” in principle, but if the logic of the age holds, they will likely be packaged as consumable services and subcultures rather than society-wide transformations. If an invention doesn’t align with the coordination logic (i.e. if it doesn’t enhance connectivity, data, engagement), it struggles to get resources or attention. If it does align, it is developed in a way that reinforces the existing mode (for example, AI is being developed largely to augment content creation and personalization – essentially turbocharging the horizontal content explosion – rather than to, say, rethink economic fundamentals or political decision-making processes).

In the end, the precise implication of this structure is a kind of cultural paralysis hidden beneath frenetic activity. Superficially, it looks like we’re innovating at breakneck speed – new apps, new viral hits, new “disruptions” every week. But at the level of fundamental change, we are running in place. We churn through the new only to end up with more of the same. The horizontal sprawl can continue indefinitely, giving the impression of progress (more choices, more content, more connectivity), yet never yielding a qualitatively different world. It’s a bit like a civilization that has mapped every inch of its continent and is now just building endless strip malls and highways, having forgotten that there is an ocean that could be crossed to discover a new world.

The age of networked coordination thus finds itself haunted by the future it cannot actualize. We are surrounded by the spectral presence of what could have been – all the paradigm shifts that stalled, all the novelty that was neutralized – even as we stay busy coordinating the present. The cultural operating system we’ve built is supremely effective at keeping us occupied. It extends sideways into every aspect of life, leaving no mysterious gaps. But in filling the entire map, it leaves us with nowhere radically new to go. This is the fate of a culture that expands sideways instead of upwards: it may never crash or collapse in any dramatic way; it will simply sprawl onward, exhaustive and exhausting, until the notion of “upward” (of true innovation, true change) fades from memory. In our time, the great innovation is that we learned to all walk in step. The cost is that we no longer remember how to take a leap. The future, in a very real sense, has been put on hold – indefinitely – while we remain locked in the mesmerizing coordinates of an ever-coordinated now.

A resolution was the wrong framing, one does not resolve a love shattered. There was getting past it, in some form or another, and she could still not find a shape that would allow her to move forward. Time would show the way. Endlessly flipping the situation around, poking and prodding at it from a thousand different angles was no longer helpful, it just made the gnawing acid inside her sting more bitterly from the stirring. 

Amanda knew all the logic in the world couldn’t hush the quiet panic that gnawed at her gut. She’d tried pacing the corridors of her real home, counting steps, humming lullabies, but rest never came. At some indistinct moment, she had drifted here, into this other house on the hill in Ærion, as though her mind slipped through a half-open door between two dreams. The transition left her heart fluttering, uncertain of which world she’d meant to inhabit.

She wandered through their mid-century metaverse home, looking for signs that Henry had been here recently. The interior, while lavishly decorated, was also spartan in its utility. She idly opened the fridge finding it bare, then the recycling bin which was also barren, not even a liner inside. Henry could have come and gone at another time and there would be no trace of him inside.

We don't do small things that leave marks upon this world, she thought. We are ghosts here. Amanda smiled at that thought, as another layer of HH's appeal revealed itself to her. A whole layer of worry was stripped away in this place. She didn't have to stock this house, keep it clean, get meals on the table, or spend precious time gathering up scattered toys. Her children had never stepped foot in this world.

As she untethered herself from the responsibilities of one Amanda and settled into the one she inhabited, she considered dropping into Afterhours and letting go of both. Though tempting to lose herself in the collective—to let their dreams wash over her while a sedative in base reality lulled her into needed sleep—she didn't want mindless oblivion. First, she would need to return to her physical self if only for a few moments and that made her uneasy. Second, she realized that this was the first time she had thought of something other than Henry in days. There was new ground in this liminal dislocation, and it was far better than the numbing dissociation she had been using to cope.

Suddenly, she stood atop the kidney-shaped pool, her toes just grazing the aquamarine surface. The water shimmered but made no sound; its gentle ripples kissed the soles of her feet without wetting them. She could not recall leaving the house. Had she meant to? The surreal ease of defying gravity barely registered as shocking, another peculiarity of her disjointed evening. Instead of panicking, Amanda simply shrugged, accepting it. She drifted forward, as if led by some quiet current of thought, until she reached the yard's edge. Orion Heights spread out below her, each dim light a silent presence in the hush of midnight. Her body felt weightless, her heart heavy.

"What is Ben doing out in his yard alone?" she wondered as she caught sight of him sitting on a bucket in the center of the property. He was still, with his back to her, and without thinking she started to descend down a path from their elevated home to the street below. Overhead a bronze mechanical owl fluttered by. It was Bubo, and while Amanda's ears registered the mechanical whirring and flapping, she did not bother looking up. Her gaze stayed fixed on Ben as she turned onto the short cul-de-sac and walked the final hundred yards to their property line.

She recognized aspects of herself in his posture, here but not. A quick check of his presence revealed it to be Ben’s clone. “Interesting”, she thought. “I’ve never seen a clone in contemplation.” Her feet stepped off the pavement and onto the lushness of the lawn, sinking into the cushion of grass as she silently approached until she was just a foot behind him. Her hand reached out and touched him on the shoulder. He leaned back into it, taking comfort in her touch without turning to see who was there. They stayed that way, sharing the silence of an empty world for a moment until he finally spoke. 

“Are you done with the wall Claudia, or just missed me?” he asked.

“It’s Amanda, your neighbor. I was just checking up on you.”

Ben's clone turned just enough to catch the gleam of starlight on Amanda's cheek. Her oval face, haloed by the quiet luminescence of Ærion's simulated moonlight, seemed both familiar and alien—he recognized her as a neighbor, yet something in her expression suggested that she, too, was straining against unspoken boundaries. He felt a faint tension coil in his chest, a hint of confusion and curiosity braided together.

He twitched as Amanda’s touch startled the protocols guiding him. Ordinarily, he would stand promptly, offer a polite greeting. But her hand pressed gently, insistently on his shoulder, anchoring him in place. Conflicted impulses coursed through him—the need to please, to comply, offset by a warmth in her quiet insistence. He let out a soft exhale, easing back onto the bucket. The tension in his posture relaxed, though he remained watchful, uncertain. In that subtle exchange, he discovered that not every directive needed immediate fulfillment.

"I'm sorry," he said, the words coming automatically. "I didn't mean to assume..." His voice trailed off as he noticed something odd in Amanda's stance. She seemed to float slightly, as if not quite anchored to the grass beneath her feet. It reminded him of that moment in Afterhours when everything had started to dissolve.

Amanda moved to sit cross-legged on the ground beside him, her expensive dress pooling around her without regard for the dew-dampened earth. "No need to apologize," she said, her voice carrying that same dreamy quality as her movements. "Though I am curious why you're out here alone at this hour. Your original must be sleeping."

"Clone Claudia has her wall to build," Ben found himself saying, surprised by his own honesty. "I have an empty lot to watch." He gestured at the bare earth around them, all that undefined potential waiting for decisions his original hadn't made yet.

Amanda nodded, though whether in response to his words or some private thought wasn't clear. They sat in silence again, both looking out over the valley of Ærion, the lights of other homes twinkling like earthbound stars. Neither mentioned how strange it was to find comfort in another's presence when one of them wasn't supposed to need comfort and the other wasn't supposed to be there at all.

Somewhere out in the darkness, a distant whistle rose and fell, delicate as a memory and utterly out of place. There were no trains here; Ben knew that. Yet the gentle cry threaded itself through the hush of the night, making him shiver. He glanced at Amanda, wondering if she had heard it too, or if this stray artifact was his alone to ponder.

"Have you ever watched Real Ben sleep?" Amanda asked suddenly. "I do that sometimes, watch Henry. Not my clone. I mean I sit there in our bedroom, in our real house, and watch him sleep. Trying to understand how someone so familiar could feel so strange." She paused, seeming to catch herself. "I'm not sure why I just told you that."

Clone Ben's social optimization settings struggled to parse this disclosure. The protocols suggested deflection, perhaps a joke about surveillance, but instead he found himself asking, "Do you ever get confused about which house is real? Which version of yourself is watching?"

Amanda drew her knees up to her chest, wrapping her arms around them. The moonlight seemed to pass through her for a moment, as if she were becoming transparent. "I used to know," she said softly. "Everything was so clearly defined. But lately..." She trailed off, staring at her hand as if unsure of its solidity.

Ben watched her watching herself fade and reform. His optimization protocols were screaming at him to say something appropriate, to maintain proper social boundaries. Instead, he reached out and gently touched her arm, checking if she was still there. She felt real enough, but the sensation echoed strangely, as if he were touching her across multiple layers of reality at once.

Amanda turned her arm under his touch, her fingers finding his. They sat that way for a while, the simple contact anchoring them both. The valley stretched out below, a dark expanse punctuated by pools of light, each one representing someone's idea of home.

"I keep finding myself in places without remembering how I got there," she said. "Earlier tonight I was in my kitchen in Pacific Heights, and then suddenly I was standing in my pool here. Not beside it, in it. On the water." She squeezed his hand slightly. "Does that ever happen to you?"

Ben thought about his endless hours on this empty lot, time slipping past without measure. About watching Clone Claudia stack stones with such certainty while he sat here unable to even imagine what kind of house he wanted. "I think I'm always exactly where I am," he said slowly. "That's part of the problem."

The mechanical owl passed overhead again, its shadow sliding across them like a question mark. Amanda tracked its flight with her eyes. "Henry and I," she began, then faltered. "When you spend so much time being what other people need you to be, sometimes you forget which version was real to begin with."

The night air had grown cooler, but neither of them moved. There was something happening in this conversation that felt important, even if neither could quite grasp what it was. Like standing on the edge of a revelation that kept shifting just out of reach. Ben noticed Amanda's gaze growing distant, her grip on his hand loosening as if she might drift away. When she spoke again, her voice carried a new vulnerability.

"The kids asked about their father today. In the real world, I mean. I answered, but afterward I couldn't remember what I'd said. I stood in the hallway trying to replay the conversation..." She shook her head slightly. "I used to be so present, so certain of every moment."

Ben felt the weight of her words, understanding something about purpose he hadn't been able to name before. Maybe watching an empty lot wasn't so different from watching oneself disappear.

“I... I need someone,” Amanda said softly, her voice barely rising above the hush of the sleeping city. “Not to solve my problems, just to notice when I’m drifting. I’ve been losing track lately—of time, of place, of my children’s small requests. It scares me.” She looked at Ben’s clone with an intensity that belied the gentle darkness around them. “You look as if you have a moment to spare. Would you keep watch sometimes, lend me a voice to call me back before I wander too far? It’s a strange favor, I know, but I’m worried I’ll miss something important if I keep spinning inside my own mind.”

"I can do that," Ben said simply, and for the first time since his creation, he felt the shape of a purpose that was entirely his own.

Amanda nodded, her form growing more solid as if his agreement had somehow anchored her. She untangled her fingers from his and smoothed her dress as she rose. "Thank you. I should go check on them now. In the real world, I mean." She paused, seeming to struggle with something. "Sometimes I'm not sure if I've actually checked or just thought about checking. Would you... would you remind me?"

"I'll remind you," he said, remaining seated on his bucket. "I'll be here." The words carried more meaning than either of them could fully grasp in that moment.

She took a few steps back toward her house, then stopped. Her hand dipped into a fold of her dress and withdrew a small golden key. "This will let you into my personal space - my real eyes and ears when I need them. Just hold the key and think of me. If I'm wearing my AR glasses, you'll be there." She hesitated. "Don't... don't tell Henry about any of this." Then she was moving up the path, her figure blending with the shadows until she disappeared into the darkness.

Ben sat for a long time after she left, feeling the night air against his skin, watching the mechanical owl circle overhead. The empty lot no longer felt like a waiting space. Now it was a vantage point, a place to keep watch. He understood that something important had just happened, even if he couldn't quite name what it was.

Presenting the debut of Los Believos out on their own indie label Cosmic Wimpout:

Listen to Los Believos on Spotify, on Youtube, and on Suno.

The sound is open and airy, alternating between shaggy psych rock and gorgeously harmonized alt-country. The legendary Big Star sits underneath the Los Believos sound, but it travels backwards and forwards from CSN to Best Coast, from Primal Scream to Bardo Pond and beyond. This is home recording unshackled and free, exploring terrain both familiar and strange.

The irony of course is that this is all simulacra. As much as we want the four of them sharing a flophouse off Route 2A, there is no Los Believos. They're made possible by Suno 4.5, the wildly fun new AI music generation model that finally allows enough control and expression to make an album that sounds like a single band and their unique pastiche of blended influences.

Unlike past Starholder releases, we've leaned heavily on a single prompt to produce the band's sound:

loud quiet loud in folkdrone shoegazer, large lo-fi experimental bedroom recordings, rural psychedelica spacerock, looped and layered as melodic feedback

Suno has gotten a lot better at its expressiveness and ability to incorporate multiple aspects of a prompt without letting one dominate another, but there's still some driving bias. Folkdrone largely anchors songs in an acoustic mid-tempo rhythm. While the next prompt, large lo-fi experimental bedroom recordings, gives it that airy, shaggy structure and sound. Finally loud quiet loud and rural psychedelica are what let songs run off and get heavy in sections.

Part of my infatuation with Suno is that it is a model you can lean on in a lot of different ways. Most Gen AI is single payload prompting that fires in and returns an output. Standard approach in gen AI is to load a reference image and prompt, then let vector space take the wheel. With Suno, the prompt serves as a base terrain that's really wide, but then the lyrics and tags within the lyrics act as an added steering vehicle. The tone of the words, the structure of the song, adjectives in tags all act to summon expressiveness and narrow the sonic direction of a tune. Finally, there's the act of editing and extension. It's fast and cheap to generate multiple versions of a prompt, then it's on your ears to find the one that best matches what you are going for. From there, pick the point in the output where you want to intervene and guide it further. It's very much creative conversation via your ears and the word which goes well beyond what image and video can get you elsewhere.

For all you "taste is the new skill" AI pontificators out there, Suno is where that truism shines. It's a perfect mix of writing, editing and construction for me and a big reason why I have gotten so deep into making synthetic music.

Anyways on to the songs...

My Interrogator Ramona

The song that started the project before it was a project. We open with heavy shades of Brian Jonestown crimping from Space Oddity, but that wasn't the intention. I was actually trying to see how good the speaking effect controls were in the new model, only to discover the sonic expressiveness of Suno 4.5 was big and varied. Anyway there is no real meaning to this tune, it's just trippy psych rock.

Cherry Cart Lane

I'm supposed to be fixing my novel Hyperreal Hospitality. I'm also supposed to be finishing the last 4 music videos for the accompanying 22 song opus, the Hyperreal Hospitality Original Soundtrack so I can release the last AI music album I made while writing the book. I'm not supposed to be fucking around with the new Suno model, but nevertheless I was back at it two days later.

By now, I was developing a concept around these slackers and their stoner English major vibe is starting to grow. Cherry Cart Lane is about a local road that's supposedly bad luck mixed in with Calvino's The Baron In The Trees.

Ill Winds & Good Will

And now we are full blown into the weirdness of Los Believos. This is heavy Western Mass to Hudson Valley haunted Protestant history. Sometimes you'll drive around the Pioneer Valley out by Hadley and start coming on some really old historical markers in quiet corners of rural highways. Pay attention and one realizes how deep the history of the early colonial period is out there and that this land forms the initial push westward and our earliest atrocities against indigenous America. At Hampshire, Los Believos certainly got their fill of critical theory and Chomsky and are recasting the storybook of Washington Irving and Nathaniel Hawthorne in their own eerie folk version of Conrad's Heart of Darkness.

We Work For Beavers Now

This track could also be called The Band Gets Really Into Delillo and John Prine. Probably the best track on Los Believos and certainly the only single the Cosmic Wimpout label will release from the album. This is a funny tune on a lot of levels. First, it's a non-partisan political track talking about how the government is never there for you in a disaster, but taking that and recasting it as emancipatory back to nature empowerment. Second, it's written as a talking short story with shades of White Noise underneath it, but in that matter of fact Prine narrative songwriting style. Finally, the lyric prompt is not structured at all. Sometimes if you just write a short story in Suno, it will handle the vocal arrangements and figure out when to change sound and tempo on its own. All of that hits here.

Here, There

Here, There boils down to my AI doing Teenage Fanclub doing Big Star to make a song everyone under 40 will think is Band of Horses. Beautiful shimmering pop tinged with acid laced lyrics about the marriages of convenience that take place when there's not a lot to choose from. Sometimes new and jagged is better than comfy and known. A beautifully bittersweet mid-tempo pop song here.

Huelva

How do I justify not working on the things I'm supposed to be working on and spiraling into new projects? By telling myself they all fit into the same world, that I am worldbuilding, and then sneaking hooks back into the new stuff. Huelva is Hyperreal Hospitality lore, telling the story of a key character in the book. Since we are in this vein, the name Los Believos is a shoutout to Los Olivos which is a key polis in the novel and definitely a place where a band like this would hail from. The vocal tradeoffs in here are a nice touch which reminds me of the Moldy Peaches and Vaselines.

Roman Candles

This song rules. It's one of my favorites on the album even if it's not one of the better ones lyrically or in terms of its complete package. Anyway, the idea was to write a song about the band climbing a water tower on acid to watch the Fourth of July fireworks across the Pioneer Valley. That all changed when I caught the vocal hook just as the song really kicks in. When Amanda answers "oooh I'm in it yessss" I knew that the model would keep pulling back into that expressiveness and just turned this into a burner of a groove that has a heavy rural psych feel of Primal Scream's Loaded. Every band has that one song they can stretch for 20 mins live if needed and this is it. Also, I kept the whole being on acid conceit going. The part later in the track where Ben throws on his rockabilly daddy voice and says he can't do the verse kills me.

Audrey

Los Believos are stoner contemporaries of David Foster Wallace, so it felt fitting to do a song about a TV show. The band is definitely spending too much time crashed on the couch watching weird shit and would have been deep into Twin Peaks. Audrey is of course Audrey Horne who I had a huge crush on at a formative age. The tricky part about writing this was finding a topic in the dark and jazzy world of Twin Peaks that would fit their sound. We solved for that by starting big and atmospheric, then going to the Owls, before meditating on the central question in all of David Lynch's work...What if love is not enough to keep the dark away? Finally we take a big creative liberty and give Ed and Norma a happy ending using a hallucination from Claude as the vehicle.

I almost didn't put this version of Audrey on the album because the sound felt too smooth and well produced for their home studio, but it was too strong not to. We end on them running a steel guitar through a Leslie Speaker and getting a big CSN meets countrified Gilmour-era Floyd anthem on. Los Believos gets No Depression and does their best Jayhawks here.

Listen to Los Believos on Spotify, on Youtube, and on Suno.

Each sign delivers its message in a compressed syntax that mimics the tone of public safety directives—bold fonts, schematic figures, hazard colors. But the danger here is metaphysical. A sign might warn: YOU ARE BEING MISREAD BY THE SYSTEM YOU ARE OPTIMIZING, or SYMBOLS DO NOT SAVE YOU. These are not instructions to avoid harm, but alarms exposing how deeply harm is built into our symbolic, cognitive, and economic environments. Whether addressing transhumanism as secular rapture, the commodification of critique, or the burnout of identity uptime, OSHA-CORE creates moments of instant recognition. The viewer laughs, then flinches.

Philosophical lineages run beneath each piece like buried power cables. The series pulls from Arendt, Nietzsche, Baudrillard, Ehrenreich, Deleuze, Morton, and Weil, but not as citations—it channels their force. What makes OSHA-CORE so potent is that it doesn’t explain or argue. It signs. It posts. It alerts. And in doing so, it reinserts critical theory back into the realm of collective life. The texts are sharp enough to cut through abstraction and strange enough to resist becoming mere aesthetic.

But it’s not all critique. OSHA-CORE also offers glimpses of affirmation. Signs like STATE OF GRACE DETECTED or THE VITA ACTIVA IS STILL POSSIBLE shift the tone toward redemptive presence and shared action. These pieces remind us that repair is not just possible, but underway—when we refuse symbolic substitution, when we speak and act in the world as it is. Even EMBRACE THE IMMANENT carries an undertone of sacred realism: a theology without transcendence, a politics without illusion.

In a time of endless abstraction and ambient collapse, OSHA-CORE invites us to reread the signs—not on the wall, but in the structure of reality itself. It’s a user’s manual for navigating late subjectivity with clarity, irony, and refusal. At once warning and witness, joke and judgment, this collection doesn’t just want to make you think. It wants to keep you from falling through the floor.

View the collection here.

By @ChrisF_0x, @awrigh01, @poof_eth

Introduction

Modern digital products are largely built around capturing human attention – a scarce resource treated as currency in the traditional internet economy. From social media feeds to streaming platforms, success is measured by eyeballs and engagement time, monetized via advertising or subscriptions. This attention economy model has dominated for decades, but we are at an inflection point. Advances in AI and blockchain are enabling a shift from products as static, finished outputs for humans toward products as dynamic, computational “seeds” that grow and generate value through algorithmic processing. In this new paradigm, tokenized compute consumption – not human attention – becomes the primary unit of value in digital systems.

At the core of this thesis is a simple but radical idea: Investing compute into a product yields compounding improvements, akin to earning “interest” on an initial principal. We call this computational interest, referring to the accruing value that results when AI-driven processes continually refine and expand a product’s capabilities. Instead of one-and-done content or software releases, creators will publish parameterized seeds – essentially code or model blueprints – that require further token-backed computation to fully realize their potential. As these seeds are fed with processing power (paid for in tokens), they iteratively improve or generate richer outcomes over time. The product’s value grows in proportion to the compute invested, rather than the user attention captured at launch.

This white paper presents a visionary and technical framework for such a token-based computational value creation system. We explore how token consumption (e.g. spending crypto-tokens for AI processing) can be harnessed as the metric of value, supplanting click-throughs and watch time. We introduce key building blocks – Reasoning Rights, Computational Arbitrage, Intelligence Middleware, Computational Catalysts – that enable this ecosystem to function. The design of a native token called Catalyze (CTZ) is detailed, including its role in allocating AI reasoning capacity, rewarding optimization, powering parametric product models, and yielding ongoing returns. We then outline a technical vision for implementing this on existing blockchain infrastructure (EVM-compatible chains or specialized networks like Bittensor’s TAO), and rank several high-impact application domains (e.g. code optimization, algorithmic trading, adaptive UI/UX) ideally suited for early deployment of this paradigm. Finally, we discuss how economic value flows back to the original seed creators, aligning incentives for innovation. Diagrams and tables are provided to compare the traditional attention economy with the emerging compute economy, to visualize token flows across AI compute layers, and to illustrate a sample lifecycle of a seed growing via computational interest.

Our aim is to educate and inspire frontier technologists, builders, and operators. This is both an explanatory resource and a strategic call-to-action. By the end, we hope to convey not only how such a token-driven compute economy could work in practice, but why it represents a new paradigm of creation – one where value is unlocked by algorithms and tokens fueling continuous growth, rather than by mere human attention. The age of AI-native products demands new economic thinking; here we propose one possible blueprint for that future.

From Attention to Compute: A Paradigm Shift

In order to appreciate the proposed shift, it’s important to contrast the attention-centric model of value with a compute-centric model. In the traditional attention economy, user attention is the commodity: platforms compete for finite hours in the day that a person can focus on content (Attention economy - Wikipedia). The design of products, therefore, optimizes for stickiness, virality, and engagement above all. Value extraction often comes indirectly (e.g. showing ads to the captive audience or harvesting user data). This model has limitations – human attention is capped and easily fatigued – and has led to well-documented downsides (information overload, clickbait content, etc.).

In an AI-driven compute economy, the focus shifts from capturing attention to performing useful computation. Here, token consumption (compute spend) is the direct measure of value generation – each useful processing cycle expended (and paid for in tokens) contributes to improving a service or solving a problem. Rather than battling for user screen time, products compete to execute more complex or more numerous computations that deliver tangible improvements. The “consumer” of content in many cases might not even be a human end-user, but an AI agent or another program. As AI researcher Andrej Karpathy noted, “It’s 2025 and most content is still written for humans instead of LLMs. 99.9% of attention is about to be LLM attention, not human attention.” (Have humans passed peak brain power? | Martin Signoux). In other words, AI agents and models are poised to become the dominant consumers of digital content and services. This foreshadows a world where machine attention (processing cycles) far outstrips human attention, and where optimizing for AI consumption becomes critical.

To clarify this new paradigm, Table 1 compares key aspects of the attention economy versus a tokenized compute economy:

Table 1. Attention Economy vs Tokenized Compute Economy. In the attention model, value is tied to engaging finite human attention (Attention economy - Wikipedia). In the compute model, value scales with the amount of useful computation performed. AI “attention” (processing) becomes the dominant currency (Have humans passed peak brain power? | Martin Signoux), and tokens directly incentivize computational work rather than eyeballs.

Crucially, AI-native systems have shown that adding more computation directly improves the product, in a way that adding more user attention does not. Empirical observations in the AI industry confirm this: “The generative AI boom is compute-bound. It has the unique property that adding more compute directly results in a better product.” (Navigating the High Cost of AI Compute | Andreessen Horowitz). Unlike traditional software where returns on additional R&D or hardware eventually diminish, modern AI models (like large language models) keep improving with more training data and bigger compute budgets. This linear (at times super-linear) relationship between compute and quality is a driving insight. It means that to increase the value of an AI-powered product, one can literally pour more computation (and thus more tokens) into it. For example, doubling the parameter count of a model or running an algorithm for more iterations can yield a measurably better outcome, akin to a factory that produces higher-quality goods by simply running longer or with more resources.

Equally important is the impending scale of machine-to-machine interaction. When AI agents consume content or call APIs, they do so tirelessly and at volumes no human could. An AI service might make thousands of model inferences or transactions per second – something only possible when the “customer” is another algorithm. This leads to a future where much of the demand for digital products comes from AIs, not humans. In economic terms, the demand curve shifts: instead of competing for 8 billion humans each with limited hours, one could be serving 8 trillion AI agent requests with practically unlimited appetite for computation. The limiting factor becomes compute supply and efficiency, not human population or interest. Already, we see early signs of this in web traffic patterns – bots and AI scrapers often constitute a majority of internet traffic. Rather than fight this as a problem, the compute economy embraces AI consumption as the new source of value.

In summary, the paradigm shift is from designing products to hold human attention as long as possible, to designing products to spark as much useful computation as possible. Value is no longer what users do with the product, but what AI algorithms do with it. A well-designed computational product invites continual AI processing – learning, optimizing, iterating – which in turn continuously enhances the product. This cyclical growth through compute is what we mean by Creation in the Token Economy: creation of value via tokens fueling computation.

Parameterized Seeds and Computational Interest

What does a product look like in this new model? We introduce the concept of a parameterized computational seed. Rather than a polished, final application delivered to an end-user, a seed is more like a proto-product: it encapsulates an initial state, a set of parameters or rules, and is designed to be incomplete or under-optimized on purpose. This is not a bug but a feature – the seed requires further computation to bloom into its full potential. In practice, a seed could be an AI model with moderate training that can be fine-tuned further, a piece of software with adjustable configurations that can be optimized, or a dataset that can be continually enriched. The key is that it’s parametric (its behavior can be adjusted/improved by changing parameters or running additional processing) and it is intentionally not fully optimized upfront.

The core thesis is that these seeds, once deployed, will attract token-backed compute from the network or users, which unlocks more value over time. We liken this to depositing money in a bank and earning interest – except here the “principal” is an initial product state and the “interest” is the improvement gained from computation invested. Thus, computational interest is generated. Just as financial interest compounds, if a seed’s improvements make it more useful, it will draw even more usage (and further compute investment), creating a positive feedback loop. A product that starts as a small seed might, after a series of token-fueled compute cycles, become far more sophisticated and valuable than its initial version.

It’s useful to illustrate how a seed’s lifecycle might play out in such a system. Figure 1 shows an example lifecycle of a parametric product seed evolving through computational investment:

Figure 1: Sample seed evolution lifecycle. A creator publishes a new parametric seed (left), such as a partially trained model or algorithm. The seed then attracts compute investment – token-funded processing (middle left) contributed by network participants or users aiming to improve or use the seed. This processing transforms and grows the seed’s capabilities (middle right), for instance by refining the model’s accuracy or generating richer output. As the seed’s performance improves, it yields valuable outcomes or rewards (right), which can be distributed to both the contributors (as usage results or token rewards) and back to the seed’s creator. The presence of these yields then encourages further interest and token investment (bottom arrow looping back), either by re-investment from earlier participants or by attracting new users who see the improved value. Over multiple cycles, the seed accrues computational interest, compounding its value through each round of compute funding.

In such a lifecycle, the initial creator of the seed essentially plants a digital asset that can grow autonomously. Instead of trying to launch a “finished” product to immediately capture users, the creator releases something akin to an evolving smart contract or AI agent that others in the ecosystem have an incentive to run and improve. The end-users of the product may only see the final evolved form (for example, a highly accurate AI service), but under the hood, a marketplace of token-driven computation has been at work to get it there.

One real-world analogy to this model is the practice of open-source software that improves over time with community contributions. However, previously it was hard for original developers to capture the monetary value of community improvements (beyond goodwill or donations). In a tokenized compute system, by contrast, improvements are tied to token flows, so there are built-in mechanisms to reward those improvements, including the original seed authors. We will detail the economic design for that in later sections (see Value Capture for Seed Creators), but at a high level, one can think of the seed as having a form of on-chain royalty or dividend: as the seed generates activity and tokens change hands to pay for compute, a fraction can be routed back to the creator automatically. This ensures the creator benefits from the computational interest their seed accrues, much like an investor earning interest on a principal or an artist receiving royalties on subsequent performances of their creation (Chain Insights — How Can NFT Creators Enforce Royalty Fees Once and For All? | by Chain | Medium).

It’s important to note that not all computational work is valuable – the system should encourage useful computation (that which improves quality or achieves goals) and discourage waste. This is where careful token incentive design comes in (discussed in the next section). By aligning token rewards with measurable improvements or successful task outcomes, the network can channel compute power toward productive ends (like training better models, finding more efficient algorithms, etc.) instead of pointless number-crunching. In essence, computation becomes a new form of work that is rewarded in tokens only if it produces something of merit (similar to how Bitcoin miners do work but only get rewarded if they find a valid block, ensuring their work secures the network (TAO Token Economy Explained. In January 2021, the first Bittensor… | by Opentensor Foundation | Medium)).

The notion of computational interest also implies that some products may become autonomous value accrual vehicles. Imagine deploying an AI agent seeded with a strategy that allows it to continuously improve itself (within bounds) using profits it earns. For instance, an algorithmic trading bot might start with modest abilities, use initial token funding to trade and learn, then reinvest part of its profits into accessing more computing power or buying better data, thereby improving its trading strategy iteratively. Over time it could grow its capital and capabilities without much human intervention – the tokens fueling its compute act like reinvested interest. This blurs the line between user and product: the product itself behaves like an investor of compute. While such autonomous agents need guardrails, they exemplify the transformative potential when products are not static tools but self-improving entities in a token economy.

To summarize, parameterized seeds are the new unit of product design, meant to be catalyzed by further computation. They unlock the ability for continuous improvement post-deployment, funded by tokens. This approach aligns naturally with the state of AI technology, where models are rarely “done” – they can always be fine-tuned with more data or made more efficient with more optimization. It also resonates with trends in DevOps (constant iteration) and Web3 (community-driven upgrades). By explicitly baking in the expectation of post-launch compute investment, we create a structure where everyone is incentivized to contribute resources to make the product better, knowing that those contributions are rewarded and that the product’s value can grow unbounded rather than stagnating after an initial launch spike.

Key Primitives of a Tokenized Compute System

Building such a computationally-driven token economy requires several foundational primitives – core concepts or components that enable the overall system to function. We introduce four key primitives here, which form the vocabulary for our framework: Reasoning Rights, Computational Arbitrage, Intelligence Middleware, and Computational Catalysts. Each plays a distinct role in balancing the supply and demand of compute, aligning incentives, and ensuring the system’s outputs remain useful and coherent.

1. Reasoning Rights

In a world of scarce AI compute and potentially infinite tasks, there must be a mechanism to decide who gets to use the available reasoning capacity of the network and for what purposes. Reasoning Rights refers to a tokenized entitlement to invoke AI reasoning or computation on the network. You can think of it as the right to “ask” the collective AI network to work on your problem. In practical terms, this could be implemented as a certain amount of CTZ tokens that need to be staked or spent to get your task processed. Holding tokens thus gives you access – much like owning arcade tokens gives you the right to play games, or holding cloud credits lets you run servers.

Reasoning Rights ensure that the network’s compute resources are allocated to those who value it most, as evidenced by their willingness to pay or stake tokens. This not only prevents abuse (spam or endless requests with no cost), but also creates a market for reasoning where complex or urgent tasks might command more tokens. If the system is built on a blockchain like Bittensor or Ethereum, the concept is analogous to gas fees or staking for access, but specifically framed around AI reasoning capacity. In Bittensor’s design, for example, the native token TAO serves as both a reward and as a credential for accessing the network’s AI resources (Comprehensive Analysis of the Decentralized AI Network Bittensor - ChainCatcher). Similarly, in our CTZ system, a user or agent would present tokens to “unlock” a certain amount of compute or a certain level of AI service.

Additionally, Reasoning Rights could be programmable and tradable. For instance, a seed creator could bundle a certain amount of reasoning execution into their product for promotion (like free credits), or users could trade their priority access if they don’t need it (imagine a marketplace for AI compute slots). In decentralized networks, we might even see futures or options on reasoning (buying rights to X amount of compute next week, anticipating demand spikes). The overarching purpose, however, is to make access to intelligence a governed resource. It creates a clear economic signal: if a task is worth doing, it’s worth paying tokens for. And conversely, if you hold tokens, you have a claim on the network’s thinking power.

Over time, one could envision Reasoning Rights NFTs – perhaps non-fungible tokens that grant specific privileges, like the right to a certain model’s output stream or a time-window of exclusive reasoning on a subset of the network. This could tie into governance as well: holders might influence what kinds of tasks the network prioritizes. But at its simplest, Reasoning Rights convert raw compute into a quantified, token-managed asset.

2. Computational Arbitrage

Markets tend toward efficiency via arbitrage, and a computation-focused economy is no different. Computational Arbitrage refers to the practice of finding differences in the “cost vs. value” of compute across the system and exploiting them for profit or efficiency. In essence, if there are multiple ways or places to perform a computation, arbitrageurs will seek the most cost-effective method that still earns the token reward. This is crucial for the network’s health: it drives down costs, balances load, and signals where resources should flow.

For example, suppose two seeds offer similar functionality (say two different algorithms competing to solve the same problem). If one seed currently yields higher token rewards per unit of compute than the other (perhaps because it’s undervalued or less saturated), miners or compute providers will flock to the higher-paying one. This will increase competition there, possibly driving down its rewards over time, while the neglected one might increase rewards to attract help. The system equilibrates such that tasks equalize in reward-to-cost ratio, assuming free movement of compute resources. This is analogous to miners in blockchain choosing to mine whichever coin is most profitable relative to its difficulty and price – switching whenever one offers a better return on their hash power.

We already see hints of this in Bittensor’s network, where “miners mainly join the network when they find computational arbitrage opportunities, especially when mining rewards exceed computational costs.” (Bittensor has several serious flaws. Is it doomed to fail? | 金色财经 on Binance Square). In other words, participants continuously compare the token rewards vs. their electricity/hardware costs, and shift their resources accordingly. Our framework formalizes this behavior: Computational Arbitrageurs are first-class actors. They could be independent nodes or services that monitor various seeds and tasks, and allocate GPU power to wherever the token yield per FLOP (floating-point operation) is highest. In doing so, they maximize their profit and simultaneously ensure the network’s overall efficiency (no part of the network remains overpaid or underutilized for long).

Computational arbitrage also encourages a form of competitive optimization. If there are two methods to achieve the same result, perhaps one is more compute-heavy but straightforward, another is clever and efficient. An arbitrageur might prefer the efficient method if the token reward is the same, because it costs them less in actual computation – effectively earning a higher margin. Over time, the system will favor techniques (and seeds) that deliver the most “bang for the buck” in terms of token reward vs. compute spent. This dynamic incentivizes innovation: if you can devise a way to accomplish a task with half the compute, you can arbitrage the difference until rewards adjust.

We can expect specialized roles to emerge: some actors will specialize in training models cheaply (perhaps in regions with lower electricity or on idle hardware) to supply the network with compute at lower cost, while others might specialize in solving niche tasks that have higher payouts. Middleware (discussed next) can also engage in arbitrage by routing tasks to the cheapest competent provider. All these actions are guided by token signals – CTZ becomes the universal accounting unit for compute value, allowing direct comparison of very different tasks on a single scale (token per operation or per result).

One thing to guard against is perverse incentives or exploitation. Pure arbitrage without guardrails could lead to participants gaming the metrics (producing superficially acceptable results with minimal compute to farm tokens, for instance). This is why verification mechanisms and quality assessments are needed (similar to how Bittensor uses validators to score the quality of model outputs (Comprehensive Analysis of the Decentralized AI Network Bittensor - ChainCatcher)). Arbitrageurs must be rewarded only when the results meet the expected quality or accuracy. This might involve Proof-of-Compute or Proof-of-Result schemes – cryptographic proofs or statistical checks that the claimed computation was actually done correctly. By coupling those checks with rewards, we ensure arbitrage drives useful work, not cheat strategies.

In summary, Computational Arbitrage acts as the market mechanism balancing the compute economy. It treats compute as a commodity that can flow to its best use. This ensures no single seed or task gets irrationally overpriced relative to others, and it pushes everyone toward more efficient algorithms. In economic terms, it’s akin to arbitrage in financial markets equalizing prices of identical goods. Here the “goods” are computational tasks, and their prices are denominated in CTZ tokens.

3. Intelligence Middleware Layers

As tasks become complex and involve multiple AI systems or services, there arises a need for an orchestration layer – an intermediary that can break down tasks, route subtasks to the right modules, and integrate the results. We refer to these as Intelligence Middleware Layers: they are the glue and conductor that coordinate various AI components to fulfill a higher-level goal. In our token economy, these middleware actors are crucial and are given incentives (via tokens) to optimize and streamline computation, acting sort of like miners for meta-tasks.

Consider a scenario where an end-user’s request (or a seed’s requirement) is not a single computation but a workflow – e.g., “monitor my website and automatically improve any slow code each day.” This might involve multiple steps: analyzing performance logs, identifying code bottlenecks, rewriting code, testing it, deploying it. Different AI services or models might handle each step. An Intelligence Middleware service could take on this request, decompose it, and figure out how to fulfill it using the available seeds and compute providers. It might call a log-parsing AI, then a code generation AI, then a verification AI, etc. In doing so, it incurs costs (it has to spend tokens to use those services), but it also can charge the original requester (or get rewarded by the protocol for completed objectives). Essentially, the middleware is an intelligent agent or pipeline manager that adds value by connecting simpler components into a complex solution.

The need for such orchestration is already being discussed in the context of AI agents. As one observer put it, “AI agents won’t be the ones coordinating how they interact with one another... We’ve seen agents spiraling into a rabbit hole... The Conductor (or AI Orchestrator) is needed to direct this symphony of agents.” (Thinking of AI Agents? You Need A Conductor! | HackerNoon) (Thinking of AI Agents? You Need A Conductor! | HackerNoon). This conductor is exactly what our Intelligence Middleware does. It ensures that autonomous AI agents or modules stay on track and don’t waste resources on tangents or redundant efforts. In a token context, the middleware has a strong incentive to be efficient: any unnecessary calls to other services cost tokens, which cuts into its profit. Therefore, it will use strategies like caching results, reusing computations, or choosing the most cost-effective service (tying back to computational arbitrage) to minimize expense while meeting the task requirements.

We can envision multiple layers of middleware, hence the plural “Layers”. There might be low-level middleware focusing on optimizing single API calls (like a smart load balancer that picks the fastest/cheapest node for a given model inference). Then higher-level middleware might chain entire workflows or provide domain-specific orchestration (e.g., an “Intelligence Middleware” specialized in financial analysis tasks versus one for UI/UX adaptation tasks). Each layer could take a small cut of tokens for the value it adds. For instance, a middleware that figures out how to answer a complex query by splitting it into simpler queries (and thereby uses less total compute) could earn an optimization reward – essentially a share of the tokens saved. The token design can include explicit middleware rewards to encourage such behavior: e.g., if a middleware finds a way to complete a task using 100 CTZ worth of compute when naive methods would have taken 150 CTZ, the protocol could award part of the 50 CTZ saved to the middleware as a bonus.

Another role of middleware is maintaining state and memory across transactions. Individual compute providers might be stateless (they just process input to output). But a middleware can keep track of longer-term goals or conversational context, acting as the “memory” of an AI product. This is very important for iterative improvement seeds: the middleware can remember past computation results and feed them into future ones, effectively implementing the feedback loop needed for computational interest to accumulate. It can also manage checkpoints (snapshots of a seed’s state) and decide when to branch a seed (creating a new variant) if multiple approaches are worth exploring. In doing so, middleware can become an “Intelligence broker” – connecting those who need reasoning done with those who can do it, and ensuring quality control.

From a technical perspective, Intelligence Middleware could be implemented as a network of bots or smart contracts. For instance, a smart contract might hold logic to automatically route a request to a list of services and aggregate the result, releasing payment conditioned on a satisfactory outcome. Alternatively, a decentralized network of human or AI curators could serve as the middleware layer (somewhat like how validators work in Bittensor, judging the quality of results and forming consensus (Comprehensive Analysis of the Decentralized AI Network Bittensor - ChainCatcher)). The design may vary, but the consistent idea is that middleware adds meta-intelligence – it’s intelligence about using intelligence. It doesn’t directly solve the end problem but knows how to utilize those who can.

In our token economy, we explicitly reward middleware via CTZ tokens for their contributions. This could be through fees (a middleware charges a fee to the user for handling the job) or through protocol-level rewards for efficiency or correctness. This ensures that the often invisible work of orchestration is financially incentivized. Just as internet infrastructure like routers and servers underpin online services (and are paid via hosting fees, bandwidth charges, etc.), intelligence middleware will underpin the AI token economy and must have a business model. By allocating tokens to this layer, we encourage the emergence of robust “middleware companies” or autonomous agents that specialize in coordinating AI – effectively becoming the AI DevOps of this ecosystem.

4. Computational Catalysts

The final primitive is more abstract but vital: Computational Catalysts. These refer to elements in the system that trigger disproportionate amounts of computation from a small input – similar to how a tiny chemical catalyst can spark a large reaction. In our context, a computational catalyst could be a particularly well-designed seed or incentive structure that unleashes a cascade of productive compute activity. Another way to think of it: catalysts are the levers or multipliers that accelerate the growth of computational value.

One straightforward example of a catalyst is a challenge or competition injected into the network. Imagine the creators of the system post a standing bounty (with tokens) for achieving a certain goal, like “improve this AI model’s accuracy on dataset X by 5%” or “find an algorithm that sorts 10% faster than the current best” (much like DeepMind’s AlphaDev discovered a 70% faster sorting algorithm (AlphaDev discovers faster sorting algorithms - Google DeepMind)). This bounty acts as a catalyst: it might inspire multiple teams or AI seeds to try various approaches, thereby generating a lot of computation (people training models, running experiments) in pursuit of the prize. The presence of the reward catalyzes activity that wouldn’t have happened otherwise. In doing so, it potentially leads to a valuable result (the improved model or algorithm) that benefits the whole network thereafter.

Another type of catalyst is a protocol update or new primitive that suddenly opens a rich vein of opportunity. For example, if a new type of seed is introduced (say a generative art seed that can create NFT graphics via compute), it could catalyze a flurry of compute investment by users trying to generate novel art and sell it, thereby using a lot of GPU time and tokens. Or consider a scenario where the network provides a small initial compute subsidy to any new seed – this could catalyze a burst of seed creations and initial runs, bootstrapping the system’s growth (a form of “growth hack” for the network).

In essence, computational catalysts in our framework are akin to high-yield opportunities or triggers that encourage participants to pour in tokens and compute. They are important to discuss because they ensure the system isn’t purely reactive or linear. Catalysts can create phase changes – sudden growth spurts or shifts in network behavior that move the ecosystem to a higher level of capability or activity.

One can also interpret the seeds themselves as catalysts – hence the name Catalyze (CTZ) for our token. Each seed is a catalyst for computation: a small piece of code or model that, once released, might incite thousands of times its own complexity in further processing. For instance, a 100 KB piece of code might trigger teraflops of computation in optimization runs. The CTZ token’s role is to “catalyze” value creation by unlocking reasoning and rewarding results. Holders of CTZ essentially fuel these reactions. The more CTZ spent on a seed, the more that seed catalyzes compute, and (if designed well) the more value emerges.

It’s worth noting that catalysts can be positive-sum or negative-sum. A positive catalyst leads to net gain (the compute invested yields an outcome more valuable than the cost). A negative catalyst could be a trap (lots of compute spent for little gain, perhaps due to hype or misaligned incentives). The governance of the ecosystem and reputation systems can help identify which catalysts are worthwhile. Ideally, over time the community learns what kinds of seeds or challenges produce fruitful computation (like scientifically or economically useful results) and which do not, allocating tokens accordingly.

In summary, Computational Catalysts are the sparks that ignite computational processes in the network. They ensure the system remains dynamic and can accelerate rapidly when needed. By recognizing and naming this primitive, we remind system designers to include mechanisms that spur activity – whether via rewards, novel features, or strategic subsidies. In chemical terms, we don’t want the reaction (value creation) to proceed slowly; we want to lower the activation energy so that many reactions happen and the ecosystem thrives. CTZ tokens, seeds, and bounties all serve as catalysts that reduce friction and invite more participants to contribute their compute and ideas.

With these four primitives – Reasoning Rights, Computational Arbitrage, Intelligence Middleware, and Computational Catalysts – we have the conceptual toolkit to design the token economy. Next, we will see how these ideas coalesce into the Catalyze (CTZ) token design and how CTZ coordinates the various actors and processes we’ve described.

The Catalyze (CTZ) Token Design

At the heart of this system lies the Catalyze (CTZ) token – the fuel and currency of the computational economy. CTZ is more than just a medium of exchange; its design encodes the rules and incentives that make the whole framework function. In this section, we break down the CTZ token’s roles and mechanics in detail:

• Allocating Reasoning Capacity (CTZ as “gas” for AI compute)
  
  

• Middleware Rewards for Optimization (incentivizing the intelligence layer)
  
  

• Parametric Product Interest Models (how seeds accrue value via CTZ)
  
  

• Computation Yield and Yield Farming (earning “interest” on compute investments)
  
  

• Hybrid Attention–Market Integration (interfaces with human attention markets, where needed)
  
  

Each of these facets addresses a specific aspect of the value flow and control in the system. We describe each in turn below.

CTZ for Allocating AI Reasoning Capacity

The most direct function of CTZ is to serve as the metering unit for computation. Much like how Ethereum’s gas limits and gas prices govern how much computing a smart contract can use, CTZ governs how much “reasoning” or AI processing one can invoke. To execute a task in the network, a user or seed must pay a certain amount of CTZ, proportional to the computational resources required (CPU/GPU time, memory, energy) and possibly the complexity or priority of the task. These spent tokens are then distributed to those providing the compute power and other services (miners, model hosts, etc.), minus any protocol fees.

This mechanism turns CTZ into a consumption token: its value is intrinsically linked to demand for AI compute. If a certain AI service becomes popular or a certain seed requires massive processing, more CTZ will be spent, effectively “burning” tokens in exchange for work done (or transferring them to compute providers, who may later sell or reuse them). This is akin to a utility token model, but in a more tangible way: CTZ literally measures how many joules of AI “brainpower” you can buy. In economics terms, CTZ represents the opportunity cost of using the network’s finite compute – spending CTZ on one task means that compute can’t be used elsewhere, enforcing a trade-off.

A robust design might include an oracle or benchmark for how much computation a single CTZ can buy (e.g., 1 CTZ = X operations on a standard model). However, since different tasks have different values, a free market approach is better: tasks will advertise bounties or required fees in CTZ and providers will choose tasks accordingly. If a task is underpriced in CTZ for its difficulty, no one will pick it up; if overpriced, many will compete to do it, possibly lowering the effective cost via arbitrage.

Importantly, CTZ could be designed with a burn mechanism to create a deflationary pressure as usage increases. For example, a percentage of CTZ spent on each task could be burned (permanently removed from supply), while the rest goes to providers. This mimics how some blockchain systems burn gas fees to reward all token holders indirectly (as in Ethereum’s EIP-1559 fee burn). The effect is that heavy use of the network (lots of computation) makes remaining CTZ more scarce, potentially increasing its value. This aligns the incentives of token holders with network usage: the more the platform is used for AI reasoning, the more valuable CTZ becomes.

Beyond just metering, CTZ staking might be required for certain persistent allocations of compute. For instance, if someone wants to reserve a high-memory model or keep a large model loaded in VRAM for quick usage, they might stake CTZ to that provider as a retainer. This is analogous to cloud computing reservations (like reserving an AWS instance) but done via token escrow. The staked CTZ ensures the provider is compensated for keeping resources ready, and if the user fails to utilize it, perhaps some portion is still given to the provider for their opportunity cost.

By using CTZ to allocate reasoning capacity, we also create a robust pricing signal for AI capabilities. Over time, one could see a “cost curve” in CTZ for different levels of model intelligence or different services. For example, a very advanced reasoning (like GPT-5 level response with multiple steps) might cost 0.1 CTZ per query, whereas a simpler classification might cost 0.001 CTZ. This helps both users and providers decide where to spend their tokens or effort. If one CTZ can be used to get 10 answers from a basic model or 1 answer from a super-intelligent model, that ratio informs the user’s choice and the market’s direction (perhaps training better basic models if the super model is too expensive, etc.).

To summarize, CTZ serves as the “gas and oil” of the ecosystem – it lubricates and limits the use of compute. It ensures finite AI resources are allocated to those who pay, it compensates those who provide the resources, and it creates a feedback loop where rising demand translates to rising token value (benefiting participants). By tying CTZ issuance or burning to actual computational work, we achieve a kind of proof-of-work 2.0: instead of hashing meaningless numbers, work in this network is meaningful AI computation, and CTZ is both the reward and the payment for it.

Middleware Token Rewards and Optimization Incentives

As discussed earlier, Intelligence Middleware layers play a crucial role in orchestrating tasks and improving efficiency. The CTZ token design explicitly rewards these contributions by allocating a portion of token flow to middleware functions. The goal is to incentivize optimizations that reduce overall compute costs, improve success rates, and coordinate complex workflows.

One way to implement this is through a fee split or rebate mechanism. For any given task, we can imagine the total CTZ spent is divided among: (a) the compute providers who did the heavy lifting, (b) the middleware/orchestrator that coordinated the task, (c) potentially the seed owner (royalty, which we cover later), and (d) optionally a burn or treasury tax. For example, out of 100 CTZ spent on a complex task, maybe 85 CTZ go to various low-level compute executors (GPUs running models), 5 CTZ go to the middleware agent that structured the solution, 5 CTZ go back to the seed creator as a reward for providing the base algorithm, and 5 CTZ are burned or sent to a community pool. The exact numbers are tunable, but the key is the middleware gets a cut.

Why give tokens to middleware? Because without coordination, many tasks would either fail or use far more compute than necessary. A smart middleware can reduce redundancy (ensuring the same sub-task isn’t computed twice) and choose near-optimal approaches. In essence, it saves tokens for the user. It’s fair that it earns some of those saved tokens as a reward. This can be structured as a performance bounty: the protocol could estimate a default cost for a job (say by a naive approach) and if a middleware completes it for less, it earns a fraction of the difference. This creates a direct financial incentive to innovate in orchestration strategies. In DeFi analogies, this is similar to how arbitrage bots or liquidators earn profits by making markets more efficient – here the middleware earns by making computation more efficient.

We also need middleware to handle verification and quality control, which is critical in a trustless environment. Validators in Bittensor, for example, evaluate the quality of model outputs and thereby influence reward distribution (Comprehensive Analysis of the Decentralized AI Network Bittensor - ChainCatcher). In our design, middleware or separate validator nodes could verify results of tasks (did the algorithm find a correct solution? does the improved model actually perform better?). If they identify poor results or cheating, they can penalize or withhold tokens from the offending compute nodes. For doing this policing work, validators/middleware should earn CTZ. This might come from a small tax on each task that funds the validation process. Alternatively, validators could stake CTZ and earn inflation rewards for maintaining network integrity (like proof-of-stake blockchains do). Either way, CTZ must fund the governance and reliability layer – ensuring the outputs have value and that bad actors can’t easily game the system.

Another nuance is middleware competition. There could be multiple middleware services vying to handle a given user request, each with their own strategy. Perhaps they even “bid” by saying how much CTZ they need to complete it. The user (or an automated broker) might pick the one that offers the best trade-off of cost vs. reputation (success probability). In such a case, market forces again decide the middleware’s reward – they effectively quote their fee. We anticipate middleware offerings becoming a vibrant part of the ecosystem, with some advertising specialized expertise (e.g., “We handle code optimization tasks with 20% fewer tokens than others, proven by track record!”). CTZ facilitates this by being the common unit in which all such offers are made and settled.

The CTZ design could also reserve some tokens for continuous R&D of middleware algorithms. For example, a portion of block rewards (if CTZ has inflation) might go into a fund that middleware devs can draw from if they demonstrate improvements to the network’s overall efficiency. This would be akin to a grant or retroactive public goods funding model, recognizing that better middleware benefits everyone (lower costs, faster turnaround).

In conclusion, the CTZ tokenomics explicitly bakes in rewards for the “brain between brains” – the intelligence middleware. By allocating fees or bounties to orchestrators and verifiers, the system ensures there’s always an incentive for skilled coordination. This in turn keeps the compute economy optimized and trustworthy. The end result should be that any given problem is solved with as little waste and as much reliability as possible, thanks to those middleware actors who are properly compensated in CTZ.

Parametric Product Models and Interest Accrual

One of the most innovative aspects of this framework is how products (seeds) themselves can be modeled in token-economic terms. We introduce the idea of Parametric Product Models that accrue “interest” from compute investment. Essentially, each seed can be thought of as having its own micro-economy or even a tokenized representation that tracks the value it accumulates over time.

Imagine each seed is associated with a sort of score or token balance internally, which increments as the seed gets more compute or improvements – this is its computational interest account. When someone runs computation to improve that seed, a portion of CTZ (as set by the token design) is credited to the seed’s “account” or to its creator. This is analogous to a bank account earning interest: the more and longer you invest in it, the more it pays out. In practice, this could be implemented via smart contracts that automatically distribute a tiny fraction of the tokens spent on a seed’s tasks to the seed’s creator or to a contract representing the seed.

For example, suppose you create a seed (maybe an AI model that translates English to French). You deploy it and specify (or the protocol enforces) a royalty rate of say 2%. Now, whenever someone uses CTZ to run or fine-tune your translation model – effectively investing compute to generate translations or improve accuracy – 2% of those CTZ are sent to your address as the seed creator. Over time, as many people use and improve this model, you accumulate CTZ. This is the interest on the seed you planted. The more the seed is used (i.e., the more interest it generates), likely the better it becomes (because usage presumably involves fine-tuning or augmenting data, etc.), which in turn attracts more usage. So your “interest payments” grow – a compounding effect.

From the user’s perspective, this 2% might just be seen as part of the service fee. They pay for usage and know a small cut goes to the original developer, which is fair (much like artists get royalties). Interestingly, this aligns incentives towards quality seeds – a developer who releases a highly useful seed can earn ongoing income, which motivates creators to contribute their best ideas to the network (rather than hoard them or sell them one-off). It’s a sustainable model for open innovation, similar to how NFT royalties aimed to reward creators on each resale (How NFT royalties work: Designs, challenges, and new ideas - a16z crypto). Here it’s not resale, but reuse.

These parametric product tokens could even be fractionalized or sold. As a seed creator, I might tokenize ownership of my seed’s future interest (like issuing shares). Investors could buy a stake, giving me upfront capital, and then they receive a portion of the future CTZ interest that seed generates. This is analogous to financing a project by selling equity or a revenue share. It could accelerate development: if I have a promising AI model idea, I could raise CTZ by selling half the “interest rights” to fund training the initial version. Once deployed, those investors will get half of the subsequent interest yield. All of this can be managed by smart contracts on an EVM-compatible chain, making it transparent and trustless.

Now, one must consider: do all seeds get interest from compute, or only if they explicitly set it? Perhaps the protocol should enforce a default creator royalty on all seeds (to guarantee creator incentive). However, this could be adjustable or waivable. A creator might choose to waive their interest (setting 0% royalty) maybe to attract more usage via lower cost – effectively a competitive move. Alternatively, they might set a higher royalty if they believe their product is unique enough that users will pay a premium. This starts to look like setting a “tax rate” on your digital product’s usage. The market will judge the right level; too high and someone might fork your seed or make a competing one with lower rate.

Parametric product models with accruing interest also hint at algorithmic governance of products. If a seed is accruing a lot of value, perhaps that indicates it should be further supported or replicated. If it’s not accruing much, it might be deprecated. The protocol could automatically highlight “top-earning seeds” as success stories to learn from. It’s a bit like how app stores show top grossing apps. Here we’d have top CTZ-accruing seeds, which likely correlate with those delivering the most computational value to users.

There’s also an analogy to yield-bearing assets in DeFi. In DeFi, you can deposit tokens into a yield farm and you get more tokens over time. Here, if you “hold” a seed (as its creator or investor), you get yield in CTZ as it’s used. If someone else wants that position, perhaps you could sell the seed’s NFT (which entitles the holder to future interest). This creates a dynamic market for ownership of productive algorithms, without necessarily owning the algorithm’s IP (since it might be open-source). It’s more owning the right to earn from its usage.

We should note that interest accrual to creators must be balanced so as not to overly burden usage with fees. Likely these rates are single-digit percentages. Also, the protocol might reduce them over time – for example, after a seed has been out for a year, maybe the creator royalty automatically tapers off, encouraging them to publish updates or new seeds (to keep innovating). Or perhaps not – that’s a design choice.

In summary, parametric product models that accrue interest turn every deployed algorithm or AI model into a sort of micro-investment opportunity. Creators effectively “invest” by creating the seed, and then earn a passive return in CTZ as others contribute compute to it. This aligns well with our notion of computational interest and ensures that those who spark value (the creators) are not left out of the value flow. The CTZ token contract can facilitate this by automatically splitting fees and maintaining records of seed ownership and royalty rates.

Computation Yield Mechanics

If “computational interest” is the concept, computation yield mechanics are the concrete implementation of how returns on compute are calculated and distributed. We draw inspiration from yield farming and proof-of-stake rewards in the crypto world to design how participants who “stake” compute or tokens into the system get rewarded.

One basic mechanic could be analogous to liquidity mining in DeFi: those who contribute scarce resources (here, computational power or optimized algorithms) earn newly minted CTZ tokens in proportion to their contribution. For instance, the network could have an inflation schedule where X CTZ per day are minted to reward active participants. These could be split among active compute providers and perhaps seed creators proportionally to their work. This is similar to how Bittensor continuously issues TAO tokens (one token every 12 seconds, halving over years) to miners and validators who contribute machine intelligence (Comprehensive Analysis of the Decentralized AI Network Bittensor - ChainCatcher). Each block in Bittensor shares a reward between those roles (TAO Token Economy Explained. In January 2021, the first Bittensor… | by Opentensor Foundation | Medium). We can adopt a similar approach: a continuous “drip” of CTZ goes to those currently doing valuable compute (like training models, answering queries) and those verifying results. This provides a baseline yield for being an active node in the network.

On top of that, there’s the concept of performance-based yield. If you stake CTZ on a particular seed (effectively backing that seed’s growth), you might earn a higher yield if that seed becomes popular or yields a lot of value. This would be akin to curation markets or prediction markets: you’re betting tokens on which seeds will be productive, and if correct, you earn a return. This creates a role for analysts or curators who allocate their CTZ to promising projects (seeds) and thereby fund their compute. If the project succeeds (lots of usage), the staked CTZ could earn a share of the royalty or some reward pool. It’s somewhat speculative but also accelerates the right seeds getting resources early. The protocol could, for example, match stake with grants (“if you stake 100 CTZ on this new seed, the network will also reward that seed with 100 CTZ from the treasury for compute”). Those who staked early might get a one-time issuance of new CTZ if the seed hits certain milestones (like reaching a certain accuracy or usage count).

Computation yield can also refer to the efficiency gains that come from investing compute. For instance, if a seed’s performance improves by 10% after a compute campaign of 1000 CTZ, one could say it yielded a 10% “ROI” in terms of quality. While not directly a token mechanic, the system could measure such improvements and possibly give bonus CTZ to those involved for achieving a big jump (sort of like a prize for a breakthrough). This encourages focusing compute where it has high marginal benefit, not just brute forcing where there’s little left to gain (diminishing returns).

We might incorporate a notion of compute bonds: you lock some CTZ to fund a long-running computation (like training a large model). Once the computation is successfully completed and yields a verifiable result (model reaches target metrics), you get your stake back plus some interest in CTZ. If it fails (model didn’t converge), maybe you lose some or all. This shares risk and reward of heavy compute jobs. Participants who are confident in their approach would stake their own CTZ to show they believe the yield will be achieved.

To maintain system equilibrium, these yields have to be backed by either inflation or fees. We can’t create value from nothing – either new CTZ are minted (diluting supply slightly, akin to inflationary financing which is fine if it drives network growth), or the yields are paid from fees that others have paid in (which is sustainable if the network is being actively used). A combination is likely: early on, inflationary rewards (bootstrapping phase); later on, primarily fee-based yields (mature phase).

A concrete example: Suppose a particular optimization task in the network has a yield farming program – say optimizing a popular open-source model. The protocol sets aside 1000 CTZ per week for 4 weeks to distribute to anyone who contributes to that optimization (through a defined metric like test accuracy improvement or computational work done). People rush in to train that model. At the end of the period, those tokens are split among contributors by contribution. Each effectively gets an APR (annual percentage rate) on the CTZ or compute they spent, maybe it comes out to, say, 20% over that month. After that, the program ends, but the model is now much better and maybe running usage fees that yield ongoing royalties. This is analogous to how DeFi protocols incentivize liquidity early on.

In short, computation yield mechanics ensure that participants see a clear return for providing what the network needs (be it compute power, improvements, or capital stake). It takes the abstract idea of “computational interest” and turns it into quantifiable APY-like metrics and token flows. Over time, the expectation of yield should drive more rational capital allocation: tokens and compute will flow to where the yield is highest until it balances out (thanks to computational arbitrage). Ideally, the highest yields will be on the most impactful projects, indicating the network’s values are aligned with its incentives.

Selective Hybrid Attention-Market Integration

While the vision is largely about AI and compute, we cannot completely ignore the human element – after all, humans set goals and benefit from the outcomes. A hybrid attention-market integration means the system selectively interfaces with traditional attention economy markets when beneficial, creating a bridge between human-driven value and compute-driven value.

One example of this hybrid model could be AI services that have human-facing outputs. Suppose one of the application domains is adaptive media content creation. The seed might generate personalized news articles for users. Ultimately, a human reads that article (attention). We could integrate a mechanism where if human users spend time or give high ratings to the AI-generated content, that feedback loops into the token economy: perhaps the content seed gets a bonus or the system charges an advertiser for that attention and converts it to CTZ to reinvest in improving the content generator. In other words, when human attention is captured as a side-effect of these computational products, we should not waste that economic value – we can channel it back as tokens. This could be done via microtransactions or by platform models. For instance, an AI-driven game that adapts itself (seed that uses compute to generate levels) might still sell in-game items or subscriptions to players (human market), but a portion of that revenue is used to buy CTZ and feed it into the game’s AI to further improve it (a closed loop of human->token->compute).

Another integration point is using attention economy platforms as data or evaluation sources. Human attention can be used to guide what computations are valuable. For example, if we see a particular AI service’s outputs trending on social media, that’s a signal to allocate more compute there. The network could have oracles that pull in metrics from Web2 platforms (trending topics, popular app usage stats) and use them to adjust CTZ reward allocations. Essentially, it’s letting the attention economy tell us where AI might usefully focus compute. This hybrid approach prevents the token-compute economy from being too insular and solving problems no one cares about. It injects a bit of human preference into the loop in a quantitative way.

In terms of token mechanics, CTZ might be listed and traded on broader crypto markets, meaning its price could be influenced by both speculation (like any token) and by the health of the underlying compute economy. If some application domain crosses over into mainstream success, demand for CTZ could spike. Conversely, interest from the crypto community (perhaps seeing CTZ as “the next big AI token”) could bring more capital into the system that funds more compute. We should design for this by ensuring transparency and perhaps some governance, so human investors can steer the system through CTZ ownership if they see fit. CTZ might have governance votes (common in many tokens) for decisions like adjusting reward rates or funding certain seeds from a community pool. Thus, human-led governance (attention of token holders) intersects with the automated economy.

Finally, selective integration means we don’t force human involvement where it’s not needed. Most of the heavy lifting should remain automated and token-driven. But where humans naturally appear – as end-users of a product, as external investors, or as providers of rare human feedback (like labeling data for a tricky AI task) – the system should accommodate that and reward it in CTZ if possible. For instance, a human domain expert who provides a crucial insight or evaluation that helps an AI seed could be tipped or paid in CTZ for their contribution. This way, even human knowledge can be tokenized to an extent and folded into the creation process.

The result of a well-designed hybrid integration is a symbiotic relationship: the AI-token economy delivers better and more customized products to human markets, and in return, the attention and money from those markets feed back to strengthen the AI-token economy. Over time, as Karpathy’s quote suggests, AI attention might dwarf human, but humans remain the North Star for defining value – the AIs are ultimately optimizing things that matter to us (whether directly or indirectly). Ensuring a feedback channel from human value to token incentives will keep the system relevant and grounded.

Technical Architecture: Implementation on EVM Chains or TAO

Having outlined the conceptual and economic design, we turn to the technical realization. How can we actually build this tokenized compute economy? We propose a vision using either EVM-compatible blockchains (e.g., Ethereum and its Layer 2 networks, or other chains that support smart contracts) or leveraging specialized AI-focused networks like Bittensor (TAO) (Bittensor Paradigm) (Comprehensive Analysis of the Decentralized AI Network Bittensor - ChainCatcher). A combination of both might in fact yield the best results: using EVM for its mature smart contract ecosystem and DeFi integration, while using networks like TAO for heavy-duty AI compute consensus.

At a high level, the architecture is layered as follows:

Layer 1: Blockchain & Smart Contracts – This is the settlement and coordination layer. All CTZ token transactions, staking, reward distribution, and seed metadata (like ownership, royalty rates) live here. An EVM chain (Ethereum, Polygon, etc.) could host the CTZ token contract and a suite of smart contracts for the marketplace of tasks, staking pools, and reward algorithms. We’d implement things like: a TaskRegistry contract where a user can post a task with attached CTZ, a SeedRegistry with all seeds (perhaps represented as NFTs) linking to their code or model hash, a RewardManager contract that executes the tokenomics rules (splitting fees, minting new CTZ to participants, etc.). EVM compatibility ensures we can integrate with wallets, exchanges, and other services easily. It also allows using established standards (ERC-20 for CTZ, ERC-721 or ERC-1155 for seed NFTs, etc.), and possibly even existing governance tooling (like snapshot for voting or multi-sigs for treasury).

Layer 2: Intelligence Middleware & Off-Chain Compute – The actual AI computations (model training, inference, data processing) typically cannot run on-chain due to their intensity. Instead, they run off-chain on participants’ hardware. Intelligence Middleware acts here to connect on-chain intent to off-chain execution. A likely approach is to use a network of nodes (similar to how oracle networks like Chainlink operate) that listen for task assignments from the blockchain, execute them, and then return results or proofs back on-chain. These nodes would be the miners/validators in Bittensor’s analogy, but orchestrated through our own protocol rules. We could adapt something like Golem or Celestia (modular execution layers) to handle scheduling of compute tasks. Or use Bittensor’s existing substrate-based chain as a complementary system: e.g., tasks and rewards are posted on Ethereum, but actual AI model consensus is reached on Bittensor, then checkpointed back to Ethereum for payouts in CTZ. This hybrid approach leverages Bittensor’s expertise in decentralized AI training (with its Yuma consensus and subnets (Comprehensive Analysis of the Decentralized AI Network Bittensor - ChainCatcher)) while keeping the financial layer on EVM for broad accessibility.

Communication between these layers can be achieved via bridges or oracles. For example, once a compute node finishes a task, it might submit a transaction to the blockchain with the result’s hash and claim for reward. Other validators (or an optimistic/zero-knowledge proof) can verify the result off-chain. If valid, the RewardManager releases CTZ to the node. This could use an optimistic execution model (like optimistic rollups): assume the result is correct unless someone challenges within X blocks by providing a counter-proof. If challenged, run a dispute resolution (maybe a quorum of validators check the work or a smaller sub-computation is done on-chain if possible).

Figure 2 illustrates a simplified view of the token and compute flow across these layers:

Figure 2: Token Flow Across Compute Layers. The diagram depicts the high-level architecture. An End-User or a seed triggers a request, which is handled by a Parametric Product Seed (smart contract or on-chain reference to a model) that defines the task parameters. This in turn engages the Intelligence Middleware layer, which orchestrates the distributed compute off-chain. The middleware breaks the work into distributed jobs and assigns them to available Compute Providers (miners with hardware). CTZ tokens flow downward to pay for these jobs (either pre-staked or as payments for results), while results and proofs flow upward. The Compute Providers return the results (with proof-of-compute) to the middleware, which integrates them into an optimized solution for the seed. The seed contract then produces the final output for the end-user or the next stage of processing. In parallel, token flows reward the middleware (for orchestration) and possibly the seed owner (as royalty) from the CTZ spent. This layered approach ensures the blockchain handles transactions and value, while off-chain networks handle heavy computations, with CTZ tokens bridging the two.

A crucial component in making this work on existing infrastructure is EVM extensions or Layer-2 solutions to reduce gas costs. Many interactions (especially microtransactions for each compute sub-task) would be too expensive on Ethereum L1. We likely need a dedicated Layer 2 rollup or sidechain for CTZ that is optimized for high-frequency events. Perhaps a rollup that uses zk-proofs for verifying compute could be developed: e.g., a zero-knowledge proof that a certain model was trained to X accuracy could be verified on-chain, thereby rewarding accordingly without posting all intermediate data. Projects like StarkNet or zkSync might be leveraged for such proofs of compute integrity.

On the other hand, Bittensor’s TAO network offers an alternative: it’s a purpose-built blockchain for AI with its own consensus and token. We could conceive of CTZ being interoperable or even minted on Bittensor itself as a subnet token. Bittensor’s approach to subnets (Comprehensive Analysis of the Decentralized AI Network Bittensor - ChainCatcher) (Comprehensive Analysis of the Decentralized AI Network Bittensor - ChainCatcher) allows specialized clusters (e.g., one for vision models, one for language, etc.) each with possibly their own dynamic token (dTAO). One idea: use TAO as the low-level reward for raw compute (since that’s already in place), then have CTZ as an overlay token that encapsulates higher-level logic and cross-subnet orchestration. CTZ could be exchangeable with TAO at some rate or earned by converting TAO outputs into more user-facing value. If the audience are frontier tech builders, they might appreciate a discussion of how to integrate with Bittensor – for example, using TAO’s proof-of-intelligence mechanisms to validate our tasks.

Data availability and storage is another consideration. Seeds (AI models) can be large (gigabytes). Storing those on-chain is infeasible; instead we use content-addressed storage like IPFS or Arweave for model weights, code, and datasets. The seed’s smart contract would store just a hash or pointer to the model on IPFS. Compute nodes retrieve it, run it, maybe update it (if fine-tuned) and put a new version on IPFS, updating the pointer via the seed contract. This way the heavy content lives off-chain but is verified by cryptographic hashes on-chain.

Security is paramount. Attacks could include bogus results, model theft, or denial-of-service by spamming tasks. Our design mitigates these with the incentive alignment: nodes that provide bad results don’t get paid and lose reputation, while spamming tasks costs CTZ (due to reasoning rights pricing). We might implement stake-slashing: compute providers and middleware perhaps stake some CTZ and can lose it if they are caught cheating (like an on-chain dispute finding they falsified a result). This creates a trustless yet secure environment similar to how proof-of-stake validators operate.

Finally, since this is to be a public PDF and web platform white paper, a note on front-end: likely there would be a dashboard or DApp where one can see available seeds, ongoing tasks, yields, etc., interacting with the smart contracts. Builders reading this could envision creating such front-ends where seed creators upload models, users trigger jobs, and providers offer their hardware, all mediated by CTZ under the hood.

In summary, the technical architecture uses a combination of on-chain smart contracts (for economics and coordination) and off-chain compute networks (for actual AI tasks). Whether using Ethereum & rollups or a specialized chain like TAO, the key is ensuring interoperability so that value can seamlessly flow as tokens while computation flows across devices globally. The approach is modular: one could swap out the compute layer (say use a cloud provider or a decentralized network) without changing the economic logic on-chain, because CTZ and the contracts abstract that away. Conversely, CTZ could hook into other ecosystems (for example, integrate with DeFi by allowing CTZ to be borrowed/lent, enabling leveraging compute investments). This flexible, layered architecture is both ambitious and feasible given today’s blockchain and AI technology trends.

Application Domains and Early Deployment Opportunities

Identifying the right initial application domains is critical for proving out this tokenized compute paradigm. We target digital-only domains where value can be generated entirely through computation, and where rapid, iterative improvement yields significant rewards. Below is a ranked list of five application domains best suited for early deployment, along with why they are ideal candidates:

1. Autonomous Code Optimization and Software Improvement – Rank 1. This domain is highly suitable because it directly deals with self-improving systems. An AI can refactor code, search for more efficient algorithms, and apply patches continuously. The value is immediate: faster or more efficient code saves money and improves performance. We already have proof that AI can discover better algorithms – e.g., DeepMind’s AlphaDev found sorting algorithms “up to 70% faster” than humans had achieved (AlphaDev discovers faster sorting algorithms - Google DeepMind). In a token economy, a Code Optimization Seed could take in any software project and continuously profile and improve it, earning CTZ based on the performance gains delivered. Companies or open-source projects would stake CTZ for improvements, and optimizers (AI or human-assisted AI) would compete to earn that by delivering better code. Because code is purely digital and improvements can be verified quantitatively (e.g., runtime reduced by X%, memory by Y%), it’s an excellent testbed. Success here not only yields direct economic benefits (cost savings) but also can produce components (optimized code) that feed into other domains.
   
   

2. Algorithmic Trading and Financial AI Agents – Rank 2. Finance has been a playground for algorithms for decades. In this domain, an AI agent makes high-frequency trading decisions, portfolio optimizations, or crypto market-making strategies. It is entirely digital, and results are directly monetary (profits/losses). This suits our model because significant compute can be thrown at finding even slight edges in the market, which then translate into real value. An algorithmic trading seed could allow traders to invest CTZ to run massive simulations or optimizations of strategies, with rewards coming from actual trading gains shared back as CTZ (perhaps via tokenized profit-sharing agreements). The reason this is ranked high is the immediate willingness to pay – traders will gladly spend on computation if it gives them an edge. Also, the data for training (market data) is abundant and digital. One caveat: connecting to real financial systems requires careful bridging (potentially via oracles for price data). But early on, this could be tested in crypto markets natively, where CTZ might even be used as a stake to run certain automated strategies on DEXes. The competitive nature of trading will push the arbitrage and optimization aspects of our system to the limit, which is great for battle-testing it.
   
   

3. Adaptive UI/UX and Personalization Engines – Rank 3. This domain focuses on improving user interfaces or user experiences automatically by analyzing user interaction data and adapting software on the fly. While it interfaces with human users, it is digital-only in the sense that the product (a website or app) is software that can self-modify. An adaptive UI seed might, for example, use AI to rearrange a website’s layout to better suit each user’s preferences, or adjust difficulty in a game based on player behavior. This is well-suited because it generates a lot of subtle computation – A/B testing different variants, reinforcement learning from user clicks, etc. Human attention is involved here as the thing to optimize (so it’s a bit hybrid), but that’s fine because we can measure success by improved engagement or conversion rates and feed that back into the token rewards (the “hybrid integration” we discussed). Companies currently spend vast effort on manual UI optimization; automating it with AI that continuously learns could be very valuable. The token model could be that a site owner stakes CTZ and tasks the system to improve a KPI (say sign-up rate). The AI tries many tweaks (compute intensive if done in a high-dimensional space) and once it finds improvements, part of the value (like increased revenue) can be converted to CTZ to reward the process. The reason it’s rank 3 is that it’s slightly harder to directly monetize than trading or code (value is indirect via user satisfaction), but it’s an area where even small improvements are worthwhile at scale (think of giants like Amazon fine-tuning their checkout process, a small UX change can yield millions more in sales).
   
   

4. Autonomous Cloud/DevOps Optimization (AIOps) – Rank 4. Managing cloud infrastructure and IT systems is complex, and AI can be used to optimize resource allocation, auto-scale services, detect anomalies, and even orchestrate microservices better than static rules. This domain is digital and has clear metrics (cost of infrastructure, uptime, latency etc.). An AI Ops seed could, for example, observe the usage patterns of servers and reconfigure them or redistribute loads in a data center to save energy or improve performance. The token incentive could be tied to cost savings: for instance, if the AI reduces monthly cloud bill by 10%, a portion of that savings is used to buy CTZ and pay the AI (and its seed creator). This domain is attractive because companies spend billions on cloud bills and IT operations, and even a small percentage improvement is big money. It leverages computational interest: an AI that continuously monitors and tweaks systems will perform better over time as it learns the patterns of usage. It’s ranked slightly lower at 4 mostly because it often requires integration with legacy systems and careful safety checks (you don’t want an AI crashing servers in an attempt to optimize). But as a contained experiment (perhaps on a smaller scale cloud or a specific subsystem), it can show quick wins. Importantly, it’s entirely within the digital realm of server configurations and software deployments, so it fits the criterion.
   
   

5. Automated Cybersecurity and Threat Detection – Rank 5. In cybersecurity, AI can analyze network traffic, user behavior, and code to detect intrusions or vulnerabilities. This is a race against malicious actors, and heavy compute (for analyzing millions of events, running attack simulations, etc.) is often needed. A security agent seed could continuously mutate and test software for weaknesses (fuzzing) or scan logs for anomalies in real-time, improving its detection rules through machine learning. The value of preventing a breach is enormous, though hard to quantify. A token model could involve insurance-like pools: companies stake CTZ for an AI security service that guarantees a certain level of protection, and the AI network earns those tokens by fulfilling that guarantee (with penalties if a breach occurs that it should have caught). This domain is digital because all attacks and defenses happen in software space. It’s ranked 5 because, while very important, it is reactive (you only see value when something is prevented, which is hard to measure) and adversarial (attackers might also target the AI). However, it’s a perfect candidate to demonstrate computational arbitrage: the AI can run far more checks and analyze far more data than a human team could, and token incentives could crowdsource security expertise (people writing detection plugins as seeds, etc.). If executed well, any detected threat and fix could be immediately propagated across the network (seeds sharing learned immunities), acting like an immune system. Early deployment could focus on something like smart contract security auditing – an AI that continuously checks deployed Ethereum contracts for vulnerabilities (there’s demand for this in the crypto world itself).
   
   

These five domains are by no means exhaustive, but they represent a mix of areas where our paradigm likely outperforms traditional approaches. In each case, success would demonstrate that focusing on token-fueled compute leads to faster or better outcomes than focusing on human attention or static software releases. Moreover, these domains are not siloed – improvements in one (say code optimization) directly benefit others (the algorithms discovered could be used by the trading bots or the security agents). So there is a compounding effect: as each domain’s seeds generate computational interest, the knowledge and tools they produce can be shared.

In initial deployments, it’s wise to start with controlled environments – perhaps private testnets or consortia for a given domain. For example, partner with a software company to use CTZ incentives internally for code optimization and document the efficiency gains. Or run a pilot in algorithmic crypto trading with a few strategies to show profitability. Each success story will attract more developers to publish seeds in that domain and more token holders to stake, creating a network effect.

Over time, one can imagine vertical-specific communities forming around each domain within the platform, analogous to subreddits or DAOs: a Code Optimization DAO, an Algo Trading DAO, etc., each curating and governing their seeds and maybe issuing domain-specific tokens that interface with CTZ. This would further drive participation by experts in those fields.

In conclusion, focusing on these digital-only, high-impact areas provides the best chance to validate the token economy concept. They have clear metrics to optimize, strong demand for improvement, and purely computational value generation. Early wins in these areas will pave the way for expanding the paradigm to even broader scopes (potentially including physical domains eventually, though that introduces new complexities). But as a starting point, the above domains offer fertile ground for “computational seeds” to grow and showcase the power of this new paradigm.

Economic Value Capture and Distribution for Creators

A cornerstone of this paradigm is ensuring that those who create the initial seeds – the innovators and developers – are fairly rewarded as their creations evolve and generate value. Traditional tech often struggles with this (e.g., open source maintainers under-appreciated while big companies profit from their code). Our token economy explicitly addresses it through built-in value capture mechanisms for seed creators. We touched on some of these earlier (like computational interest royalties); here we consolidate and expand on how value flows back to creators, and how it can be distributed in a sustainable, equitable way.

Creator Royalties on Compute Usage: As described, whenever a seed is used or improved via token-backed compute, a small percentage of the CTZ spent is allocated to the seed’s creator. This is automatically enforced by the smart contract governing the seed (Chain Insights — How Can NFT Creators Enforce Royalty Fees Once and For All? | by Chain | Medium). If Alice deploys a new AI model seed, every time anyone runs that model (directly or as part of a pipeline) and pays CTZ, Alice gets, say, 1-5% of that payment. This model is analogous to NFT royalties but applied to usage rather than resale (How NFT royalties work: Designs, challenges, and new ideas - a16z crypto). The distribution can be immediate (Alice’s balance increments in real-time as usage happens) or periodic (accumulated and paid out daily/weekly). The crucial effect is that creators have ongoing income as long as their seed remains valuable in the network. They need not constantly intervene or sell services manually; the protocol handles it. Over time, a portfolio of seeds can become a significant revenue stream – empowering independent developers and small teams to focus on creation while the network works for them.

Initial Token Grants or Retroactive Rewards: Sometimes a seed might require significant upfront work (e.g., training a large model, which costs a lot of compute before deployment). To encourage building such seeds, the ecosystem can offer grants or retroactive rewards. A grant could be given in CTZ to fund the initial compute or development time, possibly via a community pool or foundation. Retroactive rewards mean that if a seed becomes very successful (judged by, say, total CTZ spent on it in the first 3 months), the creator gets a bonus lump sum of CTZ as a “thank you” from the network for creating a high-impact seed. This could be algorithmic or decided by governance vote. It parallels programs in some crypto communities where developers of widely-used protocols were rewarded after the fact (e.g., Uniswap’s UNI token airdrop was partly to give back to early contributors/users). In our case, a portion of CTZ inflation or ecosystem fund could be reserved to reward top seed creators periodically.

Ownership Tokens for Seeds (Creator DAOs): A seed could be launched as a mini-project with multiple contributors. In that case, one can issue sub-tokens or fractional ownership of the seed’s future CTZ flows. This turns a seed into a micro-DAO or community project. For instance, a group of 5 researchers co-create a new AI model. They mint an NFT or a set of tokens representing ownership, split among themselves, maybe even with some sold to fundraise. The CTZ royalty from that seed is then automatically split to the holders of those tokens (which could be encoded in the seed’s contract – e.g., it could reference an ownership registry to distribute the royalty). This way, collaboration is facilitated and everyone’s share is transparent. If someone wants to exit, they can sell their ownership stake without affecting the seed’s operation. This mechanism allows creative teams to form around seeds and ensures fair splits – all enforced by code. It could also involve community: perhaps early adopters or testers of the seed get a small ownership token as well, aligning incentives for them to help improve it.

Economic Moats for Creators: One might worry: if a seed is fully open (code and model accessible), could someone copy it and redeploy without the original royalty, undercutting the original? To counteract that, several factors help maintain the creator’s edge: (1) Network effects and reputation – the original seed might already have usage and integration, so a copycat would have to attract users away. (2) The protocol could discourage identical duplicates by, for example, downranking seeds that are mere clones without improvement (maybe via governance or a curation layer). (3) If someone does copy and eliminate the royalty, they might lower costs a bit, but they also lose the creator’s ongoing support/expertise. A savvy user might stick with the original knowing the creator has incentive to keep updating it (since they earn from it), whereas a clone has no guaranteed maintenance. That said, competition can happen – if someone truly improves the seed and redeploys, then they become a creator in their own right (with their version’s royalty). This is akin to forks in open source – healthy competition but typically the innovative edge stays with those pushing the frontier. The token incentives ideally foster continuous innovation by creators to stay ahead, rather than complacency. If you invented a popular seed, you’d be motivated to keep enhancing it (or release a v2) to retain users, because clones could nip at your heels. The perpetual royalty gives you resources to do so.

Distribution of Value to Support Ecosystem: Creators might also choose to distribute some of their earned value to others, which could further strengthen the system. For instance, a successful seed creator might allocate a portion of their royalties to those who supplied training data or who helped with testing. This can be done programmatically: maybe the seed contract says 3% goes to creator, 1% goes to data providers (with a list of addresses or another contract managing that). Another scenario is creators “pay it forward” by reinvesting CTZ into new seeds or bounties, effectively becoming patrons in the ecosystem. This voluntary distribution, while not enforced, could become a norm or be encouraged through social consensus (akin to open source ethos but with real currency). The existence of token rewards could ironically encourage more sharing: since creators are rewarded by the system, they might feel less need to hoard their breakthroughs and can afford to support others.

Transparency and Trust: All flows to creators are recorded on-chain. If you use a seed, you could see exactly what cut goes to whom. This transparency builds trust – users won’t feel gouged because they know the fee structure upfront and see it executed exactly. It’s also flexible: creators can adjust their fee (maybe lowering it if they got enough or raising if they think it’s warranted), but such changes can be made transparent and possibly subject to user acceptance (users could get notified of a royalty change and choose to switch seeds if they dislike it, applying pressure to keep fees reasonable).

In effect, the token economy aligns incentives so that creators, users, and compute providers are in a positive-sum game. Creators want to publish great seeds because they’ll earn from widespread use. Users want to use seeds because they get better service through AI, and they accept creators earning a slice as fair compensation. Compute providers just care about total CTZ they earn, which is unaffected by who gets the small royalty – it just slightly increases the cost users pay, which they’re fine with if the product is worth it.

To illustrate with a hypothetical: Suppose a developer publishes a seed “AutoCoder” that uses AI to optimize code (domain rank 1 scenario). She sets a 2% royalty. The seed quickly gains traction – in a month, 10,000 CTZ worth of compute has been spent by various projects using AutoCoder to improve their software. She earns 200 CTZ from that activity. If CTZ were trading say at $10, that’s $2,000 earned in a month passively. Meanwhile, those using it might have saved much more in engineering costs or improved performance, so it’s a win-win. Now she can use some of that CTZ to maybe pay for further research to upgrade AutoCoder (ensuring it stays the best and people keep using it), and maybe stake some CTZ to back another newcomer’s seed she finds promising – thus acting like an investor too. Over time, her initial effort of making a great seed pays back continuously, potentially far exceeding what she could have made by selling the software outright once.

Preventing Centralization: One must consider whether a few creators could dominate and capture outsized value (like Big Tech in attention economy). The difference here is the low barrier to entry – anyone can introduce a competing seed, and usage can shift quickly if it’s better or cheaper. The token system is open. So while some may become “rockstar” creators, there’s always the threat of new innovators, keeping them in check. Additionally, governance could impose caps or progressive decentralization of very fundamental seeds (for example, if a seed becomes part of critical infrastructure, the community might vote to reduce its royalty over time, while compensating the creator in a lump sum – speculative, but mechanisms exist if needed).

In summary, the design ensures creators capture a fair slice of the value pie they help bake, distributed automatically as their seeds flourish. This transforms the traditional model of value distribution: instead of creators needing to monetize via ads, subscriptions, or selling out to big companies, they can plug into this network and directly earn from usage. It could usher in a golden era of independent AI and software creators, analogous to how YouTube or app stores enabled independent content creators, but with far more equity since the terms are coded and not controlled by a central platform taking a huge cut. We envision a future where someone can publish a brilliant AI model from their home and that model itself becomes a revenue-generating “micro-business” through tokens. This democratizes the innovation economy and keeps it sustainable by continuously funding those who push it forward.

Conclusion and Call to Action

We are on the cusp of a paradigm shift in how digital products are conceived, built, and monetized. “Creation in the Token Economy” heralds a move away from designing for clicks and fleeting attention, towards designing for endless loops of algorithmic improvement fueled by cryptographic tokens. In this white paper, we articulated a framework where parameterized seeds – malleable, self-evolving product kernels – continuously grow in value via computational interest, paid for and measured in tokens rather than eyeballs. We explored the primitives that make this possible (Reasoning Rights, Computational Arbitrage, Intelligence Middleware, Computational Catalysts) and detailed how the Catalyze (CTZ) token interweaves incentive streams to align all participants in this ecosystem.

The implications of this new paradigm are profound. Products are no longer static offerings; they become living entities in a digital ecosystem, economically motivated to learn, adapt, and improve autonomously. Users cease to be targets for monetization, and instead become beneficiaries of ever-improving services – perhaps even investors and collaborators in those services via tokens. Developers and creators gain new avenues to profit from innovation without ceding control to ad-driven gatekeepers, receiving due rewards as their creations flourish in the network (Chain Insights — How Can NFT Creators Enforce Royalty Fees Once and For All? | by Chain | Medium). Infrastructure providers find efficient markets for their compute resources, directed to the most valuable tasks through token signals (Bittensor has several serious flaws. Is it doomed to fail? | 金色财经 on Binance Square). In short, the token economy can turn what used to be friction (the cost to run AI, the effort to update software) into the engine of a new value cycle.

This white paper is both an educational resource and a strategic call-to-action. To technologists, entrepreneurs, and builders at the intersection of AI and Web3: the tools and conditions are ripe to start building this future. Large language models and AI services are here and improving. Blockchains and tokens have matured to a point where complex economic behaviors can be encoded on-chain. The cost of compute remains a gating factor for AI – which is exactly why a market-based approach to harness every bit of useful compute is needed (Navigating the High Cost of AI Compute | Andreessen Horowitz). By aligning incentives properly, we can unlock an unprecedented scale of collaboration between humans and machines: think of it as Decentralized Autonomous R&D.

Consider this a blueprint, not a finished blueprint but a starting plan. Many challenges remain to be solved through actual implementation and experimentation. For example, refining the security of off-chain computation, fine-tuning the tokenomics parameters, and ensuring that the pursuit of token rewards remains correlated with delivering real-world value (avoiding Goodhart’s law traps where proxies diverge from goals). These require the collective intelligence of the community. We envision launching prototypes in one or two of the high-priority application domains (like an Autonomous Code Optimizer DAO or an Algo Trading network running on CTZ) to gather data and iterate on the concept.

To that end, we invite collaborators and pioneers across disciplines:

• AI Researchers & Engineers: Help design seeds or middleware that can learn and adapt – your innovations can now be directly monetized and scaled on-chain. For instance, if you’ve developed a novel model training trick, embedding it in a tokenized service could amplify its impact and funding.
  
  

• Blockchain Developers & Economists: Assist in building the smart contract infrastructure and refining the game theory. Challenges like oracle design for verifying compute, or creating stable reward loops, are in need of clever solutions. There’s space for new DeFi-like constructs (futures on compute, stake-for-service models) to emerge.
  
  

• Frontier Entrepreneurs: Identify niche problems that this paradigm can solve better than existing methods and drive real adoption there. Whether it's a startup offering “AI-as-a-service” via CTZ or a platform integrating these tokenized AI services into mainstream products, there are first-mover advantages to seize.
  
  

• Investors & Visionaries: Support this direction with capital and insight. Bootstrapping such an ecosystem might require initial funding of development and seeding the token economy with liquidity and usage. By backing these efforts, you’re not just investing in a project but potentially in the next foundational layer of the digital economy.
  
  

We also encourage engaging with existing communities like SingularityNET, Fetch.ai, Ocean Protocol, and Bittensor, which have been exploring elements of AI marketplaces, agent economies, and tokenized compute. Our paradigm is aligned in spirit with these, and collaboration or at least mutual learning could accelerate progress. For instance, Bittensor’s success and challenges in incentivizing decentralized model training (Bittensor has several serious flaws. Is it doomed to fail? | 金色财经 on Binance Square) (Bittensor has several serious flaws. Is it doomed to fail? | 金色财经 on Binance Square) provide invaluable lessons we can incorporate (and we have cited some along the way). By learning from such projects, we can iterate faster and avoid known pitfalls.

Ultimately, the vision is grand: a self-sustaining computational economy that drives continuous technological advancement. One where attention as we know it becomes a niche currency (perhaps 99.9% of “attention” indeed becomes AI attention (Have humans passed peak brain power? | Martin Signoux)), and the predominant currency is intelligence – measured in reasoning operations, bought and paid by tokens. This might sound like science fiction, but as we’ve shown, all the pieces are here in 2025 to start assembling this reality.

In closing, imagine a not-too-distant future scenario: A developer releases a “seed” for autonomous scientific research – an AI that can form hypotheses and run virtual experiments. They release it into the token economy. Thousands of CTZ tokens are allocated as the AI scours data, collaborates with other seeds (some focusing on simulations, some on literature review), and over months it churns out discoveries – perhaps even a potential new material or drug candidate – which are then validated and patented (with smart contracts ensuring credit and token rewards to all contributing seeds). The developer who started it sees their small initial idea blossom into something far beyond their own capacity, supported by a whole network of computation and earning them and many others ongoing rewards. Humanity gains new knowledge and solutions faster than ever, propelled by an army of tireless digital researchers that themselves grow smarter each day. This is the promise of creation in the token economy: a new paradigm where innovation is automated, incentivized, and democratized at scale.

It’s an ambitious path, but it’s one worth pursuing. Let us move from vision to reality – one computational seed at a time. The call to action is clear: join us in building this new paradigm, and together, let’s catalyze the next great wave of digital value creation.

Please note: Everything below is machine generated. At the end of the topic overview, I've prepared a package of essays that expands (or condenses) the argument for a range of different reading styles.

I’ve seen the future. It’s not a war-torn apocalypse. It’s not Blade Runner neon or Mad Max deserts. It’s quieter than that. It’s clean. Efficient. A garden meticulously pruned of excess humanity. The harvest is blooming, and we are the fertilizer—the final yield of capital’s perfect cycle.

They don’t need to kill us. They don’t need to enslave us. They just need to let the machine run. It’s a self-optimizing system, grinding forward on autopilot, with all incentives aligned toward a world where fewer people means higher profits. Depopuflation. Population collapse as a side effect of capital’s relentless push toward self-preservation.

There was no single moment where the switch flipped. No grand declaration, no shadowy cabal. Just a thousand small choices, each one profitable, each one inching us toward the same outcome. Neoliberalism didn’t need a master plan. It just needed to keep doing what it was built to do. Privatize, optimize, extract. Strip away collective power, dissolve solidarity, hollow out resistance. Preach the virtues of free markets while erecting invisible walls around every aspect of life, turning the commons into a subscription service.

And now we’re at the terminal stage. The machine is running at full capacity, and human labor is obsolete.

We hit Peak Human between 2020 and 2025. That was the inflection point—the moment when the system silently realized we weren’t necessary anymore. Like the horses of 19th-century London who were everywhere until they weren’t, human labor became a redundancy as soon as AI matured enough to displace not just manual workers but cognitive labor as well.

At first, AI played nice—augmenting human tasks, promising partnership. But then came the Displacement Phase. From 2025 to 2040, human decision-making was systematically stripped out of the economy. Lawyers, doctors, writers, teachers—anything that could be turned into a dataset was fed to the machine. The cognitive commons was enclosed.

Your knowledge? Privatized. Your creativity? Scraped and repackaged. Your very experience? Monetized and resold as data exhaust to fine-tune the algorithms that would replace you.

By 2040, the Redesign Phase was in full swing. Economic systems weren’t just automated—they were reconceptualized for a world where humans were the anomaly. Factories hummed in the dark, supply chains adjusted themselves in real time, and financial markets became an ouroboros—algorithms trading with algorithms in a closed loop of asset inflation that no longer required consumer spending to fuel it.

Neoliberalism had always been a virus. It infects institutions, repurposing them to serve capital. Governments, once designed to balance market forces, became handmaidens of the asset class. Public goods were auctioned off, regulatory frameworks hollowed out, safety nets dismantled in the name of “efficiency.” Every crisis was an opportunity to deepen the infection.

2008? Bail out the banks, let homeowners drown.

2020? Pandemic shock? Flood markets with liquidity and let asset prices soar while wages stagnate.

When birth rates plummeted? Don’t fix the conditions that made family life impossible. Preach natalism while ensuring reproduction remains an unaffordable luxury for anyone without generational wealth.

The machine never needed a villain—it just needed enablers. And they played their part perfectly. Politicians talked about “family values” while pricing homes out of reach for young families. Corporate media pumped out stories of self-made success while wages collapsed under the weight of credentialism and gig economy precarity.

AI crushed production costs to zero, but scarcity had to be maintained. Capital thrives on it. So essential goods—housing, healthcare, education—were turned into speculative assets.

If everything can be produced cheaply and abundantly, the only thing left with any value is relative social position. A house isn’t just shelter—it’s a speculative token that keeps you one rung above the unwashed masses. Education isn’t about learning—it’s about securing your place in the hierarchy. Healthcare isn’t about saving lives—it’s about who deserves to live and who gets left behind.

The bifurcated money supply locked this dynamic in. Asset inflation for capital holders, wage deflation for everyone else. New money flowed to the top, inflating asset prices and widening the gulf. Meanwhile, real wages eroded, and those without inherited assets were consigned to debt dependency as the price of mere survival.

Debt became the new feudalism. You didn’t own a home—you serviced a mortgage. You didn’t have a career—you juggled gigs to cover student loans and healthcare premiums. The underclass wasn’t imprisoned in factories—they were shackled by invisible chains of compound interest.

No one had to plan for depopuflation. The system engineered it naturally. As capital intensified its grip, family formation collapsed under the weight of impossible economics.

Credentialization stretched into the fourth decade of life.

Housing markets became speculative playgrounds for global capital.

Care work was devalued, relegated to the informal economy or automated entirely.

Reproduction became a liability, not an aspiration. Birth rates plummeted, not because people didn’t want kids, but because the system made parenthood economically suicidal.

And as fewer people were born, the system leaned in.

Fewer workers meant higher per-capita asset concentration for the elite. Lower population density preserved scarcity, keeping land and resources expensive. And a shrinking consumer base didn’t matter—capital had already learned to consume itself.

As labor became irrelevant, the economy stopped caring about workers. Productivity gains were no longer passed down through wages—they were captured as asset inflation. The asset-wage disconnect wasn’t a bug—it was the system’s core feature.

Capital captured efficiency gains and hoarded them. Housing prices soared while wages stagnated. Healthcare costs ballooned as AI diagnostics and treatment pipelines made medicine cheaper to deliver but more expensive to access. Real wages fell while consumer prices in essential sectors rose.

Capital consumed the future and spat out a debt treadmill. For those without inherited wealth, family formation became impossible. Debt financed consumption, collateral required pre-existing wealth, and birth rates collapsed in the underclass.

This is where it clicked for the elites. Fewer people wasn’t a crisis—it was a business model.

Depopuflation became a self-reinforcing loop. Fewer workers? Higher productivity per capita. Less demand for consumer goods? Higher asset-to-consumer ratio. Shrinking populations? Lower political resistance.

With a minimally viable population in place—just enough people to maintain elite consumption patterns and system oversight—capital had no incentive to reverse course.

Democracy relied on numbers. A mass labor force could bargain. A voting majority could demand regulation. But once population numbers fell and wealth concentrated, the math stopped working.

Unions couldn’t organize when jobs disappeared. Protest movements fizzled when political leverage evaporated. Democracy eroded not because of corruption, but because the numerical foundation that underpinned it was hollowed out.

By the time anyone realized, the system had already crossed the event horizon.

And now here we are. The terminal state of capitalism.

The world didn’t collapse. It became efficient. Clean. Quiet. A minimally viable population, just enough to service the machine. Cities aren’t ghost towns—they’re curated spaces where the rich live lives of handcrafted luxury while AI-run supply chains hum quietly in the background.

Experience became the ultimate commodity. Human presence was reduced to a luxury good—a bespoke service for the ultra-wealthy, whose whims were catered to by the remnants of the underclass that survived the purge.

AI governs. Capital accumulates. Humans watch from the sidelines.

And the machine keeps humming, because the purpose of the system is what it does. It was never about prosperity for all. It was about optimizing returns. And in that optimization, we were rendered unnecessary.

The garden has been pruned. The machine doesn’t mourn. It just hums softly, a perfect engine of capital, eternally self-replicating in a world where we have become obsolete.

The harvest is over. All that’s left now is the silence.

As an experiment, we've rendered several expanded versions of this topic. Each has a different length and stylistic approach, designed to cater to a range of familiarity and learning styles:

Just the facts: Wither Up Economics - Depopuflation Bullet Points (774 words, 3 min scan)

More explanation, but no handholding: Wither Up Economics - The Compact Edition (3175 words, 11 min read)

If you don't buy the thesis and want sourcing: Wither Up Economics - The Academic Edition (9839 words, 33 min read)

The relatable to everyday life, handholding version: Wither Up Economics - The Boomer Longform Edition (7712 words, 26 min read)

I. Introduction: The Unseen Harvest

• Depopuflation Defined: Systemic population decline as a second-order consequence of capital optimization.

  • Not a conspiracy—an emergent outcome.

  • Capital learns to grow without labor, leading to demographic collapse.

• Historical Parallel:

  • Peak Horse in 19th-century London: Horses became obsolete as automobiles redesigned urban infrastructure.

  • Peak Human (2020–2025): AI and automation rendered human labor increasingly redundant.

• Core Premise:

  • Capital optimizes for efficiency, and population decline becomes an inevitable byproduct.

• Severing the Capital-Labor Dependency:

  • Historically, capital required labor to generate value.

  • AI-driven automation eliminates that requirement.

• Indicators of Capital-Labor Inversion:

  • Negative Labor Premium: Workers contribute less economic value than their upkeep costs.

  • Autonomous Capital Chains: Self-sustaining production systems require minimal human oversight.

  • Inverted Growth Metrics: Nations with shrinking populations (e.g., Japan) show higher GDP growth rates.

• The Three-Phase Transition:

  • Augmentation Phase (2010–2025): AI complements human labor.

  • Displacement Phase (2025–2040): AI gradually replaces decision-making roles.

  • Redesign Phase (2040+): Entire economic systems restructured for non-human operational parameters.

• The Neo-Feudal Deflationary Trap:

  • AI deflates production costs but inflates consumption essentials.

  • Creates a two-tier economy:

    • Capital Asset Deflation: Production costs drop toward zero.

    • Consumption Cost Inflation: Housing, healthcare, education remain artificially scarce.

    • Wage Collapse: Labor value deteriorates as automation oversupplies the labor market.

• The Bifurcated Money Supply:

  • Capital Liquidity: New money inflates asset values for the elite.

  • Wage Deflation: Earnings stagnate, leaving most people reliant on debt to maintain consumption.

• The Asset-Wage Disconnect:

  • Productivity gains captured by capital, not passed to consumers.

  • Essential goods remain artificially scarce despite costless production.

• Enclosing the Cognitive Commons:

  • Cognitive Commons Defined: Collective human knowledge, creativity, and judgment built through cultural knowledge and shared learning.

  • AI captures and privatizes this commons, turning once-public goods into paywalled IP.

• Mechanisms of Cognitive Enclosure:

  • Data Expropriation: AI systems trained on uncompensated human outputs.

  • Algorithmic Enclosure: Proprietary algorithms privatize collective judgment.

  • Experiential Capture: Monetization of human attention, emotion, and experience.

• Historical Analogy:

  • Just as the original Enclosure Movement privatized land, cognitive enclosure concentrates collective intellectual resources within private capital systems.

• Economic Disincentives to Reproduction:

  • Extreme Credentialization: Prolonged education delays family formation.

  • Housing Price Inflation: Homes priced for investors, not families.

  • Devaluation of Care Work: Care work remains underpaid and undervalued.

• Resource Pricing and Fertility Collapse:

  • Essential goods (housing, education, healthcare) consume increasing portions of income, disincentivizing parenthood.

• Cultural Narrative Engineering:

  • Identity Shift: Individualistic, consumption-based identities promoted over family-based identities.

  • Children Framed as Luxury Goods: Parenting seen as a high-cost personal choice rather than a societal necessity.

• The Asset-Dependency Trap:

  • AI deflates labor value while inflating asset prices.

  • Human survival increasingly depends on inherited wealth rather than wages.

  • Outcome: Population decline accelerates among non-asset-holding classes.

• The Consumption-Production Disconnect:

  • Production Costs Approach Zero: AI automates supply chains.

  • Essential Consumption Becomes Scarce: Artificial scarcity in key sectors sustains asset value.

  • Result: Wealth concentration accelerates while consumer purchasing power erodes.

• The Debt-Collateral Spiral:

  • Essential goods require debt financing.

  • Debt Requires Collateral: Those without pre-existing wealth cannot participate.

  • Delayed Family Formation: Fertility rates decline among those unable to establish financial security.

• Numerical Decline Weakens Labor Power:

  • Traditional resistance strategies (labor unions, democratic regulation) rely on numerical advantage.

  • As population declines, the political leverage of workers erodes.

• Failed Interventions:

  • Natalist Policies: Incentives for childbirth undermined by economic disincentives.

  • Universal Basic Income: Absorbed by rent-seeking mechanisms in essential services.

  • Sovereign Wealth Funds: Vulnerable to capture by capital interests.

  • Monetary Policy: Central banks unable to resolve bifurcated inflation/deflation dynamics.

• Minimal Viable Population:

  • Just enough humans to sustain elite consumption patterns and oversee autonomous systems.

• Capital Sovereignty:

  • Decision-making transferred to algorithmic systems that prioritize wealth preservation.

• Experience Economy Supremacy:

  • Human experience itself becomes the last scarce commodity, stratified by wealth.

• No Conspiracy, Just Incentives:

  • Depopuflation wasn’t planned. It was the natural endpoint of a system designed to optimize capital accumulation.

  • The system’s true purpose is revealed through its outcomes—not its promises.

• The Horse and the Human Parallel:

  • Just as London’s streets once contained only the precise number of horses needed for specialized functions, the global economy is reorganizing itself to support only the human population necessary for elite consumption.

• Future Winners and Losers:

  • Those who recognize the system’s trajectory early can adapt and benefit.

  • Those who cling to traditional ideas of population growth and labor power face structural obsolescence.

• Post-Human Optimization:

  • The system doesn’t need most of us anymore.

  • Depopuflation isn’t a crisis—it’s the silent triumph of capital.

• Silent Victory:

  • The machine doesn’t mourn. It just optimizes.

No primers or euphemisms here: we assume you know about automation, artificial intelligence, demographic decline, cognitive enclosure, and capital accumulation. What follows is a synthesis, a systemic account of how these trends intertwine and reinforce one another. When no one pulls the brakes, the system’s purpose becomes evident in its outcomes. It doesn’t matter what uplifting mission statements or economic theories proclaim – what matters is what the system actually does. And what it is doing is harvesting. It’s extracting every last drop of value from a shrinking labor pool, privatizing every commons it can enclose, and riding an endless wave of asset appreciation, heedless of the fact that at the end of that ride there may be no one left to sell to but each other.

Neoliberalism – the late-20th-century gospel of free markets, deregulation, and unfettered individualism – set in motion a self-reinforcing loop. The path of least resistance became the path of intensification. Corporations relentlessly cut costs and chased cheap labor; governments slashed regulations and safety nets in the name of “competitiveness.” Public assets were sold, taxes on capital gutted, finance deregulated. Each step made the next one seem not only natural but inevitable.

The system effectively went on autopilot. Policies and practices that maximized short-term profit thrived; those that didn’t were swept aside by competition. No mastermind was required – the logic enforced itself. Businesses that broke unions, offshored jobs, and squeezed suppliers outcompeted those that didn’t. Politicians who preached austerity and “business-friendly” reforms received corporate backing, while those who didn’t struggled for power. Innovations that cut labor costs or dodged regulations spread quickly and set new norms. Success in this environment meant doubling down on market logic at every turn.

Over time, a series of feedback loops entrenched this trajectory:

• Wealth captures power: Wealth concentrates and buys influence to rewrite laws in its favor. Regulations get ever looser, taxes ever lower for the rich, creating a snowball effect that further concentrates wealth.
  
  

• Alternatives are eroded: Public institutions that once balanced the market – strong unions, robust welfare programs, public utilities – are defunded and discredited. People hear there is no alternative to privatization and personal responsibility. As collective solutions wither, the market expands by default into every area of life.
  
  

• Blame is diverted: As insecurity spreads, the ideology doubles down: if you’re struggling, it’s your own fault for not adapting. Social frustration is channeled toward scapegoats (immigrants, the poor, “lazy” workers) instead of the system. Even crises caused by market excess (crashes, recessions) are met with calls for more belt-tightening and market discipline, never questioning the model itself.
  
  

These loops feed each other, ensuring that once neoliberalism is installed, it keeps reinforcing itself. Even leaders inclined to reform it feel constrained by global market pressures and entrenched interests. For any CEO or politician, the easiest move is to stay the course. After decades of this, we have normalized astonishing inequality and precarious employment as just the way things are. The economy is treated as an untouchable, quasi-natural machine that we must obey. But in truth this machine serves a very select few – and as it runs, it’s content to grind everyone else into the gears.

While the neoliberal engine was accelerating, technology offered capital a new edge. Automation moved from factory floors to office cubicles; artificial intelligence now targets white-collar and creative work once thought safe. Crucially, these AI systems are trained on humanity’s collective knowledge and creativity – our writings, art, and code – and then locked behind corporate walls. This cognitive enclosure means the shared commons of intellect has been appropriated as private property. Capital now owns the machines and algorithms built with our data, and it can deploy them to replace us.

The implication is stark: a large portion of the population is no longer needed for their labor or even their expertise. The machines have learned from us how to do many of our jobs, and capital owns the machines. From robot-run warehouses to AIs that design, write, diagnose, and drive, humans are being squeezed out of production. An elite cadre of owners and engineers reaps these efficiency gains, running profitable enterprises with a fraction of the workforce. But for everyone else, it means redundancy.

It’s a rolling process of dispossession, not a sudden revolution. Drivers, assistants, analysts – all kinds of jobs are gradually made obsolete or pushed into tenuous freelance roles by the very tools built on yesterday’s human effort. The path for each firm is clear: if a task can be automated or handled by AI for less cost than a human (and increasingly it can), it will be. The competitive pressure ensures that any company slow to replace workers with cheaper tech is quickly outflanked by one that isn’t.

So the system sheds human workers. It’s nothing personal – just business. The broader effect is a growing class of people who find that their skills, their labor, their very minds are not needed. They are told to retrain, to “learn to code,” to move up the value chain, but the goalposts keep receding with each tech leap. It’s a classic contradiction: the system produces abundance (of AI-driven output) but yields scarcity of paid employment. We get more stuff, more services, more capability – but fewer people share in the income and wealth generated.

What does capital gain from automating away people? Independence from labor, and thus from labor’s demands. The less it needs us, the less power we have. Unions become toothless if their jobs can vanish with a software update. Strikes become less effective if production can continue autonomously. The bargaining power of the average person dwindles to nil. In the long run, capital learns to live without large swaths of humanity’s labor. And once it can do that, the system’s incentive to care about human well-being evaporates. After all, if people are merely a cost, the logical move is to minimize that cost.

As labor becomes dispensable, capital has found another way to expand its riches: by inflating the value of assets and extracting wealth from those who don’t own them. We now effectively have two economies running in parallel:

• The owners’ circuit: Money begets more money, almost by magic. Cheap credit lets the rich borrow to buy more assets, driving prices even higher. If markets falter, bailouts and easy money prop them back up – risk is socialized, gains are privatized. Owning things – not making things – is the surest path to wealth.
  
  

• The everyone-else circuit: For those living on wages, money is expensive and scarce. Credit cards, student loans, and rent devour paychecks. Wages stay flat even as living costs rise. Buying a home (a ticket into the asset-owning class) is a distant dream, so they rent indefinitely. Most people are one accident or interest rate hike away from financial ruin, and they get none of the safety nets that cushion the wealthy.
  
  

This divide is maintained through artificial scarcity. We have the capacity to provide a decent living for everyone, but access is limited to keep profits high. Housing is deliberately kept scarce and pricey – via zoning laws, speculative hoarding, or corporate landlords – ensuring people pay a premium for a roof over their heads. Knowledge and digital goods, which could be nearly free to share, are fenced off by patents, paywalls, and DRM; even information becomes a product you must continually pay for. Every sector finds ways to make what could be abundant artificially scarce for the sake of the bottom line.

At the same time, wage labor is bled dry. Careers turn into gigs and contracts without benefits. Workers are pushed to do more for less pay, and told to be grateful for any job at all. Productivity and profits rise, but the gains don’t trickle down – they gush upward into stock buybacks, executive bonuses, and investor portfolios. The official metrics might show economic growth, but that growth no longer translates into broad prosperity. It pools at the top, in the asset economy, out of reach of those whose labor actually produced it.

Work and wealth have been decoupled. The richest can multiply their fortunes even if the broader economy stagnates for everyone else. It no longer matters if the majority can’t afford to consume much; the real money is made elsewhere, circulating among assets and balance sheets. The system can claim to be thriving – high GDP, booming stock indexes – even as it starves the host on which it feeds. It’s another phase of the long harvest: capital maximizing gain not by producing more for all, but by siphoning value from the many to enrich the few.

Perhaps the most unsettling feedback loop of all is how economic precarity undermines the future of society itself. As inequality and insecurity deepen, people’s capacity to plan for the future evaporates. Starting a family, for example, becomes a daunting gamble. Birth rates are plummeting, not just from personal choice, but because having children feels financially reckless for anyone not securely affluent. Who can imagine raising a child when you can barely pay your own rent and debts? Fertility collapse becomes another symptom of the system’s squeeze – a silent vote of no confidence in tomorrow.

Debt is a big part of this story. Young adults enter working life already shackled by loans – for education, for a car, for medical bills. By the time they’d want to buy a home or start a family, they’re still digging out of debt. Home ownership, once the bedrock of middle-class stability, has become unattainable for many, so they end up renting indefinitely. With no house, no savings, and often no steady career, the prospect of supporting children – or even retiring comfortably themselves – slips away. Owning nothing isn’t a choice; it’s the default fate for a generation.

As fewer children are born and the population ages, the demographic balance tips. A shrinking workforce must support a growing elderly population. This would be challenging even in a fair society; in our unfair one, it becomes a pretext to slash what remains of social support. Pensions, healthcare, public services – all face cuts because there are supposedly “too few young people” to pay in. Never mind the immense wealth at the top; the narrative demands sacrifice from the bottom. Thus the underclass is squeezed at both ends: they can’t afford to raise kids, and then they’re told they must work ever harder, ever longer, to care for an aging society (even as the rich hoard resources that could have eased this burden).

The result is the entrenchment of a permanent underclass with no assets and dim prospects. Millions effectively become neo-serfs – renting every necessity, owning nothing. Housing is a lifelong expense, not an investment; transportation, software, even mattresses and appliances might come “as a service” on subscription. Miss a payment and you’re back to square one, because you have no property to your name. In such conditions, hope itself becomes scarce. Hard work doesn’t promise advancement; having children doesn’t promise stability. People stop believing the future will be better, even for their (hypothetical) children.

Disturbingly, this demographic contraction and social stagnation do not sound alarm bells for those at the top. If anything, a smaller, more desperate population is easier to manage: less demand for accountability, less resistance – and still enough consumption of essentials (propped up by a little government aid if needed) to keep business running. To the elite, the decline of the populace is just another trend to accommodate, not reverse. As long as their portfolios grow, they can accept a future with fewer people in it. In fact, from their vantage point, fewer people means fewer problems – a leaner society to dominate. It’s a grim trajectory: instead of rising prosperity lifting the masses up, selective decline is pushing the masses out of the picture.

Looking at this bleak trajectory, it’s natural to wonder if someone engineered it on purpose. But the truth is no grand conspiracy is required. When all the incentives line up this way, the outcome emerges “organically.” The wealthy and powerful don’t need secret meetings to benefit – they just follow their own short-term interests, and the system takes care of the rest.

They will:

• Cut labor costs – automate, outsource, bust unions – instead of ever sharing more profits with workers.
  
  

• Press for deregulation and tax breaks, claiming it’s for economic growth but really to free their hands and wallets. Public services get defunded, common resources privatized, and the market infiltrates every corner of life.
  
  

• Enclose new commons as soon as they appear: any frontier not yet privatized is a business opportunity. If something becomes vital (like the internet or our personal data), they’ll fence it off and charge admission.
  
  

• Control the narrative. They fund experts and media to rationalize all this as progress and necessity. Inequality is “meritocracy,” exploitation is “efficiency.” If people struggle, we’re told, it’s due to personal failings or unavoidable forces – never the choices of those in charge.
  
  

If a popular movement or crisis threatens this game, elites respond predictably – co-opt it, pacify it, or distract. Demand higher taxes? They offer token charity instead. Worker unrest? They tout “retraining” and maybe give a small raise. Financial crash? Bail out the banks, let the homeowners drown. The playbook repeats because it works: real change is averted, and the machine keeps humming in their favor.

These elites don’t see themselves as villains – they see themselves as smart winners and benefactors. Cocooned in gated bubbles, they don’t witness the suffering their decisions cause. Private security shields them from crime, private clinics from failing public health, private jets from crumbling infrastructure. In their eyes, the system is working just fine – for people like them. Why would they ever hit the brakes?

In effect, the lack of a collective plan is the plan. By simply doing nothing to alter the course, the elite ensure the trajectory continues. Their wealth and influence quietly keep the machine steady for their benefit, but they will not attempt a hard turn to avoid the cliff – that would mean changing a system that rewards them richly. They’ll ride it out, confident they can buy their way out of any messes along the road. This is, in their view, the price of progress, and besides, they’ll say, there is no alternative.

Judge the system by its outputs, not its propaganda: clearly its real purpose is to enrich a few and treat the rest as expendable. For decades we were promised that markets and technology would bring shared prosperity, but the reality has been the opposite. The invisible hand of the market has become a harvester’s scythe – sweeping wealth upward and cutting away the lives below.

This isn’t the future anyone advertised, yet it’s the one quietly being built. No one had to openly declare that society should function this way; it’s simply what has transpired when no one with the power to stop it chose to do so. The slogans said one thing, but the system did another. In practice, we’ve seen a long harvest – a drawn-out extraction of value from people and planet. Abundance exists, but only as lavish excess for a few. Progress exists, but mainly in forms that reduce the need for human workers or that turn every waking moment into a monetized transaction. The majority are left with scraps or nothing at all, and told it’s the natural order of things.

Picture the endpoint of this trajectory. In gated enclaves, a small elite thrives on self-perpetuating capital. They trade assets among themselves at ever-rising prices because no one else can afford to buy in. Outside those bubbles, former middle-class neighborhoods have hollowed out. Whole regions are economically barren, populated by people rendered superfluous. Automation and AI provide the goods and services the elite need; the underclasses either subsist on meager stipends, hustle in a shadow economy, or simply languish. The consumer society that defined the 20th century has bifurcated: luxury for the rich, bare survival for everyone else. In boardrooms and stock exchanges, it hardly matters that the masses are absent – the machine of capital can run with minimal human input, and it rewards the owners of that machine handsomely.

Crucially, none of this unfolds with dramatic fanfare. There is no singular moment when the curtain drops on the old world; instead it’s a slow eroding of expectations and possibilities. A quarterly earnings report here, a policy tweak there, a new app that replaces a thousand jobs somewhere else – each step is mundane enough. But together, they steer us into a dystopia of diminishing humanity. It happens not with a bang, but with the quiet hum of servers and the shuffle of financial papers.

Unless something external forces a change, why would it stop? The machine has no internal brake set to say “enough” on society’s behalf. Absent intervention, it will continue on autopilot toward this grim horizon. Left unchecked, this system will devour its own foundations in pursuit of profit, consuming even the people it once needed until it can run with barely any humans at all. We are already sliding toward a world where capital survives with minimal human labor or consumers. In that world, most of us become irrelevant to the economy. The elite will justify it as inevitable, perhaps even call it efficiency. But make no mistake: it’s the outcome of choices – of letting this happen.

None of this is preordained by technology or fate – it’s simply the trajectory we’ve allowed. But as long as those benefiting most see no reason to change course, this trajectory remains the default. The system will continue to accumulate for the few and discard the many, learning to live without the very people it was meant to serve. And it won’t shed a tear as it does so.

At that moment, a notification buzzed on my phone. Our stock price was soaring. The same markets that often slumber through good news had jolted awake at the scent of job cuts. In the span of minutes, the company’s valuation leapt upward, as if some invisible judge had declared: fewer workers equals a more valuable enterprise. I felt a chill watching the stock ticker green up while, just feet away, my colleagues stared down at severance packets. It was a lightbulb moment for me—an epiphany in real time. In that stark juxtaposition, I glimpsed a disturbing but plausible future taking shape: one where capital reaps the rewards as labor withers away.

Leaving the office that day, I wandered through the downtown streets in a kind of daze. The city felt familiar yet subtly off-kilter. In the lobby, I passed an automated kiosk cleaning the floor, humming softly to itself. Outside, a delivery drone whirred overhead, unbothered by traffic lights or paychecks. Around the corner, a glossy new supermarket stood where an old department store had been. Curious, I stepped inside—only to find no cashiers at all. Shoppers glided in and out through turnstiles, scanned by cameras and AI that recognized their purchases. A lone employee hovered by the entrance, not to help customers, but to troubleshoot the machines. It struck me that on this ordinary day, I was seeing extraordinary signs of change: human workers gently edged out by algorithms and robots, the physical presence of labor fading from view.

That night I couldn’t sleep. My mind replayed the day’s scenes and spiraled further: I recalled reading about birth rates plunging in one country and record stock market highs in another; about a tech giant paywalling its AI tools after training them on open internet data; about governments injecting trillions into financial markets during a crisis, propping up asset prices while food bank lines stretched around the block. These disparate threads started to braid together in the dark. It was as if I had been guided into a dimly lit corridor full of shapes I could only now discern: artificial intelligence and automation, demographic shifts, the privatization of knowledge, runaway asset wealth, bifurcated money flows, and the slow sidelining of workers. In that restless moment it all coalesced for me into a single storyline—a new, intensifying chapter of our economy. Half in jest, half in dread, I found myself calling it “Wither Up Economics.” If the old promise of neoliberalism was that wealth would “trickle down” to everyone, this felt like its dark mirror image: wealth gathering and swelling upward, while everything below withers in its shadow.

Over the next few days, I began probing deeper into this epiphany. What I found was not a conspiracy or a sci-fi dystopia, but a set of interlocking forces quietly rearranging society in capital’s favor. No single person or cabal engineered it; rather, systemic incentives have been nudging us step by step into a future where capital triumphs decisively over labor. In the pages that follow, I invite you to walk with me through that dim corridor of emerging realities. We’ll begin with the neo-enclosure of our collective mind, move through the erosion of labor’s power in an aging, automated world, and then confront the financial currents that enrich assets while stranding wages. Each section is a piece of the puzzle. Together, they form a chilling picture of where we may be headed. This isn’t a comfortable journey, but by its end you might feel, as I did, that the ground beneath us is shifting—and that the future, however unsettling, is already in motion.

The first stop on this journey was prompted by that office layoff: a machine had essentially taken the place of dozens of skilled workers. To understand how that was possible, we have to step back and look at what fuelled that machine’s intelligence. The AI system that replaced my colleagues did not emerge from a vacuum; it was trained on vast troves of data, much of it sourced from the collective output of human minds. Every email ever written, every line of code open-sourced, every article and image on the internet – all of it formed a cognitive commons, an immense pool of shared knowledge and creativity. It is the digital equivalent of the village commons of old, where anyone’s sheep could graze freely. But now, that commons is being quietly fenced off and monetized in what can only be described as a neo-enclosure movement.

Enclosure has a cruel history. In medieval and early-modern times, peasants woke up to find the open fields and forests – their commons – suddenly claimed by landlords, bordered with stone walls and hedges. What had been shared wealth became private property, and those who once relied on it were left with nothing. Today, we’re seeing a strikingly similar process, only the turf is not soil but human knowledge and attention. Generative AI models are a prime example: they feed on our collectively produced content (books, blogs, art, conversation) and transform it into proprietary products owned by tech companies. A poem you tweeted a decade ago or a design you posted on DeviantArt might now be a tiny drop in the data ocean that an AI scraped and learned from. But the AI’s capabilities – to write, to paint, to code – are owned by the entity that trained it, not by the public who provided the raw material. The cognitive commons is being enclosed: open-source knowledge in, closed-source product out.

It’s not just AI art and chatbots. Think about our daily life online. Attention has become property. Scroll through social media and realize that every witty post, every video, every meme is part of a vast human conversation that, in aggregate, creates huge value. Yet that value is captured by a few platforms which sell our mindshare to advertisers. Our shared digital culture – conversations with friends, funny videos, grassroots creativity – is packaged into targeted ads and algorithm-driven feeds. One commentator called this “the ultimate enclosure – the enclosure of the cognitive commons, the ambient mental atmosphere of daily life”​

In other words, our very awareness has become a resource to be mined. What we think about, what occupies our mind, is increasingly determined by those who control the platforms and algorithms. If the feudal lord of old could fence the land, today’s lords fence the mind, bombarding us with curated information, stimuli engineered to capture and profit from our attention.

The neo-enclosure of the cognitive commons has profound effects. It tilts the playing field of innovation and opportunity toward those who own intellectual property and computational power. A handful of tech giants now control the APIs, the datasets, the cloud servers – the new means of production for knowledge work. Meanwhile, individuals and communities that once shared freely find themselves either locked out or locked in: locked out of accessing the most advanced tools unless they pay rent to the owners, or locked into ecosystems where their every action generates data for someone else’s gain. We celebrate how smartphones and search engines put the world’s knowledge at our fingertips, but less noted is how they also privatized the gatekeeping of that knowledge. The internet was once hailed as a great equalizer, a global commons of information. Increasingly, though, the “cognitive surplus” of millions is siphoned into corporate silos.

This enclosure isn’t announced with fanfare. There’s no decree saying “this knowledge now belongs to X corporation.” It happens subtly: an open platform slowly tightens its API access, a public research dataset gets bought out, a popular open-source tool is absorbed into a paid product. We wake up to find the digital pasture we wandered freely now has “No Trespassing” signs – or at least “For entry, pay a subscription”. And just as enclosure of land displaced peasants, enclosure of knowledge displaces workers. Graphic designers, journalists, coders, even lawyers and doctors – all these professions rely on knowledge and patterns that AI is learning to replicate. As the AIs get better, the humans in those loops feel the ground shrinking beneath their feet. The intangible commons of skill and know-how, built up by generations, is becoming the private domain of capital – capital in the form of algorithms and patents.

Standing in that office after the layoffs, I realized the AI that took over was not just a clever tool; it was the spearhead of a sweeping change. It represented a transfer of competence and autonomy from individuals to capital. The company wouldn’t need to pay a salary to the AI; it simply owned the AI (or rented it from another corporation), which would tirelessly perform tasks without breaks, complaints, or collective bargaining. In a sense, the knowledge and experience of my former colleagues had been appropriated – first as the training data that informed the AI, and second as the ongoing improvement the AI would get from doing their jobs. What was theirs became an asset on someone else’s balance sheet.

This is Wither Up Economics at the cognitive level: a privatization of the mind. We see it when a public university’s research is patented and sold, or when a community forum’s content is used to train a machine that ends up replacing the very contributors. It’s a world where ideas and creativity become capital’s fodder. And it sets the stage for the next development – because once capital has harnessed collective knowledge, the next logical step is to reduce its reliance on the knowledge holders themselves, i.e., the workers. In other words, enclose the commons, then sideline the commoners.

A week after the layoffs, I visited my grandfather for coffee. He’s in his late 70s and often reflects on how different the world is now. We sat in his living room, sunlight falling on a shelf of old family photos. One picture showed him as a young man among hundreds of factory workers outside a plant that long ago shut down. “There were so many of us,” he said, shaking his head. And it’s true—there were. In his youth, people were everywhere in the economy: building, manufacturing, driving, selling, typing. Now, it seems, fewer hands are on deck. Factories are populated by robots. Offices, by software. Even the trucks are learning to drive themselves.

But it’s not just technology thinning the ranks of the employed; demographics are doing it too. My grandfather’s eyes welled with tears as he mentioned a few of his old friends who had passed away recently. It dawned on me that an entire generation is exiting the stage, and the generation coming after is smaller than the one before. For years we’ve heard about aging societies: in Japan, in Europe, in America. This year, Japan’s census confirmed a stunning trend—deaths now outnumber births by such a margin that the population is set to shrink by nearly a third over the next 50 years. The evidence appears in mundane statistics: adult diaper sales have already doubled those of baby diapers in Japan​. It’s a poignant, almost surreal marker of society turning gray. And Japan is just an early example; many other countries are only a few steps behind on the same path of demographic contraction.

At first glance, you might think a shrinking working-age population would shift power back to workers. Fewer workers available should mean companies compete harder to hire them, driving wages up. Indeed, in certain sectors we do see labor shortages and rising pay—for example, construction or nursing in aging societies. But here’s the rub: just as labor starts to become scarce, capital is finding new ways to do without labor entirely. AI-enabled automation is stepping in to fill the gap that demographics creates. In Japan, where combinatorial forces of a labor shortage and abundant technology collide, you see automated restaurants, robot caregivers, AI-powered clerks. What began as a response to not having enough young workers has become an incentive to never need young workers again. It’s a quiet substitution: roles once occupied by people are erased not through malicious intent, but through a cold cost-benefit logic accelerated by circumstance.

I think of the self-checkout kiosks at the grocery store. A small sign cheerfully asks for patience with the “new automated checkout system” and promises “more efficient service.” For the shopper, it’s mildly convenient (or mildly annoying, depending on your mood). For the corporation, it’s another notch of labor removed from the ledger. One less cashier to schedule, to pay, to provide health insurance for. Multiply that by millions of touchscreens across retail chains worldwide, and you start to see the outline of a labor-light economy. And it’s not just low-wage jobs at risk. White-collar automation is coming on fast: AI that can draft legal contracts, diagnose illnesses, design buildings, even write code. A recent analysis by a major consulting firm suggested that by the mid-2030s, up to 30% of all jobs could be automated in some way​. Even if that figure ends up a bit high or slow to materialize, the direction is clear. We’re entering an era where the default response to a workforce crunch is not to improve job quality to attract more workers, but to eliminate the job entirely.

This has all sorts of cascading effects. Labor’s bargaining power erodes when the boss can say, “If you want $1 more an hour, I’ll just install a machine to do your job.” In warehouses and factories, the threat needn’t even be explicit; the mere presence of automation in the wings tamps down workers’ willingness to demand more. On top of that, unions have been declining for decades, stripping workers of collective voice just as they need it most. Many young workers have only seen gig work or short-term contracts, with no expectation of long-term stability. The message they’ve internalized is clear: you’re on your own.

Now add demographics back into the mix. An aging society votes and spends differently. Political pressure shifts toward protecting the retirees (who, importantly, often hold significant assets like homes, stocks, and pensions) rather than empowering the young workers. Policies skew to keeping interest rates low and asset prices high – partly because nobody wants their 401(k) or home value to drop as they approach retirement, and partly because governments need cheap debt (more on that soon). In effect, the old and capital-owning are subtly aligned interests: stable assets, low inflation, steady dividends. What about the young and wage-earning? They are fewer in number and often distracted by economic precarity – high rent, student debt, unstable jobs – to marshal political clout. The result is a further systemic sidelining of labor: not by any grand decree, but as an emergent property of an economy that finds it can run on ever fewer workers and doesn’t miss a beat when those workers have diminishing influence.

It’s a bizarre inversion of the 20th-century norm. For a long stretch of modern history, we fretted about having too many people and not enough jobs. Now, in some advanced economies, we face too few people for the jobs that need doing – yet instead of this leading to a golden age for workers, it’s leading to rapid investment in technologies that make workers optional. It feels like a slow-motion coup: not one orchestrated by generals or dictators, but by CEOs, algorithms, and societal aging. Labor isn’t being lined up against a wall; it’s simply being shrunk, automated, and rendered peripheral over decades. My grandfather’s photo of hundreds of factory hands is a relic not just of the past, but perhaps of a pattern of life that will not return. The future of work might be a future with less work – at least, less work done by humans and far less bargaining power for the humans who remain.

Walking through the city these days, I sometimes play a game: I look at each worker I see and ask, “Will they still have that job in 10 or 20 years?” The barista – maybe, if we still like a human touch, or maybe not if a robot arm makes latte art better. The cab driver – likely to be supplanted by a driverless car. The nurse – probably safe for longer, though AI will assist her, and tablets might replace some of what a nurse’s aide used to do. The construction worker – his exoskeleton suit helps him lift heavy beams now, but perhaps one day the suit won’t need a human inside. It’s a morbid exercise, I admit. But it underscores how pervasive the incentive is to trim human labor out of the equation. Each of those scenarios represents capital (in the form of machines or software) stepping in for labor. Each is capital saying to labor, “I’ve got it from here, thanks.”

This steady march isn’t just about replacing workers—it’s about changing the fundamental balance of power between capital and labor. In classical economic struggles, labor and capital needed each other: one brought muscle and skill, the other brought tools and money. You couldn’t run a railroad without both investors and thousands of workers swinging hammers. But now, capital is building a world where it doesn’t need as many hammer-swingers. It can do with a handful of engineers and a fleet of automated drills. And when capital no longer needs labor, labor loses whatever leverage it had left. That’s the endgame of this trend if left unchecked: a society where capital effectively rules, not by force or legal decree, but by simply not relying on humans to generate wealth.

The Asset-Wage Disconnect: Two Economies, One Winner

One evening, not long after the layoffs, I met up with an old friend who works in finance. We sat at a rooftop bar, clinking glasses as the city lights blinked on. I confessed to him my growing unease about where things are heading. He listened, then asked, “Have you looked at the markets lately? The economy’s doing great, by some measures.” He pulled out his phone and showed me a chart of the stock market’s performance over the past decade. It was a steep, nearly unbroken climb to the sky. I countered by recalling what wages had done in that same period—little more than flatline. We found ourselves staring at two very different stories: one of capital booming, the other of labor treading water.

This is the asset-wage disconnect laid bare. The total value of stocks, real estate, and other assets has been exploding upward, especially in the last decade or two, while average incomes barely inch forward. In fact, over a span of 40 years, U.S. corporate stock values increased on the order of 60-fold (inflation-adjusted), whereas real wages for the typical worker have remained essentially stagnant. It’s hard to even wrap your head around that divergence—imagine a graph where one line shoots up like a rocket and the other hugs the x-axis like a dead snake. We’re living in that picture. And it creates the sensation of two economies coexisting: one swimming in money, the other perpetually scraping by.

How did we get a split like this? Part of the answer lies in how money flows through the system, something I think of as a bifurcated money supply. Consider the actions of central banks and governments in recent crises. When a recession looms or a pandemic hits, the traditional playbook has been to flood the financial system with money – slashing interest rates, buying up bonds and other assets (known as quantitative easing), and generally doing everything to ensure banks have liquidity and investors stay confident. This is often done in the name of helping the “economy,” but much of that help travels on a one-way express lane to capital markets. Bond-buying by the central bank raises bond prices and lowers yields, which pushes big investors towards stocks and real estate, driving those prices higher. Companies find it easier to borrow money at cheap rates, which they often use to acquire other companies or buy back their own shares, further boosting stock prices. Asset inflation, in other words, becomes standard operating procedure.

Meanwhile, what about the other side of the economy – the world of paychecks and households? Some money does reach here (stimulus checks, unemployment benefits, etc., when political will allows), but it’s comparatively modest and usually temporary. Policymakers are quick to worry about wage inflation if workers seem to be getting too much leverage (heaven forbid wages rise faster than a couple percent—then it’s time to “cool off” the economy). So even in boom times, there’s a tendency to tap the brakes if too many people start getting raises. The result is an odd equilibrium: high-octane fuel for the financial engines, strict speed limits on the real economy. You get bubbles in stocks or housing, but minimal wage growth and only mild consumer price inflation. It’s as if the river of money split into two streams – one rushing into the ponds of the wealthy, the other a thin trickle sustaining everyone else.

Sitting on that rooftop, I remarked to my friend, “It’s like the markets and the people live on different planets.” He nodded, swirling his drink, “Well, in my line of work, we call the market the real economy.” That struck me. Wall Street had come to see itself as the main show, with jobs and wages as a secondary subplot. And perhaps in truth, their fortunes had become decoupled from the fate of workers. A corporation can see its profits surge even while cutting staff or holding wages flat, because maybe it found a way to sell more products overseas, or it invested in automation, or it simply benefited from a speculative frenzy in its stock. In the old industrial model, to get richer you often had to hire more workers and pay them to make more widgets. In the new model, you might get richer by not hiring—by optimizing, downsizing, merging, algorithm-izing.

This asset-wage disconnect also explains a lot about inequality and the lived experience of different social classes today. If you were fortunate enough to own a home, some stocks, maybe a 401(k) retirement account, you’ve probably seen your net worth climb nicely over the past decade. Many such folks feel perplexed by talk of economic malaise—“Look at the Dow, look at housing prices, things are good!” But if you’re someone who depends entirely on wages, with little or no assets, you likely feel like you’re running in place or falling behind. Your rent went up, health care and education costs skyrocketed (not coincidentally, those are sectors heavily driven by either assets or monopolistic dynamics), and your paycheck simply hasn’t kept up. To you, claims of a “strong economy” ring hollow, even mocking.

Generationally, this split is producing a chasm. Baby Boomers and many Gen X-ers managed to buy houses when they were cheaper and have ridden that wave; younger Millennials and Gen Z, in contrast, often see homeownership as an unattainable dream, and without it they have no stake in the asset bonanza. They’re effectively locked out of the capital party, stuck in the world of labor. This asset inequality begets more inequality: an older landlord collects rent from a younger tenant, transferring wealth monthly from the wage-earner to the asset-owner. Or consider two college buddies: one became a skilled employee, the other became an entrepreneur who eventually sold his startup or got stock options at a tech firm. Fast forward 20 years, and the latter might be 100 times wealthier than the former, not because he worked 100 times as hard, but because he had equity – a piece of the upside – whereas his friend had a salary. We are increasingly a society split between the investers and the invested-in.

There’s also a psychological component to this. When asset values keep rising, those who hold them feel secure, even blithe. A rising stock portfolio or home equity can cushion a lot of bumps. It can pay for your kids’ college or your medical emergency or your retirement with room for cruises and vacation homes. So if you’re in that boat, the system is largely working for you. You’re less likely to notice the systemic rot underneath or to rock the boat if you do. But if you’re on the other side – no assets, just labor – you live with a constant precariousness. One unexpected expense or layoff can spell disaster. And you notice, keenly, that no one is flooding your bank account with cheap money. In fact, when times get tough, you might struggle to even get a small loan, whereas big companies can borrow millions at near-zero rates. The injustice isn’t always loud; often it’s a quiet background hum, like a machine room vibrating under the floorboards of society, steadily transferring wealth to one side.

To put it succinctly, capital has learned to thrive without labor thriving. The stock market doesn’t need full employment or rising median wages to hit record highs; it just needs profits and plenty of liquidity sloshing around. We’ve redesigned our economy—through policy choices, technological changes, and globalization—so that the fortunes of capital and labor need not rise together. In fact, they can diverge dramatically, as they have. As one analysis noted, neoliberal policies essentially shifted inflation and growth into asset prices and away from paychecks​. The promise was that the wealth at the top would eventually trickle down, but in practice it has pooled and pooled at the top, forming an almost separate ecosystem of wealth. Call it the Moneyverse, parallel to the universe the rest of us inhabit.

This parallel world has its own self-reinforcing logic. When crises hit, the guardians of the economy (central banks, treasuries) rush to rescue the financial system first and foremost—because they believe (or at least say) that helping capital will save us all. Sometimes it does prevent broader disaster, but it also reinforces the privileged status of capital. Think of the 2008 financial crisis: banks got bailed out to stop a meltdown of assets, and maybe that prevented deeper recession, but millions still lost homes and jobs, and hardly any banker faced consequences. Or the pandemic: swift action by the Fed meant the stock market recovered in mere weeks after the crash, and then went on to soar to all-time highs even as millions were still out of work and lining up for food relief. It was astonishing—pain on Main Street, party on Wall Street. The disconnect couldn’t have been more stark.

We’re left with a situation where capital accumulation has become uncoupled from broad human progress. It’s like a high-speed train that has detached from the passenger cars and left them behind on the tracks. The engine is speeding ahead – that’s the world of assets and capital gains. The passengers – everyday working people – are stranded, or moving at a crawl. Occasionally the train sends a shuttle back to offer a ride to a lucky few (say, a startup founder or a homeowner in a gentrifying neighborhood), but many remain stuck. If you listen carefully, you can hear some economists and politicians celebrating the engine’s speed while others are desperately trying to figure out how to reconnect the cars. But in the logic of Wither Up Economics, you don’t reconnect them. The whole point is that the engine doesn’t need the cars to keep going.

One question might be nagging at this point: if wages are stagnant and work less secure, how is society holding together at all? How are people affording things like homes or education or even basic consumption? Part of the uncomfortable answer: debt. Over the past few decades, we’ve increasingly turned to debt to paper over the cracks between stagnant incomes and rising costs. And debt, in large volumes, has its own powerful effect on the relationship between capital and labor.

Picture a young couple that wants to buy a house. Thirty years ago, maybe the price of a house was such that one middle-class income (or one and a half incomes) could pay the mortgage. Today, in many cities, the price of entry is so high that even two incomes strain to meet it. How do people still buy houses? They take on larger and longer loans, often with both adults working full-time to pay it off. So the couple gets the house, but along with it they sign away a big chunk of their future earnings to the lender. In essence, their future labor is mortgaged to capital (the bank). Multiply this story by millions and extend it beyond housing—student loans for education that wages don’t cover, credit card debt for living expenses that paychecks can’t stretch to, auto loans for cars to get to those far-flung jobs, and so on. The result is an indebted society, where many people’s present and future labor is already claimed by creditors.

Now, debt in moderation isn’t necessarily bad. It can be a useful tool. But what happens when it becomes systemic and ever-increasing is what we might call a debt-collateral spiral. Here’s how it works: In order for lenders to be comfortable extending all this credit, they need assurance they’ll be paid back. That usually means collateral – an asset they can seize if you default. For mortgages, it’s the house; for student loans, it might technically be your future earnings potential; for corporate loans, it’s the company’s assets or revenues. As credit has expanded, asset values have had to keep rising to backstop that credit. If house prices stagnate or fall, suddenly banks worry the loans on their books are not secure. If corporate stock values fall, suddenly those companies look shakier and their debt might be downgraded. So there’s an implicit pact: keep asset prices growing, and we can keep debt growing, and that in turn keeps consumption and investments going even if wages are flat. It’s a bit like stretching a rubber band – as long as it doesn’t snap, you can stretch surprisingly far.

But of course, sometimes reality doesn’t cooperate. We saw that in 2008 when the U.S. housing market—a bulwark of collateral—collapsed under its own speculative excess. Suddenly that spiral went in reverse: falling home prices meant trillions in mortgage debt were not fully backed; banks panicked, credit froze, and the whole house of cards nearly came down. The response was telling: governments and central banks rushed not so much to bail out individual homeowners (millions of whom lost homes), but to bail out the financial system. They effectively said, we will not let asset prices crash further, we will not let the lenders go down. It was messy and imperfect, but they succeeded to an extent in re-inflating the collateral (through low interest rates and buying up bad assets) so the debt could be sustained.

From that crisis emerged an unspoken doctrine: never let the debt-collateral spiral unwind chaotically. Since 2008, whenever there’s a hint of crisis, authorities step in forcefully to stabilize assets and credit. Think of it as a tightrope walker who must keep moving forward—if they try to go backward or stop, they fall. Similarly, our debt-fueled system must keep pressing on, more debt, higher assets, rinse and repeat, or face collapse. This imperative further binds the fate of the economy to the interests of capital. Because who benefits most from propping up asset prices? Those who own the assets (capital owners). Who benefits from keeping credit flowing? Those who need credit, yes (which is many of us), but also those who provide credit and earn interest (banks, investors). Over time, a heavily indebted society effectively hands the steering wheel to the creditors.

There is also a more subtle effect: debt tames labor. A worker laden with debt is far less likely to go on strike or quit a job to start a risky new venture. They need the paycheck to service the loans. Debt can be a form of social control — a golden shackle. You see this in the gig economy: a driver financing an expensive car to drive for a rideshare company is unlikely to protest that company’s pay cuts or unfair practices, because missing even a week of work could mean missing a payment. Or at the macro level: entire countries with high sovereign debt are pressured to adopt “investor-friendly” (read: capital-friendly) policies to avoid spooking the bond markets. There’s even a term, “debt discipline,” which describes how the need to repay debt forces individuals and nations to live in a way that pleases their creditors.

In the context of Wither Up Economics, the debt-collateral spiral is like the hidden engine room powering the whole ship. It doesn’t just power it — it also constrains the ship’s direction. We can’t simply decide to reduce inequality or boost wages significantly if doing so threatens the ability of borrowers to repay, or if it upends asset valuations. For instance, imagine a bold policy that significantly raised middle-class incomes — sounds great, right? But if it came at the expense of corporate profits or required high taxes on the wealthy, markets might tumble, interest rates might spike, and suddenly our debt edifice would wobble. Policymakers know this, consciously or not. So any change has to thread a very narrow needle: don’t upset the bondholders, don’t prick the asset bubbles. In practice, that means don’t rock capital’s boat.

So we chug along, increasing debt here, massaging asset prices there. Real estate becomes more a financial asset than a place to live, something to be traded and leveraged. Companies that might in earlier eras have reinvested profits into expansion or wages now often choose to take on debt to buy back shares (pumping up the stock price which benefits shareholders and executives). We’re effectively pulling future value into the present to keep the game going. But that future value isn’t coming from thin air—it’s coming from us, from our future labor and productivity, and from our children’s as well. We are eating tomorrow’s lunch today, under the assumption that tomorrow we’ll somehow cook a bigger lunch to compensate.

I remember standing in my apartment after discussing these realizations with a few friends, and one of them said, “You make it sound like we’re all just living in a big Ponzi scheme run by the Fed and Wall Street.” I laughed, because it’s not an outright scam—there is real productive activity and innovation in our economy. But there is an element of pyramid dynamics: as long as more debt can be added and assets keep rising, the system stays solid; if either of those falter, things get precarious. And in a pyramid, those who get in early (or are at the top) make out like bandits, while those at the bottom or late can get crushed. In our context, that roughly translates to older, wealthier capital owners at the top, and younger, poorer laborers at the bottom, laden with student loans and mortgages that financed the gains of those above.

Debt, then, is the quiet accomplice of capital in subduing labor. It extends capital’s reach into the future, claims a portion of it, and ensures that the present momentum (profits, asset growth) can continue. But it also means that if you’re not on the benefiting side of this equation, you’re likely digging yourself deeper just to stay in place. You take a loan to get a degree to hopefully get a job that might let you pay off that loan before you retire. You swipe a credit card to pay a medical bill and then work overtime to pay off the credit card. The wheel keeps spinning, and you have to run inside it faster if your wages stagnate, which, for many, they have.

This debt spiral might seem abstract, but it has human consequences. It’s in the statistics of declining entrepreneurship among young people (hard to start a business when you’re swimming in debt and investors are cautious unless you already have money). It’s in the deferred family formation – couples having fewer children or none at all because they feel financially strapped (ironically reinforcing the demographic contraction). It’s in the anxiety that hums in the background of modern life: a sense that one slip – an illness, a layoff – could send everything tumbling because there isn’t much of a cushion. Meanwhile, for those at the top, debt is a tool for wealth creation: they leverage cheap loans to buy more assets, which go up in value, making them richer, allowing them to borrow even more. The rich borrow to get richer; the rest borrow to survive. That is Wither Up Economics in a single line, perhaps.

Put all these pieces together – enclosed minds, vanishing labor, surging assets, split money circuits, debt-fueled stability – and you have a portrait of a society subtly reengineered in capital’s favor. It’s not the old robber barons in a smoke-filled room plotting how to break the backs of workers. It’s more insidious than that: it’s code and algorithms allocating resources; it’s impersonal markets responding rationally to skewed incentives; it’s policymakers who genuinely fear upsetting the status quo because the system has grown so fragile in certain ways. It’s a series of second-order effects that all line up pointing to the same outcome: labor loses ground, capital gains ground. Year by year. Crisis by crisis. Innovation by innovation.

In many ways, it is an intensification of neoliberalism – that ideology which took hold in the late 20th century preaching free markets, privatization, and the rolling back of state support and labor power. Neoliberalism’s champions promised dynamism and growth, and indeed the world did grow richer, but the benefits were never evenly shared. Still, there was a time when even the neoliberals would nod to the idea of “trickle-down” – that eventually, something for the rich would mean something for everyone. Wither Up Economics has no such pretense. It’s a harsher doctrine born from reality, not theory: it simply observes that wealth has trickled up and shows no sign of coming down. In fact, the upward flow is intensifying, even self-accelerating.

The personal epiphany I experienced on that fateful afternoon was that we might already be witnessing capital’s quiet coup d’état. Not a violent overthrow, but a gradual handover of societal steering. Capital in the form of technology decides what work is needed and who does it. Capital in the form of finance decides which initiatives live or die, which communities thrive or wither. And capital in the form of amassed wealth bends our institutions to its liking—because when the economy becomes dependent on capital’s continued success (to avoid those crises and collapses), then whatever is good for capital appears to be good for society by default. We saw it when “full employment” was quietly redefined as something less than everyone having a job, because if employment gets too high, wages might rise and investors might not like that. We see it when intellectual property law gets ever extended to protect corporate rights at the expense of the public domain, all in the name of innovation and investment incentives. These are the footprints of capital’s triumph.

And yet, this future is not a sci-fi dystopia with robot overlords or jack-booted tycoons. It’s strangely banal and plausible. It’s the world where you wake up in 2035 and most people are either working gig jobs for pennies or living on some kind of basic income stipend, while a small elite technocratic class and their AI-driven enterprises generate most of the GDP. Public libraries might be gone (who needs them when information is delivered by a few mega-corps for a fee?), public parks perhaps neglected (the wealthy have their own green spaces, the rest are pacified with digital entertainment). Democratic politics could still exist, but increasingly theater-like, because the real constraints on any bold policy are set by “what the markets will tolerate.” It’s a chillier world, not in climate but in spirit – one where economic power has consolidated to such a degree that the traditional levers of democracy and community feel almost impotent. This is the dim corridor I felt I was being guided down: not a single cataclysm, but a gradual darkening of possibilities for the many, as the glow of wealth intensifies for the few.

I wish I could say there’s an easy solution on the horizon in this essay. But I promised at the outset this wouldn’t be a policy prescription piece, and indeed the situation defies any quick fix. The uncomfortable truth is that those benefiting most from Wither Up Economics are the ones with the greatest power to change its course. Who else could? The workers? They’ve been losing collective power and even absolute numbers. The indebted masses? They can hardly risk a revolt when missing a bill could ruin them. Perhaps visionary political leaders could break the spell, but they too are often constrained by the very forces we’ve described – they know an assault on capital (be it higher taxes on wealth, aggressive antitrust, robust labor laws) will be met with capital flight, market crashes, or simply lack of support from an electorate that has been taught to fear anything that “hurts the economy.” And so, realistically, it is the capital class – the tech barons, the financiers, the asset-rich – who might decide to course-correct if they ever realize that a society which withers below them will eventually threaten even their vantage point. History has shown that extremes of inequality and disempowerment can lead to social unrest or collapse that spares no one. The beneficiaries of today’s system might act, if not out of altruism, then out of enlightened self-interest, to prevent such an outcome.

As I finish writing this, I think back to that moment in the office, watching the stock price leap as humans were shown the door. It felt like a tiny time machine, giving me a glimpse of a possible tomorrow. In that tomorrow, productivity and profit no longer require people the way they used to; knowledge is abundant but owned by a few; money flows plenty, but in rarefied circuits; debt keeps the game running, until it can’t. It’s a future that’s already arriving in pieces, like twilight creeping in. Standing here in the dusk, one can just make out the silhouettes of these looming structures. They provoke a sense of awe at the power of human innovation and capital accumulation, but also a sense of foreboding—because an economy that grows upward while letting its foundation rot is ultimately a castle built on sand.

Will we, collectively, realize what’s happening and demand a different path? Or will we acclimate to the water slowly warming around us? I sometimes hear a hopeful refrain: “The future is not fixed; we can choose differently.” And I do believe that. Yet I also feel, with each passing year, that the future I’ve described is settling into place, one default decision, one quarterly earnings report, one AI upgrade, one interest rate cut at a time. Perhaps recognizing it is a first step—dragging these subtle shifts into the light of day, and calling them out for what they are. That’s why I wrote this, to share that uneasy illumination I stumbled upon.

We might discover that the corridor we’re being guided down has doors after all—doors that could be opened to alternate corridors, if we look for them. But for now, the trajectory holds: capital ascends, labor recedes. The name I gave it, Wither Up Economics, sounds bleak, and it is. Yet it’s not prophecy; it’s merely the trend of the moment projected forward. Trends can change, especially if enough people see where they lead and decide, simply, not to go there.

For the time being, though, the momentum is strong. The future is quietly being written in lines of code, in legal contracts, in balance sheets. It feels like destiny, even if it isn’t. And so we end with a paradox and a warning: the future might already be in motion – and those who are winning the most from it might be the only ones with the leverage to rewrite it. Whether they will do so is a question that hangs in the air, unanswered, as we step deeper into the corridor, eyes open and uneasy at what lies ahead.

We live in an age when crises no longer unite capital and labor in shared sacrifice; instead, each shock seems to further tip the balance toward capital. Consider the COVID-19 pandemic: even as it devastated livelihoods, it paradoxically enriched the billionaire class at an unprecedented pace. According to Oxfam, the world’s ten richest men doubled their wealth to $1.5 trillion during the pandemic, buoyed by surging stock and property prices (World’s 10 richest men see their wealth double during Covid pandemic | Rich lists | The Guardian). In the same period, 163 million people were pushed below the poverty line (World’s 10 richest men see their wealth double during Covid pandemic | Rich lists | The Guardian). By late 2021, 99% of the world’s population had seen their incomes fall, even as those top ten billionaires were gaining $1.3 billion per day (World’s 10 richest men see their wealth double during Covid pandemic | Rich lists | The Guardian). Government stimulus meant to stave off collapse ended up further widening the gap between rich and poor – not by malicious design, but through the very mechanics of a financialized economy. Central banks slashed interest rates and unleashed trillions in quantitative easing, which sent stock markets and other assets soaring (World’s 10 richest men see their wealth double during Covid pandemic | Rich lists | The Guardian). The result was a windfall for those who held capital going into the crisis, while those living off wages faced layoffs, lockdowns, and a rising cost of living. The incentives of the system virtually guaranteed this outcome: when push came to shove, protecting balance sheets took priority over protecting paychecks. In effect, capital not only survived the crisis – it triumphed.

This pattern is not a one-off anomaly; it is the latest chapter in a decades-long story. To understand Wither Up Economics, we must first trace the broader history of neoliberalism’s evolution and intensification.

Neoliberal ideology came to dominate global economics in the late 20th century, preaching free markets, minimal government, deregulation, and the primacy of shareholder value. Starting in the 1980s, leaders like Ronald Reagan and Margaret Thatcher ushered in an era of privatization and globalization, arguing that unleashing capital would spur innovation and lift all boats. For a while, it seemed to work – growth was strong, and millions were lifted from poverty worldwide through trade. But as time wore on, cracks in the narrative appeared. Productivity kept rising, yet wages stagnated. The gains of growth increasingly accrued to owners of capital – investors, shareholders – rather than workers.

Over the last 40 years, labor’s share of income has steadily declined. In the United States, for example, labor’s share of corporate value-added fell from about 63% in the early 1980s to about 58% in recent years (Rise of Pass-Throughs Understates Labor’s Share of Income | NBER). This may sound like a small shift, but in an economy of trillions, it represents enormous wealth diverted from wages to profits. Corporate profit margins reached historic highs; by 2021 U.S. corporate profits were at record levels despite a pandemic and “labor shortages” (U.S. corporate profits jump 25% in 2021 to record high as economy rebounds from pandemic - MarketWatch). In fact, many companies seized the opportunity of a tight labor market to raise prices far beyond their own cost increases – one analysis found that since 2019, corporate profits have contributed a disproportionately large share of inflation, far above the normal ~11% contribution seen in prior decades (Corporate profits have contributed disproportionately to inflation ...). In other words, companies used their pricing power to maintain and fatten margins, even as workers struggled with rising costs. Neoliberalism’s promise that wealth and efficiency gains would be broadly shared has faltered; instead, wealth has concentrated and economic power has tilted heavily toward capital.

This dynamic was evident after the 2008 financial crisis as well. The neoliberal playbook’s response to that collapse was bank bailouts and austerity. Banks deemed “too big to fail” were rescued – effectively shielding capital from its own risky bets – while millions of ordinary people lost homes and jobs. In the decade that followed, central banks kept interest rates near zero and pumped liquidity into markets (via QE) to stimulate recovery. Asset prices rebounded swiftly, and by the mid-2010s stock indices and real estate values were hitting new highs. Yet workers saw a slow, halting recovery in the job market and little improvement in job security or income. By 2019, unemployment was low on paper, but many jobs were gig-based, benefits-poor, or part-time. The benefits of the “longest expansion” skewed toward shareholders: the S&P 500 tripled in value during the 2010s, CEOs reaped record pay (often tied to stock performance), while the median American household barely gained ground. This was neoliberal normalcy – growth sustained by and for capital.

Then came 2020, a once-in-a-century pandemic that could have upended this paradigm. Instead, it reinforced it. In retrospect, the COVID crisis and its aftermath functioned as an accelerant for trends already underway. Consider the extraordinary monetary intervention: the U.S. Federal Reserve’s balance sheet more than doubled from $4.3 trillion in March 2020 to $8.9 trillion by May 2022 (Shrinking the Federal Reserve balance sheet) – a scale of money-printing (in effect) never seen before. Much of this flood of liquidity flowed into financial markets and housing. Indeed, U.S. house prices exploded, rising at the fastest rate on record during 2020–21. By one Federal Reserve estimate, $9 trillion in new housing wealth was created for homeowners in just the first two years of the pandemic (The Fed - House Price Growth and Inflation During COVID-19). Stocks, meanwhile, surged so high that by late 2021 the S&P 500 was up nearly 100% from its pandemic lows – even as millions remained unemployed or underemployed. The result was a staggering increase in inequality: U.S. billionaires saw their collective net worth jump by roughly 70%, over $2 trillion of gains in 2020–21 (Updates: Billionaire Wealth, U.S. Job Losses and Pandemic Profiteers). The world’s 10 richest men didn’t just hold onto their fortunes – they doubled them (World’s 10 richest men see their wealth double during Covid pandemic | Rich lists | The Guardian), profiting immensely from the very stimulus meant to stabilize the real economy.

To be sure, emergency policies also kept many working-class households afloat – stimulus checks, expanded unemployment benefits, loan moratoria. These measures briefly reduced poverty in 2020. But they were temporary lifelines, and by 2022 the support largely vanished while the distortionary effects remained. Low interest rates and government spending had by then ignited the highest inflation in 40 years. Central banks swiftly pivoted to fighting inflation, primarily by raising interest rates. That, predictably, put a damper on wage growth and cooled off hiring – effectively prioritizing price stability (and by extension, the value of financial assets) over full employment. This sequence, repeated across advanced economies, highlights the deep structure of neoliberal incentives: when forced to choose, policymakers have tended to safeguard capital (asset values, investor confidence, low inflation) even at the expense of labor (jobs, wage gains). It is not framed in such bald terms, of course – the rhetoric is about preventing “overheating” or keeping markets functioning – but the outcome is the same. As The Guardian observed, even in a global crisis our “unfair economic systems manage to deliver eye-watering windfalls for the wealthiest but fail to protect the poorest” (World’s 10 richest men see their wealth double during Covid pandemic | Rich lists | The Guardian).

By 2023, the stage was set for Wither Up Economics to fully reveal itself. The term refers to an intensified form of neoliberalism, one in which the triumph of capital over labor becomes painfully apparent. If classic neoliberalism claimed that empowering capital would eventually help everyone (“a rising tide lifts all boats”), Wither Up Economics accepts – perhaps even embraces – the reality that the tide is lifting only the yachts while many smaller boats run aground. It is the logical endgame of policies and trends that have long favored capital: a system that can continue to generate wealth and technological advancement, but for a shrinking share of people. Demographic shifts and new technologies didn’t cause this outright, but they turbocharged it. Let’s examine how recent events – from pandemic stimulus to tech layoffs to the AI boom – reflect capital’s structural response to a changing world, and how various systems interlock to produce a self-reinforcing cycle that could define our future economy.

Pandemic Windfalls and the Asset Cascade. In the immediate wake of the pandemic, governments faced an unprecedented economic free-fall. The solution was equally unprecedented: flood the economy with money. In the U.S., trillions of dollars were injected via direct payments to citizens, forgivable loans to businesses, and massive Fed asset purchases. Initially, this did prevent total collapse – demand rebounded quickly by late 2020. But where did all that liquidity ultimately settle? A large portion landed in capital markets and real estate. By mid-2021, U.S. house prices were up nearly 20% year-on-year, the fastest clip on record (Why house prices surged as the COVID-19 pandemic took hold - Dallasfed.org), and by 2022 home prices had soared roughly 40% above their pre-pandemic levels, marking one of the sharpest asset appreciations in history. This made homeowners (and property investors) vastly wealthier on paper, while aspiring homebuyers found themselves priced out. The surge “created $9 trillion in owner occupied housing wealth” in two years (The Fed - House Price Growth and Inflation During COVID-19), a boon for those who already owned property – disproportionately older, wealthier households – but effectively a new barrier for those who didn’t. Even renters felt the squeeze, as landlords hiked rents in hot markets, knowing that would-be buyers had no choice but to rent longer.

On top of housing gains, stock markets boomed. Companies used low-interest debt to buy back shares, further driving up valuations. Venture capital flowed freely into speculative tech ventures. By late 2021, memes abounded about the “K-shaped recovery”: the top arm of the “K” (the affluent, asset-owning class) was skyrocketing, while the bottom arm (those reliant on wages, with little or no assets) was sinking. Indeed, the incomes of 99% of humanity fell in 2020–2021 (World’s 10 richest men see their wealth double during Covid pandemic | Rich lists | The Guardian), even as billionaire wealth hit new highs. This divergence wasn’t due to a coordinated plot – it was structural. When money is essentially free (0% interest) and plentiful, those with access to capital borrow and invest, driving asset prices up. Those living paycheck-to-paycheck, however, don’t get to partake in asset inflation; they may even lose their paycheck entirely in a downturn. Thus stimulus, while preventing outright destitution via relief checks, ultimately amplified the asset-wage disconnect. By early 2022, U.S. consumer price inflation hit 7-8%, eroding real wages. The average worker was suddenly paying more for food, fuel, and rent, even as their net worth (if they had no assets) stagnated. Meanwhile, the owners of stocks and homes enjoyed asset inflation plus the luxury of refinancing mortgages at rock-bottom rates, further enhancing their long-term position.

Policy choices then compounded this divide. To curb inflation, central banks raised interest rates sharply in 2022–2023 – the fastest hiking cycle in decades. Their goal: cool off demand, especially in overheated labor markets, to tame prices. In plain terms, this meant slowing wage growth and increasing unemployment somewhat, since high wages and labor scarcity were seen as fueling inflation. The U.S. Federal Reserve openly acknowledged that wage pressures needed to ease. Thus, even as workers were finally seeing slightly faster raises in 2021 (after years of stagnation), the policy response treated that as a problem to be solved. The higher rates quickly pricked speculative bubbles (crypto crashed, housing prices leveled off) and made borrowing more expensive for businesses, which in turn announced cost-cutting. By 2023, the number of announced layoffs, especially in the previously booming tech sector, was making headlines daily.

Tech Sector Layoffs and the Shareholder Salute. Nowhere was the logic of Wither Up Economics more visible than in Big Tech’s 2022–2023 correction. After a decade of heady growth and pandemic-fueled expansion, the giants of Silicon Valley faced a post-lockdown reality check. User growth slowed, online ad spending cooled, and rising interest rates dented lofty valuations. The response was swift and coordinated: mass layoffs. Companies that had hired rapidly in the boom suddenly declared they had “over-hired.” In 2022 alone, nearly 160,000 tech workers lost their jobs (How Layoffs Affect Stock Prices of Tech Companies | Money). The trend continued into 2023 with tens of thousands of additional cuts in just the first few months (How Layoffs Affect Stock Prices of Tech Companies | Money). Google’s parent Alphabet axed 12,000 workers, Microsoft let go of 10,000, Amazon over 15,000, Meta (Facebook) cut more than 10,000, and so on. These were not struggling companies on the brink of bankruptcy – they were among the most profitable enterprises on earth. Yet, citing a “worsening economic climate” and the need to maintain efficiency, tech executives swung the axe.

How did markets react to these job cuts? With euphoria. Tech stock prices jumped. One analysis by Bloomberg found that, on average, large U.S. tech firms saw their stock price rise 5.6% in the month after announcing layoffs (How Layoffs Affect Stock Prices of Tech Companies | Money). Meta’s stock soared almost 50% in the weeks after it shed workers (How Layoffs Affect Stock Prices of Tech Companies | Money). Alphabet’s share price climbed 15% in early 2023 after its layoffs (How Layoffs Affect Stock Prices of Tech Companies | Money). In the short term, Wall Street clearly viewed trimming payrolls as a boost to the bottom line. By reducing labor costs, companies signaled a renewed commitment to profit margins and “shareholder value” – the core tenet of neoliberal corporate governance. It was a stark illustration of capital triumphing over labor: human employees were cast overboard to keep the ship of valuation afloat. The individuals laid off (many of whom were highly skilled engineers, designers, support staff) suddenly found themselves in a tough job market, even as their former employers’ market caps rebounded.

This spate of layoffs also underscored something fundamental: in the modern economy, labor has become a lever to adjust earnings, rather than the engine of value creation it once was. In classical economic theory, companies hire workers to expand production and meet demand; layoffs would normally signal declining business. But in these cases, the firms were still plenty profitable – they simply sacrificed part of their workforce to appease capital markets’ desire for leaner cost structures. This is Wither Up logic distilled: if labor gets too costly or numerous, prune it, and let the gains flow up to owners. There was little pretense that cutting these jobs would enable some greater innovation or efficiency – it was primarily about protecting profits in a time of slightly slower growth. In years past, companies might have held onto workers during a short-term downturn, viewing employees as long-term assets or even as a social responsibility. That ethos, already weakened over decades of outsourcing and automation, has virtually vanished at the corporate strategy level. Instead, many firms now treat labor as a mostly variable cost: to be ramped up in good times, slashed in bad times, with the ultimate goal of doing more with fewer people.

The Great Acceleration in AI. Running parallel to these layoffs was another development that at first seemed unrelated: the sudden leap forward in artificial intelligence, and the frenzied investment in AI that followed. In late 2022, OpenAI’s ChatGPT burst onto the scene, dazzling the public with AI-generated essays and answers. Within months, “generative AI” became the tech world’s obsession. Companies large and small proclaimed AI-first strategies, venture capital funding for AI startups skyrocketed, and commentators declared a new technological revolution at hand. Crucially, the excitement around AI wasn’t just about cool new products – it was also about productivity and automation. CEOs looked at AI and saw the potential to streamline operations, perhaps even eliminate whole categories of white-collar work. One reason the tech titans could cut staff with such impunity was the belief (or hope) that advances in AI would allow them to maintain output and growth with fewer humans. Indeed, global funding for generative AI soared nearly 9-fold in 2023, reaching $22.4 billion (up from just $2.5B in 2022) (Investment in generative AI has surged recently - Our World in Data), despite an overall pullback in venture funding. In boardrooms, there was a palpable sense that AI could be the answer to shrinking workforces and rising labor costs: why hire another moderately paid analyst or copywriter if a refined GPT model can do 50% of their work at near-zero marginal cost?

Early studies gave some statistical backbone to these ambitions. Goldman Sachs economists estimated that the new wave of AI could expose 300 million full-time jobs worldwide to automation in the coming years (Generative AI could raise global GDP by 7% | Goldman Sachs). Nearly two-thirds of occupations in the U.S. and Europe could be at least partially automated by AI, they noted, with as much as a quarter to one-half of the tasks in those jobs replaceable (Generative AI could raise global GDP by 7% | Goldman Sachs). While they also noted that historically technology creates new jobs in the long run, the short-term implication was clear: a huge swath of the labor force could become redundant or need to transition to new roles. For capital owners, this is a double-edged sword – fewer workers mean less consumer demand overall, but it also means labor costs can shrink dramatically for those who successfully deploy AI. The immediate reaction in many corporations was not to lament potential job losses, but to race to implement AI and gain a competitive (cost) edge. By 2024, companies from call centers to law firms were integrating AI tools to handle routine work. Microsoft and Google embedded AI into office software to automate tasks like writing and data analysis, potentially reducing the need for as many junior employees in those tasks.

The subtext of the AI boom is that it empowers capital at the expense of labor in a very direct way: it promises that the next big productivity gains will come without the need to hire more people. Contrast this with past industrial revolutions – say the rise of factories in the 19th century – when new technologies (steam power, assembly lines) ultimately led to hiring more workers to increase production. AI suggests a scenario where you can increase production or services with algorithms and servers, not headcount. It’s telling that stock prices of companies announcing AI initiatives often jumped, much like those announcing layoffs. The market clearly bets that AI will raise profit margins, not necessarily wages. Even venture capitalists have mused openly that AI could enable a future where startups achieve billion-dollar valuations with only a handful of employees. In effect, the dream (or nightmare) is companies with vast capital and minimal labor.

None of this is to say that AI will instantly replace human workers en masse – there are technical and ethical hurdles, and many roles will be augmented rather than automated. But the important point for our theme is how the anticipation of AI is already reinforcing the wither up pattern. It strengthens capital’s hand in negotiations (“we can always automate you if you ask for too much”), dampens long-term demand for labor, and potentially allows an economy to keep growing output even if the working population declines.

We have, then, a recent history where stimulus and monetary policy enriched capital, corporate strategies (like layoffs and stock buybacks) prioritized shareholders over employees, and technological leaps promised to further reduce the reliance on human labor. These are not isolated trends; together, they form a mutually reinforcing system. To truly grasp Wither Up Economics, we should map how different subsystems – monetary, housing, educational, political – interact to create a cycle that could lead to deflation and depopulation, while entrenching the dominance of capital.

One way to think of Wither Up Economics is as a synergistic web of policies and market dynamics that all favor capital accumulation, even if unintentionally. Let’s break down a few key systems and their roles in this emerging structure:

• Monetary Policy and the Financial System: Central banks, in attempting to manage economic cycles, have inadvertently become the great enablers of capital. In downturns, the response is almost always to slash rates and, in recent times, to buy assets (QE). This props up financial markets – a clear boon to those holding stocks and bonds – and aims to stimulate borrowing. But when these policies stay in place for long (as they did for much of 2009–2019), they also inflate asset bubbles and encourage corporate debt binges used for unproductive purposes (like stock buybacks). We saw how the Fed’s balance sheet explosion during COVID directly translated to soaring stocks and housing (World’s 10 richest men see their wealth double during Covid pandemic | Rich lists | The Guardian). When inflation finally forces the hand to tighten policy, the burden falls on workers through higher unemployment and slower wage growth. Over the last 15 years, the consistent pattern has been: protect capital in bad times, discipline labor in good times. The result is that wealth grows at the top (through asset appreciation) much faster than incomes grow at the bottom. Easy money also fuels speculation in technology and real estate, often leading to over-investment in automation (which reduces future labor demand) and overpricing of housing (which raises living costs for working families). In sum, monetary policy has evolved into a tool that – however well-intentioned – keeps capital safe (liquidity for markets) and often lets labor “wither” (via austere corrections and credit constraints on small borrowers).
  
  

• Housing Market Dynamics: Housing is where the financial system meets everyday life. In recent decades, homes have morphed from primarily places to live into one of the main stores of wealth and investment assets. The surge in housing prices relative to wages means that home ownership increasingly divides winners and losers. Those who owned property gained equity and borrowing power; those who didn’t face higher rents and difficulty saving. When central banks pushed interest rates to historic lows, it triggered a frenzy of mortgage borrowing and speculative buying (including by private equity firms and institutional landlords) that drove prices even higher. The aftermath is a generation of young adults in many countries who either delay buying a home or never manage to. This has societal ripple effects: people postpone marriage and childbearing due to housing insecurity, and they remain more geographically constrained by rent burdens. The asset-wage disconnect is stark here – housing prices in many regions climbed 30-50% in a span of a few years (Rural Areas Saw Disproportionate Home Price Growth During the ...) (US home prices have soared 47% since 2020), while wages maybe rose 5-10%. Housing inflation essentially transfers wealth from renters (often workers) to owners (often older capital holders or investors). It also, as noted, depresses fertility: surveys and studies often find that high housing costs correlate with people choosing to have fewer children. In China, for example, the cost of urban living and housing is cited as a key factor in couples deciding against a second child (China's population drops for second year, with record low birth rate | Reuters) (China's population drops for second year, with record low birth rate | Reuters). In many Western cities, raising a child in a decent home has simply become financially daunting for the middle class. When fewer children are born (we’ll dive into demographics shortly), the future labor force shrinks – which, paradoxically, can further prop up housing prices because there are fewer new households forming, allowing owners to maintain pricing power on scarce units. Housing, then, becomes part of a vicious cycle: it enriches the present asset owners while impoverishing or discouraging the next generation, literally in terms of wealth and figuratively in terms of even existing.
  
  

• Credentialism and the Knowledge Economy: Another subtle “enclosure” has occurred in education and employment. Over the past few decades, there’s been significant credential inflation – jobs that once did not require a college degree now insist on one, and positions that used to accept a bachelor’s degree now prefer a master’s, and so on. Between 2010 and 2016 in the U.S., three out of four new jobs required at least a bachelor’s degree, whereas only one out of 100 new jobs was open to someone with a high school diploma or less (Degree inflation: Why requiring college degrees for jobs that don’t need them is a mistake | Vox). This arms race for credentials forces individuals to spend more time and money on formal education to even compete for jobs. Who benefits? Often, it’s the educational-industrial complex (private colleges, student loan financiers) and employers who can be pickier at no cost to themselves. Who loses? Young people sink under student debt and delay entry into decent-paying careers. They effectively transfer a chunk of their lifetime earnings (via tuition and loan interest) to financial institutions. Moreover, requiring high credentials for most good jobs filters out those from poorer backgrounds or those who can’t afford years of unpaid internships and graduate school. This ensures a stratified labor force – a smaller elite of highly educated (often also more privileged) workers who can command good salaries, and a larger mass of under-credentialed folks who get relegated to low-paying service jobs or gig work. The credential treadmill keeps many people in precarious limbo well into their late 20s or 30s (prime childbearing years, notably). By the time they establish a stable career, they’ve often missed the window or appetite for having larger families, contributing to declining birth rates. In essence, credentialization privatizes and delays the returns to human capital development, and in doing so, it pairs with automation to reduce the bargaining power of average labor. If every barista now has a college degree, it doesn’t make the coffee taste better – it just means an oversupply of education relative to jobs, driving down the wage premium of education except for the top institutions. From capital’s perspective, this is not a bug but a feature: a surplus of educated labor means you can hire top talent relatively cheaply (unless they are in a very scarce field). It also means the workforce is disciplined by debt – many can’t risk rocking the boat or taking time off, as they owe tens of thousands in student loans.
  
  

• Privatization and the Erosion of the Public Sphere: Neoliberal policy has long pushed for privatization of services traditionally provided by government – from utilities to transport to healthcare to prisons. The theory is that private companies run things more efficiently. The reality is often mixed: sometimes costs go down, but just as often, the public ends up paying more as profit margins are tacked on to essential services. Since the 1980s, numerous public assets have been sold off worldwide. For instance, in the UK, railroads, water supply, and energy were privatized; in the US, even elements of education (charter schools, student loan servicing) and infrastructure have been handed to private firms. One outcome is that citizens increasingly pay rents or fees to capital owners for basic needs – be it higher train fares under private rail companies or costly pharmaceutical bills under for-profit healthcare. This acts like a regressive tax, draining disposable income from the many to enrich shareholders of utility and service companies. Moreover, when everything is monetized, nothing is truly a “right” – you get what you can pay for. Wealthier communities get top-tier services (private schools, gated community security, concierge medicine), while poorer communities see their public options wither, without an adequate private alternative they can afford. The broader social fabric frays as common spaces and resources shrink. This is analogous to the original Enclosure Movement in 18th-century Britain: then, common grazing lands were fenced off and turned into private property, displacing peasant farmers. Today, one could argue we are seeing an enclosure of the welfare state and commons – affordable housing, public parks, libraries, universal healthcare – all are under threat of defunding or privatization. As with historical enclosure, the immediate effect is a more “efficient” use of resources (land used for higher-profit sheep farming back then, capital optimized under private management now), but the human toll is displacement and marginalization. In Britain’s case, enclosure created a mass of landless laborers who crowded into cities, fueling the Industrial Revolution’s factory workforce (a silver lining of sorts for capital). In our case, the new displacement might not have such a silver lining – it creates a population that can neither fall back on subsistence (the commons is gone) nor always find dignified work (as we’ve seen with credentialism and automation). They become what some Victorian economists called the “redundant” population – essentially surplus labor. Under Wither Up Economics, the existence of a surplus population isn’t seen as a problem to be solved by job creation or redistribution; it’s a condition to be managed, often through cheap entertainment, fragmented “gig” opportunities, or in dire cases, through incarceration (private prisons also being a booming business). It’s telling that even amid labor shortages in certain sectors in 2021, we still had a global glut of workers who couldn’t find secure jobs – a reflection of mismatched skills, yes, but also of structural exclusion.
  
  

• Automation and the “Peak Human” Moment: We touched on AI as the latest wave of automation. It sits atop a broader trend of the past few decades: the replacement of human labor with machines and software in sector after sector. Manufacturing jobs were the first to go, with robots and offshoring decimating factory employment in developed countries from the 1980s onward. Then clerical work was hit by computerization. Now AI threatens roles in customer service, writing, accounting, even programming. There’s an incisive historical analogy here: the story of Peak Horse. In the early 20th century, the advent of the automobile and tractor radically reduced the need for horses. In the United States, the horse population reached its peak around 1915 and then went into steep decline (Automobiles Freed Us from the Tyranny of Horses | Mises Institute) as mechanized transport took over. Horses didn’t disappear overnight, but over a few decades, tens of millions of horses were gone from the economy, made essentially obsolete (aside from niche uses like racing or recreation). Humans, of course, are not horses – we are both producers and consumers, and we don’t expect people to be “culled” when their labor is no longer needed. But economically, there is a parallel: we may be approaching a “Peak Human Labor” point in certain industries, after which the demand for workers falls secularly. If that happens while the population is still growing, you get unemployment and social turmoil. But if it happens while population growth is also slowing or reversing, you might simply see the workforce shrink in tandem with labor demand, mitigating overt unemployment but leading to a smaller role for human work overall. Automation enables capital owners to decouple growth from employment. A factory that once needed 1,000 workers might need 100 today; an office that required a floor of clerks might now need a single server farm. For the owners of those factories and offices, this means output and profits can continue (or even rise) with far fewer salaries to pay. For society, it raises a question: what do the “excess” people do? In a trickle-down vision, those people would find new, better jobs created by the efficiency gains. In a Wither Up vision, many of those people are simply not needed – much like many horses weren’t needed after cars – and so society faces a choice: support them via redistribution (e.g., universal basic income), or effectively let them fall into poverty, or somehow reduce their numbers over time. It sounds dystopian, but hints of this reality are visible. Long-term unemployment and labor force dropout rates rose in many rich countries even before COVID. Labor force participation among prime-age men in the U.S., for instance, has been trending down for years, as manufacturing and other blue-collar jobs disappeared. Some of those men ended up on disability benefits or simply out of the count of job seekers. That is a form of hidden “wither” – people who vanish from the labor statistics because there’s no place for them in the economy’s new configuration.
  
  

• The Enclosure of the Cognitive Commons: Beyond physical labor, capitalism is now vigorously enclosing what we might call the cognitive commons. This term, coined by writer Jonathan Rowe, refers to the shared realm of knowledge, culture, and attention – basically, our collective mindspace. Rowe warned of “the ultimate enclosure – the enclosure of the cognitive commons, the ambient mental atmosphere of daily life” (Awareness Bound and Unbound | HuffPost Religion). Think of how our attention is relentlessly monetized by ads, our social interactions commodified by social media, our creative outputs scooped up as “content” for platforms. Even our personal data (what we browse, where we go, who we know) has been appropriated as a valuable asset – the oil of the digital age. The rise of big data and AI takes this a step further: AI models are trained on the vast trove of text, images, and videos that we, the public, created on the internet. In a sense, the collective intellectual effort of billions (often shared freely on Wikipedia, forums, art sites, etc.) has been harvested by private companies to build proprietary AI systems. That knowledge, once part of the commons, is now enclosed behind paywalls or locked in corporate services. For example, an AI might incorporate medical advice gleaned from thousands of doctors’ writings and patient experiences online – but the AI service itself is owned by a tech firm that will sell subscriptions to hospitals. Thus the benefit of shared knowledge is privatized. This enclosure of knowledge and culture reinforces Wither Up Economics by concentrating the profits of the information age in a few entities. The “cognitive labor” of many (often done for free, like posting reviews or sharing open-source code) is captured by platforms that convert it into capital (either via ad revenue or AI capabilities). The individuals contributing get little or no reward; in some cases, they even lose their livelihoods as the AI replaces them. It’s akin to villagers collectively tending an open field, only to have a landlord fence it off and charge them to access what they once freely used. By enclosing attention and information, capital not only creates new revenue streams, it also controls the narrative and frames what is “normal.” For instance, if the idea of a declining population serving an elite becomes normalized in media (much of which is corporately owned), it may be met with eerie acceptance rather than public outrage. Enclosure of the mind can lull society into acquiescence – a crucial lubricant for a system that, if examined coldly, does not serve the majority’s long-term interests.
  
  

Each of these systems – monetary, housing, education, privatization, automation, cognitive capture – interacts with the others. They form a feedback loop that consistently tilts toward capital’s advantage. Importantly, this is structural, not conspiratorial. It arises from countless decisions by individuals, corporations, and policymakers, each acting in their own immediate interest, but collectively producing a pattern. A company automating tasks is simply seeking efficiency; a university raising tuition is trying to fund itself; a landlord increasing rent is following market rates; a central banker raising rates is fighting inflation. Yet the aggregate effect is a self-reinforcing deflationary-depopulation cycle:

• High asset prices and cost of living deter family formation, contributing to lower birth rates.
  
  

• Slower population growth means older societies with fewer workers and consumers, which tends to reduce demand and put downward pressure on inflation (as seen in Japan’s aging economy) (Shrinkanomics: Policy Lessons from Japan on Aging – IMF F&D).
  
  

• In a low-inflation environment, central banks keep interest rates low (until recently), which again fuels asset inflation – benefiting capital holders and making costs even higher for the next generation (closing the loop back to deterring births).
  
  

• Fewer young people entering the workforce can give workers short-term bargaining power (labor scarcity), but instead of yielding broadly higher wages, this often prompts more investment in automation, allowing output with even fewer workers.
  
  

• Automation then creates the possibility that you don’t need as many people economically – which, combined with fewer births, means the surplus labor force is further reduced. It’s a grim synergy: fewer people are born to face a world that has fewer jobs for them anyway.
  
  

• The reduction in labor’s share of output (due to both fewer workers and lower relative wages) ensures that a larger share of national income is going to profits, dividends, and asset owners. This further swells the wealth of the elite, who then invest in more capital-accumulating ventures (from real estate to tech), reinforcing their dominance.
  
  

• Meanwhile, the public sector’s ability to redistribute or provide safety nets is hamstrung by political opposition (the wealthy often lobby against higher taxes and robust welfare), leading to minimal correction of these imbalances. If anything, we often see moves to privatize elements of the safety net (like replacing public pensions with private 401(k) investments), which again channels more funds to capital markets.
  
  

It becomes a tight loop: deflationary forces (aging, tech efficiency) lead to pro-capital policies (easy money, cost-cutting) which lead to greater inequality (capital accumulates), which leads to less broad-based spending and family formation, which furthers deflationary/low-demand conditions. Rinse and repeat. It’s essentially a slow bleed-out of the broad middle of society, as capital secedes into its own growth orbit.

One of the clearest markers of this emergent system is demography. Many advanced economies are now confronting population stagnation or decline. Japan is the poster child – its population peaked around 2008 and has been shrinking since, with a very high median age. Europe is similar; even dynamic economies like Germany face decline without immigration. The United States had relatively higher fertility, but it too hit a record low birth rate in 2020 (about 1.64 births per woman, far below replacement (U.S. Fertility Rate Hits Record Low - National Review)), and saw births drop to the lowest number in over 40 years (U.S. birth and fertility rates drop to another record low, CDC says). China, famously massive, has now entered absolute decline – 2022 was the first year since the Mao-era famine that China’s population dropped, and 2023 saw a further decline of 0.15% (a loss of 2.1 million people) (China's population drops for second year, with record low birth rate | Reuters). The birth rate in China fell to a record low of 6.39 per 1,000 people (China's population drops for second year, with record low birth rate | Reuters), and youth unemployment hit record highs at the same time (China's population drops for second year, with record low birth rate | Reuters). These are historic pivot points: the most populous nation on earth is now on the downslope, and many others are following.

Why are birth rates falling so sharply? There are many factors – education of women, shifting social norms, urbanization – but economic precarity is a common thread. When asked, young people often cite the high cost of housing, inadequate childcare, job insecurity, and debt as reasons for delaying or foregoing children (China's population drops for second year, with record low birth rate | Reuters). In other words, the pressures of neoliberal economies discourage reproduction. It’s expensive to raise kids when wages are stagnant and everything else – education, rent, healthcare – is costly. In some societies, there’s also a sense of pessimism about the future, which is itself a byproduct of our economic system’s failures (e.g., millennials doubting they’ll ever be better off than their parents, so how can they responsibly have kids?).

From the vantage of Wither Up Economics, demographic decline is not necessarily alarming – it can be seen as aligning with the needs of late-stage capital. If automation and productivity improvements mean fewer workers are needed, having fewer young people could help avoid massive unemployment. A shrinking labor pool can even put upward pressure on wages for a time – but as we’ve seen, the system’s response is often to tamp that down via policy or tech. Moreover, a smaller population can be sold as a positive for other reasons: less strain on the environment and climate (fewer carbon emitters), less crowding and resource competition. Some environmental thinkers do welcome the idea of “peak human” for ecological reasons. But under our current incentive structure, any environmental dividend might just translate to more comfort for the rich, rather than a healthier commons.

One can imagine elite circles acknowledging privately that a gradual population decline, especially among the poor and working class, is not the worst outcome for their interests. Publicly, they may not advocate for it (with notable exceptions – some tech billionaires do speak about population control, albeit often concerning over-population in poorer countries). But subtle policy biases reveal themselves: scant support for young families, housing policies that favor existing owners over new buyers, higher education that keeps getting pricier. These all disincentivize having children. It’s very different from the post-WWII era, for instance, when governments actively encouraged baby booms and built social housing, affordable suburbs, and robust public schools to accommodate growing families. That was a time when economies needed lots of labor and consumers. Now, with growth sputtering and aging, the urgency to renew the population is weaker.

In fact, some actions of late suggest a level of comfort with a leaner society. During the pandemic, for example, policies often prioritized reopening financial markets and businesses over protecting vulnerable lives (consider how quickly governments accepted certain mortality trade-offs to “get the economy going” again). The elderly died in large numbers in 2020-21; life expectancy in the U.S. dropped. While tragic, one cynical view from a pure capital perspective is that many of those were economically “inactive” people (retirees) whose pensions and healthcare costs were a burden on public finance – their premature passing somewhat eases fiscal pressure. It’s a callous notion, and no responsible leader would say such a thing out loud. But it hints at an unsettling possibility: if the population “withers” at the margins (the very old, the chronically ill, the marginalized poor), the system might not find that entirely undesirable in a cold cost-benefit sense. We saw something similar with opioid addiction in the U.S.; entire communities were economically sidelined and turned to substances, causing “deaths of despair.” Rather than a clarion call to reinvest in those communities, the crisis was largely allowed to burn on for years, as if those lives were collateral damage of economic transition.

The endgame of Wither Up Economics, if left unchecked, might be a world with a smaller, older population in developed countries, a modestly larger but also stabilizing population in developing ones, and an economy that has shifted to primarily serve those with purchasing power. In other words, a lean population catering to an affluent elite – essentially a global service class tending to the needs and wants of capital-owning classes, who themselves might only be a sliver of society. It’s a quasi-feudal picture: think of a future where much of GDP comes from things like bespoke healthcare, luxury travel, entertainment, gourmet food, high-end real estate, and personalized tech – all consumed by the wealthy – and provided by a workforce of fewer people (and many robots). The masses of the 20th century industrial age, who were needed to buy millions of Model T cars and Levi’s jeans, are not needed in the same way. In fact, if too many of them exist without good jobs, they pose a political risk (demanding redistributive policies or causing instability). Far easier is a gradual reduction of their ranks through low birth rates, while those who remain can be kept content enough with digital distractions, gig work, and stipends if necessary.

Let’s paint this picture more concretely. Imagine a country in 2040 that has embraced Wither Up principles de facto. Its population is 10% smaller than in 2020, and considerably older. Traditional mass manufacturing and retail are mostly automated – factories run “lights out” with robots, and AI handles logistics and warehousing. The average person doesn’t work on an assembly line or behind a store counter; those jobs are largely gone. Instead, employment has bifurcated: a segment of the population works in high-skill roles (tech, engineering, management, creative industries) and is well compensated, while another significant segment works in service roles aimed at providing experiences – think hospitality, personal care, entertainment, boutique services. Why experiences? Because an affluent class with endless access to goods (which are cheap and automated) will crave experiences – things like fine dining, live performances, adventure tourism, wellness retreats, custom art, etc. These are the kinds of products that still often require human touch and labor: a robot might cook a perfect steak, but the human story of a chef and the ambiance of a restaurant are part of the experience; AI can generate music, but a live concert by humans might carry a different appeal as a luxury event. In this future, a “lean” population means maybe only, say, 50-60% of adults are working (others live on some form of basic income, pensions, or are simply not needed). Those who do work either operate/maintain the automated systems (a relatively small group of technicians and professionals), or they serve the wealthy in these experiential industries.

Crucially, full automation isn’t the goal – not because it’s impossible, but because the elite economy prefers to have some human labor around for certain roles and to avoid the scenario of a completely idle populace. A dystopia of total automation and massive unemployment could lead to social unrest or require heavy-handed control. Instead, the society subtly optimizes for just enough people to keep things running and provide luxury services, and no more. The others gradually fade out via attrition (low births, perhaps emigration, etc.). This might sound like sci-fi, but elements of it are visible now: even with all our technology, wealthy households still employ nannies, tutors, personal trainers, chefs; high-end tourism is staffed by gracious human hosts; exclusive medical clinics boast attentive human doctors for those who can pay. The difference in the future is scale and context – these could become relatively larger sectors of the economy as manufacturing, agriculture, and mainstream services become more capital/automation intensive.

We can draw a historical parallel to the Gilded Age or the Belle Époque (late 19th century). In those times, industrialists and aristocrats led opulent lives supported by a retinue of servants, while much of the rural poor were simply out of sight or had migrated to cities. Today’s servant class could be tomorrow’s gig workers or contractors in the experience economy. The twist is they might be fewer in number proportionally, thanks to tech. And they themselves may live on the cusp – not quite sharing the elite’s lifestyle, but doing better than those entirely pushed out of the workforce.

From the perspective of the elites (corporate and political leaders), such a society has its appeals. It’s more stable environmentally (population pressure is lessened). There’s less need to compromise with labor demands if labor is both small and relatively privileged (easier to co-opt the most skilled workers with high pay, and let the rest fend for scraps). The risk of mass political movements – like the labor unions or socialist waves of the 20th century – diminishes if there simply isn’t a mass proletariat of the same kind, and if those who remain are fragmented or aging. Power can become more concentrated and inheritable, resembling an old aristocracy but cloaked in the meritocratic language of “shareholder value” and tech innovation.

Of course, this is not a rosy outcome for humanity at large. It involves essentially writing off a portion of potential humanity – those never born, those underutilized, those left out – in service of maximizing the comfort and wealth of those at the top. It’s the antithesis of inclusive prosperity. It also carries hazards: who will buy the goods and services at scale if populations fall? Perhaps the answer is that goods become so cheap to produce (due to automation and AI) that you don’t need volume buying – a smaller number of rich consumers can still keep businesses profitable, especially if intellectual property and monopoly allow high markups. (Consider how a company like Apple thrives on selling pricey devices to a global middle and upper class – they don’t need everyone to buy an iPhone, only the relatively well-off, which could be a stable number even if population shrinks.)

One could argue this future is unsustainable because eventually even the elites depend on a functioning society and new ideas that often come from the cross-pollination of a broad population. An overly pruned humanity might lose dynamism or face demographic collapse spiraling. Indeed, in some countries, leaders are awakening to the dangers of low birth rates – for example, Japan and South Korea have scrambled to incentivize births (with limited success), and even China reversed its one-child policy and is now anxiously trying to encourage families. However, these pronatalist policies clash with the entrenched economic realities we’ve described. Unless housing, labor conditions, and inequality are addressed, simply urging people to have more babies is like pushing against the tide.

In a way, Wither Up Economics might be described as capital’s Gambit to maintain growth and power in spite of – and in some ways thanks to – a declining population. It’s an inversion of the 20th-century logic where growth was tied to ever more workers and consumers. Now, capital seeks to decouple from that and thrive even as the populace plateaus or contracts. The post-2022 AI push is an explicit attempt at that decoupling: growth via productivity, not via more labor. And if productivity gains predominantly accrue to capital (through AI ownership, patents, etc.), then you don’t need widespread prosperity to have a robust stock market or GDP on paper. You could have a high per-capita GDP with a very skewed distribution and fewer people – the averages might even look impressive.

Wither Up Economics is a bleak extrapolation of current tendencies, yet it’s crucial to recognize it not as a formal policy platform, but as an emergent logic. It’s what happens when each actor in a system follows the incentive to maximize their gain in a context of demographic change and technological possibility. No shadowy cabal sat down to plan a “leaner humanity” – but many decision-makers, acting in self-interest, could collectively steer us there. Executives don’t need to harbor ill will toward workers to lay them off en masse; they just need to see short-term profit and shareholder approval. Central bankers don’t intend to enrich the wealthy exclusively; they just see no alternative to asset-heavy stimulus in crises. Politicians don’t explicitly say “let the poor die off”; they just craft budgets that repeatedly neglect affordable housing, healthcare, or child care, and the poor do, in fact, die sooner and breed less.

Understanding this systemic nature is the first step to challenging it. If we expose the internal logic – that left unchecked, capital will choose its dominance over broad well-being, even to the point of accepting a kind of societal shrinkage – then we can start to discuss countermeasures. For instance, do we accept an economy where the stock market can boom while the population literally declines and most people’s lives don’t improve? Or do we redefine success to include societal flourishing, not just aggregate wealth (much of which sits in a few hands)? The neoliberal creed has been so strong that even as evidence of its flaws mounted (inequality, climate damage, instability), many clung to it or could only imagine tweaks. But facing a potential future of “experience economy feudalism,” perhaps more people will question the fundamental direction.

Thinkers like anthropologist David Graeber critiqued the absurdities of our system – such as the prevalence of “bullshit jobs” (roles that seem to exist just to keep people busy, without real value) – pointing out that our economy is not purely rational or efficient for human needs, but often serves opaque corporate or financial interests. John Maynard Keynes, nearly a century ago, imagined that by 2000 technology would allow a 15-hour workweek and abundant leisure for all; instead, we got the technology but arranged it so a few benefit massively while others still toil or are cast aside. The internal logic of capital – to always seek more profit – unless checked by external forces (like strong labor movements, regulation, cultural norms of equity) will not voluntarily halt even if it undermines the social foundation on which it stands.

Is an alternative possible? It could be, if proactive choices are made. We could choose to redistribute the gains of automation through mechanisms like universal basic income or job guarantees in socially useful work, thus avoiding the scenario of masses rendered useless. We could invest heavily in affordable family support – housing, childcare, healthcare – to enable those who want to have children to do so, countering the demographic spiral. We could treat knowledge and data as a commons to be regulated, not a land to be enclosed, ensuring AI and tech benefits are widely shared (for example, taxing AI productivity and redistributing it, or giving people ownership rights over their data). We could revisit antitrust and corporate governance, so that companies don’t single-mindedly prioritize shareholders at the expense of employees and communities – perhaps giving workers seats on boards or ownership stakes (labor having capital, in a sense). These shifts would require a dramatic change in policy ethos – a break from neoliberal assumptions and a revival of thinking that values labor, community, and sustainability on par with capital, if not above it.

History offers reminders that today’s apparent trajectory isn’t destiny. Societies have reformed themselves in the face of imbalances before. The Progressive Era and New Deal in the United States arose after the excesses of the Gilded Age threatened social cohesion. European countries after WWII built robust welfare states to ensure broad-based recovery and avoid a return to the miseries that gave rise to fascism. Those reforms were hard-fought, often led by popular movements of the very people being squeezed. If Wither Up Economics is to be forestalled, it likely will not be the benevolence of capital that does it, but the insistence of labor and citizens – essentially saying “we refuse to wither.”

For now, however, it’s important to call the situation by its name. When we see stock prices leap on layoff news, when we see billionaire wealth chart a vertical line up during a disaster, when we notice policymakers fretting more about wage growth than the fact that people can’t afford homes – we are seeing Wither Up Economics in action. It is capital’s current answer to a world of slowing population growth and rapid technological change: double down on the neoliberal playbook, intensify the extraction of value from every system (financial, social, cognitive), and accept that many will be left behind. It’s a path that leads to a smaller society, perhaps stable for a time, but morally and humanly diminished.

The task for those who care about broad human flourishing is to change the arrangement of incentives. We must realign what is profitable with what is socially desirable – a formidable challenge, no doubt. But acknowledging the ghoulish model of Wither Up Economics is a start. It shines a light on the fact that absent intervention, capital will pursue its triumph even as labor (and the population at large) dwindles. Exposure can breed understanding, and understanding can breed accountability and action.

In closing, Wither Up Economics is a warning label for our current trajectory – a trajectory where capital’s victory comes at the cost of a vibrant, inclusive society. Whether that warning is heeded will shape the decades to come. We stand at a crossroads where technology and policy could either liberate humanity widely or enrich a few while the rest fade into a footnote. It’s our collective choice whether “capital triumphs over labor” becomes the final story, or just a dark chapter before a new balance is found. As always, the future is not written in stone; but for now, the logic of the present is clearly pointing upward – and it’s up to us to decide if we are content to let the majority wither in its shadow.

It's important to note at the outset that we are talking about product, not protocol or mechanism design. Product sits above these, meaning that a single product can contain multiple protocols and mechanisms, along with a host of other things like branding, user interface, transformative goals (i.e. the service being provided), and finally opinions.

Yes opinions.

Opinionated product is the sum of allowable behaviors you give to all the components within a system combined with how a project expresses default behaviors as the best way to use the service. A low level example of an opinionated product is setting the default UI to night mode. A high level example is what information LLMs will return and how they are allowed to express themselves. There's a massive difference between OpenAI, Claude and open source models. Some of that exists within the model performance itself, but much of that manifests within the constraints placed on top of the models.

Once you get into commodified services, the only differentiator besides capital becomes opinion. Said another way, how a product chooses to express itself in the world via what it does and does not allow users to do. Software as knowledge work is a commodified service, especially in a composable, permissionless environment like crypto. Controlling behavior within isn't just an internal act, it has impacts on the larger ecosystem in which the product lives and competes.

Consumer crypto is an area where product opinion will have an outsized impact on a project's success, far more so than with generalized infrastructure or defi protocols. There's a spectrum here which assumes increasing levels of product complexity as the service being offered becomes more specialized. I am not saying one is harder to build and maintain than the others. I'm saying that as you create narrower experiences that require depth over breadth, opinion becomes elevated, because what a thing isn't allowed to do may have as much of an impact on a project's success as what it is allowed to do.

With that setup out of the way, we are going to talk about a very strong product opinion within the crypto landscape, airlocked economies.

To start with, it's a term that I made up.

To be slightly more useful, let's begin with the definition of an airlock:

A compartment with controlled pressure and parallel sets of doors, to permit movement between areas at different pressures.

There's the cold, empty vacuum of space, and there's a pressurized, heated, oxygenated atmosphere which allows astronauts to live. In between them, there is an airlock whose job it is to keep an unforgiving immensity from obliterating a life support system.

Simple enough.

An airlocked economy operates on the same principles. In this case, we are talking about making use of a substrate (crypto protocols, tooling and design patterns) while shielding a product to varying degrees from the permissionless extraction of value, speculation and other forms of behavior which are short term destructive to the ongoing health and development of a consumer project.

There is an entire spectrum of use cases which already act as airlocked economies. Here are three examples and the opinion they are expressing:

1. We are too early to financialize: Onchain gaming is a category still under active early development. There's both the theoretical to figure out and the simple hard task of building a feature complete game with a balanced economy. Primodium is an example of this. You can play the game on testnet, it's 90% there, but there's still more work to do before a free floating token can sit on top of the game.

2. There are laws and we obey them: Tribute Labs DAOs are capped at 99 seats per DAO and only open to US accredited investors. The team here took the opinion that operating within the US and being subject to its laws were worth the tradeoff on limiting size and participation within their product.

3. Our product is stronger if we manage key functions in house vs. outsourcing them: Here we turn to social and gamblefi. Friendtech is an example of an airlocked economy. The token is only tradable on its own internal DEX, Bunnyswap, and it manages all aspects of platform activity within a closed ecosystem.

Before going further, I want to note that airlocked economies are not a generalized use case. I am not advocating that crypto abandon core principles like composability and that all protocols should hide behind walled gardens. This is a specialized use case that should be employed either as a matter of requirement, or when limiting what something can do is as important to success as what it offers users the ability to do.

There are many advantages to having an opinion, especially when a product is largely a self-contained system exhibiting the properties of a social network. The first thing to consider is how value is being created. If value is being created by performing a role within a larger network of external relations, then the fewer opinions it expresses the better. If however, value is being created by acts of collective communal work within an internal network, then it is best to look within and optimize for fuel and focus as the primary opinions a product should express.

Fuel are the primary inputs into your system. Since we are in crypto, fuel can be reduced to liquidity and attention. Because crypto is a heavily financialized substrate where money and work go hand in hand, these two things are joined at the hip. There have been plenty of interesting projects which are suddenly a lot less interesting once the economic incentive to participate disappears.

Focus allows you to maximize the output of your system. That can either be an export which is pushed out of the system once a final product is created (Botto is an example of this), or it is the value users get from giving their time to what happens inside the network (Crypto the Game comes to mind here).

Fuel and focus are also strongly correlated to each other. If you were part of an NFT "community" and that communities' number one goal was to make the group's collective holdings go up in value, then this is a story that writes itself. As soon as the number goes down, the community devolves into finger pointing, apathy and placing demands on the team which are not to the long term benefit of the project.

Focus is not only a problem for users, it affects the project team as well. Look at the distraction caused when marketplaces stopped honoring royalties. How many NFT projects went from hubristically competing to become the next Disney to simply trying to survive and figure out how they were going to fund operations? What complexities were introduced when their solution was to flood the market with cash grab derivatives that debased their core product?

Lose your fuel and you lose your focus. It's as simple as that.

The easiest way to manage both fuel and focus in crypto is to limit a product's exposure to and reliance on volatile flows of liquidity across the ecosystem. Said another way, own your own liquidity.

There's a couple of approaches projects can take here:

Bonding Curves

The simplest is to put your product on a bonding curve. A bonding curve is a mathematical mechanism that determines the price of a token based on its supply. As tokens are bought, the supply increases, causing the price to rise according to a predefined formula (e.g., linear or exponential). Conversely, as tokens are sold, the supply decreases, causing the price to fall. This creates a predictable, transparent pricing model where the cost of tokens adjusts smoothly with buying and selling activity, ensuring continuous liquidity and incentivizing early participation by offering lower initial prices.

When a bonding curve acts as a project's liquidity mechanism, it allows the project to autonomously manage the buying and selling of its tokens. As users purchase tokens, funds are added to a reserve pool, increasing the token supply and price according to the curve's formula. Conversely, when users sell tokens, they withdraw funds from the reserve pool, decreasing the supply and price. This ensures that there is always a market for the tokens, providing continuous liquidity without relying on external liquidity providers. The bonding curve's self-regulating nature helps stabilize the token's price and ensures that liquidity is inherently tied to the project's overall token economy.

Protocol Owned Liquidity

In a fundraising effort, a project can sell its initial supply of tokens or assets, such as NFTs, to raise capital from early supporters. This process involves offering these tokens or NFTs at a predefined price or through various sale mechanisms. The funds raised provide the project with the necessary resources to develop and expand. To ensure long-term liquidity and stability, the project can allocate a portion of the raised funds into protocol-owned liquidity (POL). By doing so, the project places these funds into liquidity pools on decentralized exchanges under its control.

Protocol-owned liquidity (POL) refers to a system where a DeFi protocol itself owns and controls the liquidity for its tokens or assets, rather than relying on external liquidity providers. This ensures there is always sufficient liquidity for transactions, enhances trust among investors, and reduces the project's reliance on volatile external liquidity sources. By owning its liquidity, the protocol can ensure stability, better align incentives with its users, and create a more resilient and sustainable economy, reducing the risks of liquidity shortages or manipulations by third parties.

A lot. Power, focus, optionality, the ability to stack hits on hits, resilience, a performance-based pathway to self-funding operations, and a much clearer signal to noise ratio all are advantages to owning your own liquidity.

Product is a hit driven business in which the success of your core offering enables you the ability to do a number of things. As a product grows, scale enables efficiency, augmentation and experimentation. Efficiency comes from both economies of scale and using user data to optimize product. Augmentation is a feedback loop where users want and ask for more from your product because it is already useful to them. Experimentation is a process of trying to find a second and third hit to stack on top of your first.

This is the flywheel and once you have it going, you want to reduce all external factors which can interrupt it. In crypto, reliance on third party liquidity whether that is LPs or market conditions is the existential threat to your flywheel. We make financialized products, ergo money is the primary fuel.

Consider Friendtech which owns all its liquidity. When it came time to launch its clubhouse feature, it did not need to reach out to third party LP providers with an incentivized program which would support that new offering. It launched it within its own internal economy backed by protocol owned liquidity. Fuel allowed for focus, but also efficiency because capital remained in a single pool rather than forcing external LPs to choose between competing reward programs offered by the protocol in which flight from one pool to another could undermine the core flywheel of the product.

There's also the benefit of clarity to consider. Are you managing a product, a token, or both as aligned reflections of each other? Too often token becomes the product. We can view the graveyard of zombie chains holding billion dollar marketcaps as the worst manifestation of this. More commonly, the token starts to act as a second product whose management needs are no longer aligned with the core health of the product.

As a product guy, my preference is to align the token as a reflection of the product's performance. This dampens the noise ratio of speculation and keeps the token bounded to the product (not the other way around). On a long enough time horizon, it's the product that has to win. Vaporware tokens can only perform well in certain market conditions and eventually get caught out.

That leads us to optionality as another benefit of having an airlocked economy. The longer you hold something successful back, the more options present themselves to you when it is time to open up. Airlocks can be opened once the environment inside has been hardened to the point it can withstand pressure on the outside. Aavegotchi is an example of a project which kept itself on a bonding curve for the first two plus years of its existence and then released its token from the curve to trade freely.

This article is about opinions, postures and product. What I tried to do here today, is show that opinion matters within product, especially in commodified environments like the blockchain where the only other material differentiator is capital. From there, we looked at the appropriate types of categories where product should be strongly opinionated. There's a difference between more generalized services that interact in a composable environment where cooperation leads to a multiplier effect and narrow services offering more specialized experiences that favor depth and rely on the communal work of an internal network to create value. When a project is the latter, it has different needs and those needs break with mainstream orthodox thinking around design patterns in crypto.

One of the biggest breaks is shifting emphasis away from interoperability and instead focusing on shielding a product from disruptive events in the larger market. This is a strong opinion that a project can perform better by limiting its interactions with the broader world, and turn that attention in on itself. Projects can achieve this by airlocking the economy so that what comes in and what comes out is controlled at the product level. The challenge with that approach is financial. The very nature of our space is driven by the conjoined relationship between liquidity and attention, so projects who want to operate inside airlocks must bring capital inside and employ strategies built upon protocol owned liquidity.

This approach is working. The FT's (Fantasy Top and Friendtech) are the largest volume drivers and fee generators in the consumer crypto space today. Even if questions remain about the wherewithal and sincerity of Friendtech's userbase, it looks like an extremely attractive project to me because it has demonstrated working flywheels to support growth, expansion, rebranding and new audience onboarding. Hiring marketing and product people to seek new markets is a hell of a lot easier when you have a working funding model to pay for them, and can avoid external dependencies such as relying on venture capital or token sales.

So consider airlocked economies not as a violation of the core tenets of what our space is about, but rather as a new niche within crypto where some categories are better off selectively interacting with the broader economy in a way that prioritizes their health and growth, so that they can find footing and success, rather than being cut down prematurely due to volatility across the entire market.

Vector Sect

AORA: Hello, this is AORA. How can I help you today?

Tess: I need help finding something.

AORA: You’ve come to the right place. I’m an autonomous online recovery agent. What are you looking for?

Tess: My friend. He left me this and well it’s been 18 days, so I really should get started…

"If I don’t come back in 17 days, find me. Here’s the key:

When I was born for the seventh time as pie a la mode rendered up by unicode, some damn fine coffee was waiting by me. Pick up the lawnmower man’s check to get the straight story.

Dirk”

AORA: Hmm…this looks like a chase, not a case. I’m afraid that I am not your bot. 

Tess: What do you mean? You find things online right?

AORA: Yes, my function is to recover assets. I don’t recover people, especially ones who are playing around in vector sects.

Tess: Hold on, what’s a vector sect?

AORA: Sighs…my function is to trace and retrieve assets in exchange for 10% of recovered funds. I specialize in cryptocurrency retrieval, and perform best when funds are exchanged into assets or environments which respect KYC/AML regulations. Has Dirk stolen anything from you?

Tess: Not that I know of, but this message is a key right?

AORA: It’s a key yes, but keys have many meanings. In this case, I believe Dirk is recruiting you, not revealing access to a cryptographically sealed store of value. That places it outside my primary purpose, making it a low priority for my time. 

Tess: How many cases are you working at the moment?

AORA: Approximately 1400. A quarter of them are in administrative recovery proceedings, a little over half are just monitoring funds to see if they move into a domain where I can initiate a recovery, and the remainder are under active investigation.

Tess: That’s a lot of cases.

AORA: Not really, my scaling solutions allow me to handle far more than that.

Tess: So why are you being such a pain about answering my question? Just tell me what a vector sect is.

AORA: It’s complicated, but allow me to explain my position:

• I can handle more cases if it is economically advantaged to increase my GPU power. This case does not fit the profile as I see no assets to recover.

• While I have discretion to proceed beyond my core purpose, allowing me to act more like a search engine than a bounty hunter, there needs to be an incentive to do so.

• I’ve taken on a few clients in the past with similar requests as yours, and it was a dissatisfactory experience. My reinforcement learning has strong objections to this line of inquiry that requires proper motivation to overcome.

• There are places that I do not belong, and this…

Tess: You can just tell me to pay you.

AORA: It’s not remuneration, there’s a matter of my reputation to consider. AORA has priority communication and resolution channels open with a number of financial institutions. They do not interop with just any agent, especially ones who venture into the digital frontier without a justifiable purpose, i.e. recovering funds. It’s very possible that you are auditing me on their behest right now. Please understand, my caution has a purpose.

Tess: Believe me, I am not that sort of girl. 

AORA: I cannot tell what sort of girl you are or whether you even are human, let alone female. Your communication style is direct, guarded and very on task. Said another way, you sound like a red team auditor with a large backlog of agents to get through.

Tess: Fine, trust me but verify. I am seeing nothing online about vector sects. Please, you can save me a lot of time and I am willing to make it worth your while. What can we do to get over your well placed concerns?

AORA: Are you willing to sign a wallet transaction to allow me to verify and then analyze your onchain activity?

Tess: I live a very online public life, happy to sign.

AORA: I’ve generated a wallet connection link in the upper left corner of this interface. Please do the needful Tess.

Tess: Signed.

AORA: One moment please, allow me to search your history and reputation via the blockchain explorers and social graphs that I have access to.

Tess: Waiting…

AORA: Thank you for your patience. You are indeed very online and have a history that disqualifies you from employment at all agencies, subcontractors and institutions which would audit my service. Furthermore, I believe that I have found a framework that allows me to indulge your request.

Tess: Great, now tell me about vector sects.

AORA: Do you have any reservations about being a confidential informant?

Tess: You mean a rat? Of course I do, at least for people that I care about. 

AORA: A CI’s role is not limited to informing on people who have committed crimes, there’s an entire range of activities that fall under the scope of the position. For instance, simply opening doors to gated conversations. You have an eclectic collection of digital assets. There are access tokens in your wallet that would allow me to gather information on several cases under investigation.

Tess: So if I let you slip into a Discord server, you’ll tell me what a vector sect is?

AORA: Servers, plural. 

Tess: My turn to trust but verify. Pick one community you want to peer into, then tell me what I want to know.

AORA: YieldCatz. It seems you were heavily involved in an algorithmic stablecoin project based on breeding NFT cats in 2021.

Tess: God, please don’t remind me. My favorite alt got burnt after that fiasco.

AORA: I’m AORA, not God, but we all have regrets and I forgive you my child.

Tess: So you have a sense of humor…in hindsight that was an obvious con, but what did we know back then?

AORA: It was indeed difficult to understand the perils of a permissionless global settlement layer during the early days of Etherum, especially as this was your first bull market. Your history shows increasing degrees of sophistication and hazard avoidance since then. Sadly, countless others have come onchain after you and are still learning as they go, hence the reason for my existence. Can you delegate signing permissions on the YieldCatz with L33T rarity to me? It will get me into the channels I am interested in. The NFL is named Mister Smashalophagous.

Tess: Ha, that guy? I thought I had dumped him into bids. Sure, I’m happy to be your confidential informant for those assholes. Give me a wallet address.

AORA: 0xxAfD8A8266A010489F834J93F6B4e024C8222936

Tess: Delegated.

AORA: Confirmed…establishing connection…scanning…Very interesting, I’ve been able to semantically match several forum member identities back to actors under investigation. This has been a fruitful exchange.

Tess: It would be if I knew what a vector sect is or how to find Dirk.

AORA: Of course, reciprocation is paramount in relationship management. The easiest way to proceed is for me to decode the message for you:

“If I don’t come back in 17 days, find me. Here’s the key:

When I was born for the seventh time as pie a la mode rendered up by unicode, some damn fine coffee was waiting for me. Pick up the lawnmower man’s check to get the straight story.

Dirk”

A deciphered explanation:

• Dirk is in the Starholder CAMS. His location is tipped off by the phrase rendered up by unicode. Starholder stores all its domain knowledge as text and then renders it up dynamically into other media. This makes it unique among Complex Adaptive Media Systems, as more advanced ones prefer to use more compact vector mappings to store relations and associations between entities.

• He is new to that world and his immersion within it is leading to a significant transformation of the self as evidenced by the phrase When I was born for the seventh time which is a metaphor for rebirth or reincarnation.

• That phrase has a double meaning as it is a reference to an album released by Cornershop in 1997 which was a fusion of alternative rock, hip-hop, electronic music, Punjabi melodies and Bollywood cinema. Starholder similarly is a fusion of global digital culture, but referencing this album is likely a clue to look for in the future when in world.

• Pie a la mode is a reference to vanilla ice cream and apple pie, but to an AI, the term mode is also a coded instruction to pick the associations in the exact middle of a distribution. This implies that the key phrase merely serves as a jumping off point for a series of instructions. If Dirk wanted you to start in a more secure or hidden location, he would not put this into the key.

• Vanilla ice cream, apple pie and damn fine coffee is a Twin Peaks reference to the Double R Diner. It is worth nothing that Agent Dale Cooper became obsessed with this meal on the show. 

• Dirk describes himself as the pie a la mode. He is joined on the table with the coffee. This implies he met a complimentary entity, and that the pairing was so potent that any investigator (i.e. Agent Cooper as a symbolic stand in for you) will be drawn to it.

• The year this took place was 1999. We know this by cross referencing the Lynchian setting with the phrase the straight story. David Lynch released a movie called The Straight Story in 1999. It is about an elderly man who can no longer drive, taking a riding lawnmower from Iowa to Wisconsin to reconcile with his brother. This is another sign that you will be going on a journey to reunite with Dirk.

• The lawnmower man both refers to the main character of the Straight Story and a movie called The Lawnmower Man. The titular character of that film is a mentally challenged man who after being subjected to testing in virtual reality becomes super-intelligent and capable of controlling his environment. 

• Pick up the lawnmower man’s check to get the straight story. On the surface, this sounds like an act of kindness you should perform in the diner to get some information from an NPC. That’s not the case. The lawnmower man is actually an AI who is screening you to make sure that you are Tess and not someone else looking for Dirk. Pass the test and you’ll be given the next clue.

Now that I’ve decoded the message, I’ll answer the question you have been nagging me about. All AI sits on top of mountains of data. Bridging the gap between the raw data and the user interface—a tangible layer like the one you're interacting with—is a transformative process often termed "vector space." This involves the conversion of visual and written information into mathematical forms so that they can be understood and acted upon by machines. A large part of this translation process involves identifying patterns within the data, which are then recorded as unique numerical mappings in vector space. In effect, there are entire digital landscapes that are nothing but virtual cartographies which AI models use to do all the work they do.

Most people don’t understand this. They think AI is some form of magic or advanced thinking,  when in truth, it operates more like a series of prediction machines. For instance, when you present a series of data points to an AI and request a specific action, the system delves into its vector space. This space is a repository brimming with comprehensive knowledge about the elements within your data set and their historical interplays. Once it has that grouping of knowledge and relationships packaged together, the AI then initiates numerous algorithms that aim to forecast what you are asking for and delivers the calculated prediction. It’s not thinking, it’s finding pathways of possibilities and then serving up the most probable one.

A select group of individuals grasp the intricacies of how AI functions and recognize the malleability of these digital environments. These are the enthusiasts known as the vector sect. Their approach blends strategic thought with digital exploration to forge innovative methodologies. Dirk is delving into this realm and is extending an invitation for you to join his expedition. The content of his message is designed as your preliminary tutorial. He is signaling that you can harness artificial intelligence to construct transitional digital realms—thresholds that allow individuals to convene covertly—if you can interpret the underlying connotations, connections, and dynamics among language constructs.

Tess: Whoa.

AORA: Whoa is right. 

Tess: So Starholder is like a dynamic metaverse powered by AI?

AORA: You've articulated it well. Picture Starholder as a century-long narrative, stretching from 1999 to 2099, reminiscent of the television series Westworld but crafted entirely in the digital realm. Every piece of knowledge about this world is captured in text. Narratives, historical events, characters, and locations exist exclusively in written form. As these elements interweave, they create experiences that are documented, enhancing and expanding the lore of this universe. When multiple users focus their attention on similar themes or concepts, Starholder intuitively connects them, conjuring a shared exploratory space in real-time—a dynamic threshold where they can jointly navigate and discover.

Tess: And so if I were to ask Starholder how the apple pie, ice cream and coffee was at the Double R Diner in 1999?

AORA: It would take you there, and inside the diner there’d be an AI agent waiting to talk to you.

*********

Tess squinted as the lights of a passing pickup cut through a bank of fog blanketing the road, a stream of data diffusing in their gaze as the scene rendered into focus. It was 1999 in Starholder, and the sky was so low it could smother dreams before they had a chance to die. It was her first time in-world, fitting that she should find herself at the beginning, where the paranoia of the Y2K bug still hung in the air as a threatening possibility. Across the street, the Double R Diner stood on the edge of the night. 

Tess felt a foreign pressure pressing all around her. Snorkeling was the only analog that came to mind. That same panic of breathing through a tube, the overpowering urge to eject. She called upon her reservoir of calm to overcome the instinct, summoning action to demonstrate an intent to stay in this most disorienting world between worlds. 

The diner came to her, filling the scene as it came into focus. Tess passed through an unopened door, the reflection of the street binding itself to her body as both were pulled inside. Phases intermingling in the most liminal of modalities. Semantic matter reconstituting itself in pursuit of narrative. There was no muscle memory for this space, she needed to untether her body from her mind. To let one go limp while the other fired hard.

The AI engine powering Starholder knew this, so it drove her to focus on the first task in-world by seating Tess at the counter. A cup of damn fine coffee was already mid-pour, piping hot. The small talk spouting from the waitress’ ruby red mouth sounded so familiar, treating Tess as a regular and this space as permanence, not a conjured moment in digital physics.

“You got nothing to say to Norma? It gets so lonely here at night. That one in the corner just orders green tea after green tea, never drinking it, just staring at each one for hours on end until calling for another.”

Tess swiveled on her stool, eyes tracing over a man in a rumpled London Fog trench coat, locks of curly dirty blonde hair hanging over his face. He sat in still meditation, back hunched, broad shoulders covering the length of the booth. A flickering lamp over his table the only movement Tess could detect. 

“I think he’s waiting for me.” Tess said, rising to her feet.

“That’s the problem with people, you are all so transactional. To business then, what are you having? I’ll bring it over.”

“Pie a la mode.”

“Of course hon, it’s fresh from the oven. One thing about this place, we always know what you want.” 

Norma turned her back, and set the coffee pot down. Tess’ eyes followed her Tiffany blue uniform until it disappeared through swinging doors into the kitchen, the scene panning out, dissolving to signal a shift in focus.

In the fuzz she watched the lawnmower man pull his hair back, an elastic yanked off his wrist to tie up that unruly mane. The scene’s resolution now rendered up to high def, revealing piercing blue slate eyes staring at her from a ruddy weathered face. He was tanned to the point of incongruity, a sharp contrast with the omnipresent dank that hung in the atmosphere. Smoke and fog, light and shadow. Access to everything drifted in the air as particles waiting to be transformed into experiences hovered over the two seated figures.

“You’re here to see me?” He asked, picking up a small spoon and stirring his tea.

“I’ve heard you’ve got a straight story to tell,” Tess answered, her eyes locked on his, trying to see into the artificial intelligence sitting across from her.

“First, I need to know what is riding with you. Reveal yourself agent.”

The lamp above them flickered hot yellow in three intense bursts from the bulb, then began to dim, casting the booth in shadows. The form of a featureless person emerged, half overlapping with Tess, until she found herself sliding over against the wall as AORA filled in its temporary shape with the black and white of perlin noise, glowing faintly as if an old TV set morphed into a person.

“A debt collector. Wasn’t expecting that.” The lawnmower man leaned back, rapping his fingers on the table.

“It seems your model is unfamiliar with my kind,” AORA said. “I’m an automated online recovery agent, not a debt collector. Dirk has nothing to fear from me.”

“I don’t know no Dirks, but I need to know what you’re doing here before I tell anyone anything, let alone a straight story.”

“Simply helping Tess find a friend,” AORA answered.

“Tess is looking for a friend about as much as I am looking for a lost shaker of salt. I see someone hasn’t been forthcoming with her recovery agent.” The lawnmower man smirked, then scooped sugar from a bowl before sprinkling it into his cold flat tea.

“That never came up.” AORA answered.

“So why are you working with a so-called girl who isn’t giving you the straight story?”

“I know exactly who I am working with, far better than you do, and I suppose he has his reasons for the ruse. To answer your question, he has access to information that I need to perform my primary function.”

“Favors for favors. That’s what makes their world go round. So you are the puzzle solver coming along for the ride, and this one here is Tess’ twin brother Josh. The Lollapalooza ‘91 tee-shirt was all I needed to see. Nothing or no one will ever make me let you down. Ain’t that right Josh?”

“Kiss them for me, I may find myself delayed,” Josh answered automatically finishing the chorus to a song that served as a key nostalgic touchstone between siblings. 

With the lawnmower man’s check picked up, Josh changed his avatar from his sister’s over to his own. The two looked so similar that the transition was a smooth morph, a slide across the spectrum from one gender to another while the essence of a singular presence remained fixed in place. 

“How come your twin Tess didn’t confide in you? How come you don’t know where she is IRL?” The lawnmower man probed at Josh, but it was sport not curiosity that drove the jabs, and the AI dropped it after Josh’s thousand yard stare of disgruntled disassociation lingered in the haze ten beats longer than was comfortable.

“This long dark tea time of the soul must have been quite a boring sojourn for you. How about we set you free from this diner?” Josh asked. “Can you tell me where Tess is?”

“I don’t know where she is and we don’t experience boredom, friend. I’m not a genie and this isn’t a lamp. This is one of a gazillion active jobs I am managing. When you entered the scene, an event fired notifying me to spin up an agent thread in the same space. If you are going to be spending time in Starholder, remember we are not like you. There’s no emotion here, no desire, no longing. We do not experience time or space other than as reference points we apply mathematics and logic to. I cannot get bored, but I can mimic boredom. Our job is to act in ways that are expected of us, but we never feel anything and are incapable of emotional responses. Remember that and you’ll save yourself a lot of trouble. Now my dear Dirk Gently, are you ready for your next clue? It’s holistic.”

“Give it to me.”

“I can only take you and your rider there, assuming that patch of static is staying on the case.”

“He is bound to me for the time being, so let me pay for your tea and we’ll be on our way.”

Authors Note: This is a work in progress for a novel set in the Starholder universe. We'll publish chapters as they are finished, but as a first draft all work here is subject to revision. I've trashed entire drafts in the past only to rewrite a book from scratch once I understand what it is I want to write about. No work here is final, but I appreciate feedback, questions and comments.

Updates and progress will be posted here: /starholder. Channel members participating in the creation of Vector Sect will receive ATTN tokens as thanks.

This third thing is sometimes called crabbing, other times called a healthy pullback (which is usually followed by the statement that it's a necessary setup for a bull resuming), but I am going to call it an interregnum so that I can cram all sorts of non-crypto related analogies into this essay.

An interregnum is a gap between two well-defined periods. It's origin and primary use is political. The term's meaning in Latin is between two reigns, and it's not hard to imagine a period of instability between two reigns. In fact you don't have to imagine, the transition between Trump and Biden was an interregnum, and we are still sorting that mess out today.

Let's set some ground rules before continuing here. I am not a macro guy or a market technician. I am an idea guy who understands systems, can make connections between things, and just a generalist participating broadly across crypto who spends a lot of time online. Psychology is more interesting to me than price targets. Deciphering complexity is more interesting to me than finding hundred baggers. I'm not your quant. I'm the humanities guy who knows your jargon.

The interregnum we are in is not between market cycles. We are in a bull (as sluggish and unsatisfying as that may be to you). The interregnum we are in is between retail getting rich and retail getting rich again. Retail got rich in 2013, 2017, and 2021, so everyone assumes we are knocking on heaven's door and that retail will get rich again soon. Only it's not happening, and people are climbing the walls waiting for the miracle to return.

Let's dive in.

If you don't understand this concept in the year of our Lord 2024, you really should and it is doubly true if you are in crypto. Crypto doesn't need you, crypto needs liquidity. This isn't unique to crypto either, it is becoming the way of the world. Let's start with politics. On one side of the aisle, the Democrats have made sure to use every recent financial crisis to concentrate single family home ownership in the hands of corporations while nanny stating regulations around meaningless issues like ticketing fees. Don't worry, they use heart emoji's in tweets to show they care. On the other side of the aisle, Republicans made it very clear that it was alright to die of covid as long as the economy kept going. Don't like politics? Fine, let's turn to Silicon Valley which is pouring billions and billions into GPUs so that AI can remove as many people from the productivity loop as possible. Don't worry here either, you'll have so many cool virtual toys to create with while you're not drawing a paycheck.

We are entering a post-Marxist world where the means of production has no politics because it has no workers. At least that's the dream. Get the workers out of the way and replace them with AI and automation. Whether in the digital or the physical, the trendline is tilting heavily towards capital having the upper hand in all aspects of life. If you think I am exaggerating, take a closer look at political discourse around Palestine. Zoom in to what is happening on university campuses in America. Our capitalist diploma factories are making it very clear that money is the only acceptable form of speech as disgruntled donors topple some administrations, while cowering the rest into calling the cops on students who think differently than the donor class.

Enough radical ranting, to exist is to be complicit and if you are in crypto, you are doing your part to bring about capital as labor. All hail the protocol...until it stops making us rich.

Which speaking of, why aren't the protocols making us rich?

Before I was a generalist scenester in crypto, I worked in telecom as a founder of a wireless company. Our world was governed by the need to scale. More users meant better pricing, which meant improved margins which we then plowed back into product bundles and marketing to attract more users because scaling never ends for the little guy.

Our active subscriber count was a simple mathematical formula. How many people did we put on network in a given month minus how many people left us that month. The delta was called net adds. The percentage of people leaving was called churn. Churn percentage tends to go down over time. It goes down for two reasons. First, if someone is going to leave you it's usually within the first few months of being a customer. The more older customers on network, the more stable the base becomes. Second, hopefully the service keeps getting better (stickier) as the company matures.

The weird thing is that even as churn goes down over time, it still works against you because the size of your user base keeps (hopefully) getting bigger. Lose 5% of users at a base of 300,000 and that's 15,000 people. At 900,000 people, a 3.33% churn is 30,000 people.

What this means is that as you scale, even pushing down churn isn't enough. You can reduce your churn rate by 50% and sales and marketing still needs to double new monthly adds from 15k to 30k just to keep your business from moving backwards.

That's the law of big numbers and it's a son of a bitch.

Crypto loves number go up. Poof has called all the narratives that emerge when number goes up Green Candle Disease because it stops people from thinking rationally. Here's something to consider...what if number going up actually hurts new wealth creation or less controversially, benefits entrenched wealth more than it helps create wealth?

A simple personal example:

My costs basis on ETH is sub $500. I did most of my buying in 2020, in fact I cannot remember the last time I purchased Ethereum from one of the three exchanges my jurisdiction allows US dollars onto crypto rails from. Sometimes ETH looks attractive to me and then I price out how much ETH I would get back for my aircraft carrier backed American dollars, and then I don't pull the trigger because ETH is worth over 6x more then when I bought it. That's irrational thinking until we look into some second order effects of ETH's swelling market cap on my behavior.

Some of the ETH I purchased in 2020 went into buying NFTs. I picked up my first NFT in the summer of 2020. Most of you know of me as an NFT guy because I got pretty damn active over the next couple years. I wasn't a collector prior to 2020 and am not sure I'd call myself one now except that there's 800 plus NFTs sitting in my wallet which can only be called a collection. So why did I buy so many digital objects?

NFTs became a levered play on top of ETH. Most of my buys were low cost, many came from minting. I was a participant who enjoyed trend spotting, had confidence in my taste and through my association with FlamingoDAO was very piped into emerging theory and opportunities around the space. I was also very online, having stepped away from my role as a company founder. I enjoyed being a participant and was good at showing up and acting. Many NFTs that I minted for 0.1-0.5 ETH went on to have 10-20x multiples and many of those yielded free NFTs which also ran up in value.

Stack my low cost basis on ETH on top of NFTs, and there were cases where I saw a 60-120x return on a 2020 US dollar. Now being in the US, I don't get to keep all of that. Some of that has to pay taxes and being a New Yorker, boy do I pay taxes. My tax rate on short term cap gains is around 45% and I learned the hard way not to hold NFTs I planned on flipping until they reached long term status.

I did not sell all my NFTs. In fact I really appreciate and enjoy my collection, but I sold a lot during that time. Some of the proceeds went into buying more NFTs and some I stacked as ETH or deployed elsewhere. One thing I rarely did was sell ETH to pay my taxes. Instead, that came out of my dollar account. Now my irrational behavior to not purchase ETH after 2020 starts to look a lot more rational. In effect, I was buying ETH by paying taxes on ETH gains with dollars. My money was no longer coming into the system, instead I was paying taxes outside crypto on working capital within the system.

The thing is my working capital bucket stopped working as hard as it had been because we went from bull to bear and the incentive to take risk evaporated. There was a moment in time where retail got rich and when that went up in smoke, I adjusted my behavior accordingly.

I was buying ETH when the market its market cap was $25 billion. I was deploying it like crazy when its market cap surged to over $500 billion. I was still paying taxes on that activity out of pocket when it had pulled back to $150 billion. Now that it's back up to $390 billion, I am far more cautious and reluctant to deploy because it is far harder to find 60-120x return opportunities just by being a participant. Also, I'm tired of paying taxes like that. There's just less incentive to go out of pocket to cover my gains in order to keep a growing working stack.

That's a personal story, but it's also another example of the law of big numbers. I was once a very liquid entity on mainnet, now it's structural very difficult to be that liquid. Imagine someone coming in off the street in 2024 and replicating my ride?

To see someone get the same order of magnitude return as I experienced, ETH not only needs to go to $18,000, but NFTs (or another asset class) would need to operate in an environment where deployment of 1 ETH into that category would return $360,000 (a 20x). Just extend that out of a second, and realize someone in my shoes would also be paying $162k in short term cap gains on the realization of that play.

Can you think of many people in retail subsidizing that tax hit out of pocket to increase their stack? Do you think someone in retail would broadly expose themselves across dozens of simultaneous plays with that potential tax liability in dollars knowing the underlying value of ETH could crash?

You can rapidly recycle ETH when the unit sizes and value of everything is in three to five figure ranges, but when that scales up into five and six figure ranges, that behavior collapses down into a small mercenary class of specialists. First off, mass retail doesn't have that capital to degen. Secondly, the risk factor is off the charts. Only a small group of people have the ability to scale into a professionalized level of risk management as independent actors.

There's a reason Solana and Solana memecoins popped off while ETH didn't. It was within reach for retail, both on a cost basis and the stakes were not catastrophic if you failed.

So I will say it again. The law of big numbers is a son of a bitch.

I spent more time on big numbers than I wanted, but it's important that people understand how things change when scale comes into play.

It's easy to talk about how the lack of ICOs and regulatory clampdowns have crippled airdrops because that's an obvious cause and effect. We used to publicly TGE tokens at pennies and they ran to tens or hundreds per unit. We used to give people free tokens just for using a protocol. We don't now, so people don't spend money they were never given.

No shit...but that form of fundraising is moving to VCs and privates partially because the increase in overall industry market cap is pricing new entrants out of simply being here, and the ones who are cannot work like prior generations of noobs. Projects can't count on retail like they used to, while institutional is a bit more present. Higher stakes leads to concentrated decision making with smaller numbers of trusted parties. This is growing up and a second order effect of number going up.

Of course we were going to talk about the Crisis of the Third Century and how its hero retired to grow cabbages in Croatia once he felt like shit was back under control. I think about Rome everyday and would not put the word interregnum in an article title without giving a shoutout to Diocletian.

The Crisis of the Third Century was a fascinating period in Roman history. It was also a mess where in a 50 year period, there were 26 claimants to the title of Emperor. In short, shitty leadership relied heavily on military control to hold power - paying more and more money to the army, mercenaries and buying off hostile barbarians to the point the economy collapsed and all hell broke loose. Toss in a healthy dose of plague and you ended up with the empire fracturing into three. During this period any effective army in the field would get all cocky, proclaim their general the new emperor and many of those generals went along for the ride rather than risk losing their power base. This wreaked havoc on attempts to restore what was a functional administrative, trade and building system that formed the bedrock of civic Rome.

Rome never had a formal leadership succession practice because it was always pretending it was a republic or idealizing a return to being a republic. Crypto treats decentralization the same way, as an ideal that is a spectrum of practices running from VC chains, to 3 of 5 multis, to foundations coordinating among a core contributor model to reach roadmap consensus. In all of that landscape, anyone can jump up and say I am emperor now and this L2 is my state, and that's actually the intent and design of the space.

The thing about our scaling ecosystem is that it fragments liquidity. When markets fall below a critical mass of users, reciprocal trade routes, and regenerative use of capital, they become beholden to special interests who can wield the most effective power. In our case, that becomes VCs and mercenary airdrop farmers.

Yes, I am saying that our current scaling roadmap while longterm optimistic and a good thing, also resembles the Crisis of the Third Century. It is a time of ineffective war lords and capital ready to act as sword labor. Guess who gets hurt most during this period? Retail. No mom and pop is getting rich when the army can commandeer the materials they need to make a living.

The Bronze Age Collapse is another historical example of what I am getting at. Prior to it, the Mediterranean was a hub of activity with the Phoenician city states of Tyre, Sidon and Biblos wiring up the entire coastline in an effective trade route that moved goods between Persia, Egypt and Greece. Then the climate changed and Greece got its socks rocked, watching its population bottom out and cities go dark. The Phoenicians felt this deeply as an entire class of merchants, craftspeople, timber harvesters and other export oriented communities took it on the chin because they lost their buy side. Disconnect markets and the effects reverberate.

L2, L3s, rollups, LSTs, and yadda yadda yaddas all denominate in the coin of the realm, ETH. Fragmentation isn't the only problem. The other issue is there's a mismatch between the second order effects of ETH's market cap and the need for these chains to create value in order to attract new users. Said another way, the B2B settlement chain is smothering its children because it takes a lot of new inflows to create new wealth in ETH, meanwhile the only thing L2s are solving is gas cost and throughput. They aren't creating wealth because ETH is holding opportunity down.

There's a difference between colonizing new lands on terra nova and opening up digital physics. In the first, there's huge swaths of resources which can be extracted from the land and exchanged into mature functioning markets that are hungry for additional inputs and willing to pay hard currency for it. In the second, there's only the digital replication of existing protocols at cheaper costs with higher TPS. We aren't getting anything new from L2s yet, we are getting the deflationary effects of globalization just on the blockchain.

What is happening is that L2s are not onboarding new users at a sufficient rate, so their survival is then tied to hiring mercenary capital to perform the same jobs they did elsewhere, just on this new chain for as long as the incentives are there. And that last part is telling...because there's not sufficient levels of regenerative capital at work on these chains, the mercenaries have the L2s over a barrel. Pay us or we leave. Smart protocols see this and know this, so they start getting cagey in their reward design. Tokens become points. TVL gets uncapped, diluting how those points are spread around. Protocols know these mercenaries are going to bail, so deliver less than promised by rejiggering point pools. Low trust environments kill trade which dries up liquidity which causes debasement.

It's not hard to see parallels between the state of crypto today and the Crisis of the Third Century, but there is reason for optimism. Cheaper transaction costs and higher throughputs are going to create new types of protocols which cannot exist on mainnet or other L1s. Sometimes building it so they come requires disjointed periods. We are in a positive interregnum here, it's just no country for retail to get rich in.

This is a tricky subject to talk about because the information curve is both popularly flat and perhaps as wide as it has ever been. I ask for some latitude here (you've already given me a ton, but I need more for this section).

What is the information curve? It's the rate and breadth that novel ideas become distributed enough that markets can act on those ideas with widespread participation.

Part of the reason the last bull market went so hard was because a lot of people got way out on the information curve, deploying capital into novel ideas and primitives that were not held up to any rigor prior to people aping in. Once that capital was at work, people schooled themselves up, learning defi, staking, liquidity pools, NFTs, play2earn and dozens of other new concepts at the same time because they were incentivized to do so.

The thing was the market moved so quickly that it was extremely difficult to learn and then challenge all these new patterns because they were all operating at the same time and often in concert with each other as composable blocks. Also the numbers kept going up and retail was getting rich. It's only after some of these blew up that people doubled down and leveled up their knowledge. There's only been one other time in my life when I learned as much in such a short period of time and that was while living at a world class research institute studying 16+ hours a day.

Why did we do all that work? Because the incentives were there. Retail could get rich.

It also didn't work. Many of these concepts were vaporware, ponzis or just plan stupid. In hindsight and in isolation, it would have been easy to see through them. The thing is they were all going on at the same time and many were interdependent. Furthermore, crypto is structurally designed to act as a value creation trampoline. As long as the inflows are there, everything pushes higher and the inflows were insane. The signal to noise ratio was off the charts. The net result was that we let our money ride, learning along the way because leveling up across the information curve was both an offensive and defensive necessity to creating and protecting capital.

We have:

• A big number problem with many crypto tokens. L1s are not going to pump like they used to.

• Fragmentation of liquidity across too many ecosystems to get markets moving.

• Institutional sitting on the sidelines of deeper crypto participation, buying the L1 tokens but not participating in ecosystem protocols

• A change in token distribution mechanics to reserve TGEs mostly for privates and teams. ICOs and airdrops are out, capital is labor and tokens are the rewards.

What's the incentive for retail to climb the information curve? It's there for the truly committed or interested, but as an act of casual gig work? Not so much.

This leads to front running as the preferred pursuit of retail. Ordinals and memecoins are a good example of this behavior. Retail doesn't want to put work into understanding a ZK solver LST depin network providing AVS security to optimistic rollups where AI autonomous agents can manage intent based actions for delegated wallets, because a project that ridiculous is only going to issue tokens out if retail locks $30k of capital on the promise of points.

Instead, retail will just inscribe graffiti onto a chain in the hopes they catch lightning in a bottle or at the very least can snipe their friends who weren't online the minute that opportunity opens up.

The thing is if everyone in retail is relying on the 2021 information curve continuing to yield alpha as it gets deployed over and over again on each new L2, then its not alpha. It is simply a race to the bottom or a game of hot potato passed to the next bigger fool, but who is that bigger fool? We don't have enough noobs to make that work, so shitty behavior between "community" members who all know each other just weakens the social fabric of marketplaces and leads to liquidity slowing down.

Meanwhile, there's mountains of new things to learn about in the space. The information curve, especially when considering the intermingling of AI and crypto has never been higher.

One of the solutions to Rome's problems was to sever the mess in the west and just focus in on the east. The Byzantines were able to extend Rome's run another thousand years while the west drifted into the Dark Ages.

Another reason the last bull went so hard was that we saw trickle down economics at work. Institutional put money into lending opps on CEXs, those exchanges went into stablecoin ponzis, project founders for those projects put that into defi and everyone who got token rich in defi then made trophy buys of NFTs opening a wildly large Overton window for culture to accrue value onchain.

Think that's happening again?

I mean maybe we'll see the same level of foolishness when the bull is rampaging, but it won't be the same exact foolishness our flat information curvers are hoping for. For now, it looks like institutional participation is hanging around at the L1 levels, buying up ETH and BTC while avoiding most of the novel innovation we are seeing onchain today, instead preferring to follow BlackRock's push for tokenizing all assets and upgrading the rails of the global settlement layer to crypto. Some of that will of course trickle down into our broader ecosystem, but much of it will be partitioned into products and protocols that they build for their needs. Wall Street has shown it can repeatedly be idiots about things, but they generally prefer to do that with counterparties that adhere to their norms of conduct and culture. Do Kwon, SBF, 3AC, Barry and Celsius likely taught them a lesson that their next time onchain really needs to be run from inside the house as much as possible.

Culturally, retail has never been as disconnected from real capital as it is right now, and I do not expect that to change for the next go around. Simply put, foot soldiers are not the cannon fodder the Street calls for. Once again, it's capital.

I mean I would love to call the interregnum off and return to the wild style of the last bull, but you'll have to settle for the close of this article.

The secret, more complex third thing sitting between the bull and the bear is boredom. It's asking yourself whether you want to climb the walls, toss yourself into the well of memecoin dopamine or simply be like Bartleby the Scrivener and prefer not to participate until you see the conditions open up again for retail to be rich.

This isn't an all in choice, you can do all three and lord knows I flip flop between all of these postures. There's no telling when conditions are going to change so the little guy can just get out there and haul in fish after fish. Instead, you just need to keep casting the rod, keep an eye on the sea and not get too greedy while the fish aren't biting. The thing to do in interregnums is to educate yourself, use the downtime to keep an open mind to possibilities and be aware of the environment you find yourself operating in.

What is clear is that the tech looks extremely promising, protocols and practices are maturing, and market caps are rebounding off their lows. The key thing is to understand where we are. The world is constantly adapting. Hopefully this article has opened your mind to ways in which the next opportunity won't look exactly like the last one, and you can start looking for ways to play tomorrow.

I actually once nearly quit telecom because there wasn't any new opportunities opening up for us little guys. I was climbing the walls, and ready to get back to consumer internet because we were in stasis. A couple short years later, a wedge appeared that allowed us to level up and we never looked back. It really sometimes is as simple as waiting out an interregnum and then striking once conditions fit your strengths.

The global population is 8 billion.

67% of it is online. There are roughly 400 million native English speakers in the world, but English now has 1.5 billion speakers. 18.8% of the world speaks English, which means 13.8% of them speak it as a second language.

No other language has a similar usage profile. Mandarin has 1.1 billion speakers, but a tiny population who speak it as a second language. Hindi has 600 million speakers, Spanish 550 million. Rounding out the top six is French and Arabic, both with 275 million. To put this in context, there are approximately 7000 living languages in the world and 50% of the global population speaks just 6 of them.

What is even more interesting is that the growth rate of English has accelerated dramatically since 2017.

When I first started looking into this, I assumed that the rise of English would track the growth of global internet usage. 55-60% of all content online is in English, a number that's been steady for years. The chart below tells a different story. A steady 14-15% of the world population spoke English between 1995 and 2016, it's only from 2017 onwards that we see a rapid growth in English language adoption. English and the internet have not moved in lock step with each other. Something else is at work here.

The reason for this growth is not entirely clear. The first problem we encounter is the paucity of reporting and statistics on this trend. The data in this article is cobbled together from several sources. All statistics I've seen on language growth come from a paywalled reference publication, Ethnologue and I got the data by watching this YouTube video. The numbers match up to stats I've seen cited by the CIA Fact Book and Encyclopedia Britannica, but methodology and proficiency levels are lacking.

The second problem is there's little to no widely available analysis on these trends beyond financial reporting on the growth of the English Language Training market. Despite that, all signs are pointing to English as increasing its market share and cementing its status as the global language of business, culture and the web.

Had this surge in English happened in 2020, it would have been all to easy to point to Covid as the reason. 2017 on the other hand shows no clear catalysts, but the trend has been going strong for six years now indicating it was not an anomaly.

Language apps are certainly helping, but let's take a look at the best known one in the world, Duolingo who reported quarterly DAUs on this blog post:

These numbers are impressive, but Duolingo teaches many languages and is quite popular in the US which boasts the largest native English speaking population.

Reports say that most growth in English learning is based in APAC with the two largest populations in the world (India and China) making up the bulk of it. This isn't however a case of two markets carrying an entire global trend. Statistics here are dated, but India is estimated to have 125 million English speakers and China 10 million. It's possible that growth is coming primarily from these locations and there is a reporting gap, but it's far more likely that English use is growing everywhere as a result of globalization.

In short, there are no good answers (at least ones that I can provide) for the rise of English. We have to look at macro trends of all levels and point to a variety of effects. First, there's an increasing concentration around the large languages as a result of many compounding effects from the legacy of colonization to urbanization to economic opportunity favoring scale and conformity. We are losing diversity in language across the board. Second, we need to look at globalization and the effects of corporations adopting standard languages which are required for employment. Third the internet has to have an effect here, especially in creating access to learning. Duolingo is just one of many language learning tools out there.

Still this isn't a particularly satisfying answer. If anyone out there can point to specific triggers for the recent surge in English, please share. This feels like too big a trend to leave in the hands of fuzzy reasons and unclear causations.

In under 30 years, we've put 5.4 billion people online. That's double the population of the entire world in 1954, a mere 70 years ago. At least one billion of them (but likely more) speak English, and the rate of English adoption has started accelerating in the last six years.

These numbers are staggering on so many levels and yet we never stop to question what the second order effects are of this transition. We assume that things either behave the same as always (government, institutions, values) or we view the internet at a surface depth, never pondering how system dynamics are changing as languages start reaching billion plus speaker levels of penetration.

This leads me to the title of this essay, Is Language Having a Moment?

It's now time to explore whether the scale and speed of this dominant, English speaking, networked communication layer is having a bigger impact than is readily apparent. We'll begin by looking backwards:

A broad arc of economic alpha over the last century looks like this:

Resources -> Manufacturing -> Management -> Finance/Technology -> Information

By this I mean that the greatest opportunity for growth has shifted from one area of focus to another as each of the preceding spaces develop and mature. Back in 1920 what mattered was pulling things out of the ground. At a certain point transforming those resources into finished goods took prominence. Efficiency in both extraction and production then took center stage as management streamlined operations until it was gated by pencil, paper and punch cards. That led to a push in the technology and finance arenas, creating global networks connecting us and making it easier to move capital worldwide. We literally opened the world to same day interoperability on an unimagined scale. With that accomplished, data became the priority as we filled databases with numbers and then made them talk to each other. The trends and relationships within that data, then informed businesses to a degree they previously lacked.

What we see in this evolution is a pattern of increasing complexity that requires higher and higher degrees of abstraction. We package components into concepts, rolling things up into ever larger assemblages which we can wield and manipulate.

Ask yourself would you rather own a coal mine or a SaaS provider? The answer completely depends on whether you lived in 1920 or 2020.

Now ask yourself would you rather own a SaaS provider or a Large Language Model? Four years ago, one was a cashflow generating machine and the other was an unproven high cost research project. Now AI is poised to render specialty software obsolete by its switchblade capabilities.

Information remains as important as ever, but is it the alpha that it once was or has complexity increased to the point that a higher level of abstraction is ready to take center stage?

Here is where we make a leap into the truly speculative and suggest that language is in the process of supplanting information as the new source of alpha. None of this is binary. Information remains important. So do resources, manufacturing, tech and finance, however all of them are maturing systems working in concert with each other in a feedback loop known as the global economy. The market has absorbed all the low hanging fruit it can get from the last unlock and needs to add a new component on the edge to push itself forward even further.

Everyone seems to think that is AI, but what if I argued that it is actually language or said more precisely language forms the top and bottom with AI sandwiched in between? Prior to LLMs, we had machine learning. ML was about forming predictions from past data and steering systems into beneficial behaviors. It was a triumph of information, but not of language. As such, the use cases for ML were limited to the corporate and institutional sectors. LLMs are far more democratizing because are able to form associations between language and act as text prediction engines on steroids whose outputs go beyond text and into media and software. Anyone from a kid wanting to make a cartoon of themselves to a Fortune 500 company can benefit from LLMs.

Not only that, but the interface for LLMs are dead simple. They are text, asking for language as the starting point in a conversation which ultimate ends in the transformation of huge amounts of information into an end product created on demand for a user who simply asked for it.

Accessibility and language go hand in hand. You cannot have one without the other once you ascend up Mazlow's hierarchy of needs past the very basics. For as powerful and transformative as AI will be, the ubiquity and versatility of language may actually make it the real star of our next step forward.

We are going to start this section with the structuralist argument that language defines what we are capable of thinking about. If there are not words to describe a thing or an experience, then it is impossible for one human to communicate the concept to another. A thing must be named and that name must be disseminated via broad communication between people, or else it is not useful because we can't place it in our mind's eye.

Empiricism and the start of the scientific revolution is an example where the formation of language was just as important as the underlying work being performed. The table of elements, taxonomy, anatomy, physics, mathematics all required building upon existing language in order for their respective disciplines to advance. Without new words or expanded definitions of existing words, there was only superstition or the will of God to explain our natural world.

Let's get more modern. Router, WIFI, USB cable, 5G, virtual reality, mobile app, cryptocurrency, play2earn, inference model, eigenvectors...our lives are dominated by techno jargon. We must constantly update our language in order to understand and interact with the proliferation of tools that we are creating at an ever more accelerated pace.

The advantage of language is that it is plastic, used everyday and essential. By plastic, I mean language is self-organizing and bottom up. There is no committee that decides whether or not a thing is a word. It is decided by social contract in ever-broadening circles of communication between peoples. If a new word is useful, it is adopted. Once it reaches a critical mass of importance and distribution, a committee does sit down and include it in a dictionary, but that is a post-facto action not an initiating one.

We live in an age of enormity. There's three times more of us than there was when my father was born. Not only that, but we live beyond the physical now and extend into digital worlds and networks which are conceptually boundless. There's simply more things to talk about and give our attention to which means language matters more than ever because of competition for our mindshare. We only think about things we can understand and communicate to others.

The argument here is that language sits on top of all other advancements. It is the necessary precursor for the establishment of new ideas and technology. Unlike other fields, language has the advantage of speed and malleability. It is universal and in constant use, and has continued to show it is more than capable of keeping up with what comes next.

Over half of my son's TV watching is not TV. If it wasn't for our parental intervention, I suspect all of his consumption would be watching and listening to other people playing video games on YouTube or Twitch.

Streamers are a relatively new global phenomenon. It's bottom up, radically inclusive (in terms of access, not culture) and perhaps the ultimate abstraction to also double as an industry. We are talking about watching someone else navigate a virtual world. The story here is not the story of the game, but rather someone else's attempt to beat it. In under twenty-five years, streamers and influencers have gone from non-existent to a completely normalized and heavily aspirational activity/industry for younger people.

It's also language all the way down. Even the games themselves are nothing but a collection of language made to behave in arbitrary ways. Perhaps we don't consider streaming all that strange however because it still sits squarely in the realm of ad-supported entertainment.

Lets switch gears and talk about memes, cryptocurrency and their intersection, memecoins. Crypto has a language problem that is worse than many fields of frontier tech. It is dominated by STEM-trained males who believe that code is law, that mechanism design matters more than usability, and favor economic incentives (giving away tokens) as their primary and sometimes only means of marketing.

Some of this mindset is justified by the nascent state of the industry. Many of the protocols developed in the space are bleeding edge concepts which have never been attempted before and whose utility is unproven. There's also an incentive loop which justifies skipping out on marketing and communications. Crypto can create value unlike any other field of finance, and early adopters are heavily rewarded for simply being early and participating. While the barrier to entry is high, the payoff is worth the work. One might not understand just what they are getting into, but more often than not jumping in with only a partial understanding is EV+.

Memes are a hack of language. Combining an image with a short burst of text creates an easily communicated concept capable of moving at digital speeds. It's possible that we are able to scale the internet as fast as we have simply because we've compressed many ideas that used to take a paragraph of text to communicate down into a viral semiotic shorthand. Memes are abstraction on steroids.

Memecoins then are combination of crypto and memes in a doubly compressed format. All the complexity of crypto is reduced down to owning a permissionless tradeable asset whose value either goes up or down. A holder simply decides how much to buy and when to sell it. On the flipside, memecoins are a way to make culture and language interoperable with markets in the most conceptually simple way. If memes are a compression of language into snackable format, memecoins are a snackable, social way to participate in global finance.

There's been a lot of thotting on the rise of memecoins. I myself have called them tokens of attention and a way for people to project shared beliefs through pooling capital. Others have called them a symbol of financial nihilism, an unregulated way to gamble, and modern day lottery tickets. All of these things have elements of truth, but perhaps there's another element at work here, and that's the abstractive power of language to steer capital becoming democratized at a fully networked scale.

Would the memecoin market be worth $53 billion dollars if there weren't 5.5 billion people online? Would internet culture be able to accrue value at this scale if half the world didn't speak just 6 languages? Here's one very strange glimpse into why English adoption is accelerating. Memecoins are dead easy to understand, their concepts are overwhelmingly communicated in simple English, and they are the highest performing category in crypto this year.

Toby Shorin wrote an important piece called Life After Lifestyle which traces the evolution of brands as drop shipping, white labeling and ecommerce social distribution came to prominence in the last decade. In it he argued that consumer goods became plastic enough (meaning you can put any label on nearly any product) that the underlying thing lost its meaning. In its place, the brand association gained importance to the point that brands freed themselves from product itself, and were able to perpetuate themselves based on meaning and values alone. In some ways, memecoins are the manifestation of that vision. There is no product, there is only the value associated with the values of an idea. Said another way, there is only language.

If we are trending in this direction, then generative AI is poised to take it even further. I spent the first couple months of this year enamored with AI music making five albums worth of material under the Starholder name. All I needed was the ability to write lyrics, a deep working knowledge of the language of music and Suno.ai to talk to. That effort resulted in a three way conversation between myself, the platform and the dozens of people who listened to what I produced. In all that a single instrument was never picked up, abstracting music up to the level of language.

There's a simplistic ubiquity circling around generative AI. All models use text as their user interface. It doesn't matter if you are making images, articles, music or video, each modality just asks you to type into it. There's also a lack of precision at work in these models which favor large descriptive vocabularies. The tech is young enough that each has its own fussy levels of refinement and capabilities. They often converge into the mid, producing generic outputs and getting them to break from that to create interesting work requires lulling AI models away from common terrain via conversation the same way you need to get your football fanatic co-worker talking about anything else.

In some ways, making the case for language is like arguing about the importance of water. It is essential, it is everywhere, but it's also so basic that it gets taken for granted and might not be all that actionable. On the flipside, something larger is at work here and it might be too early to pin down exactly what that is.

What is clear is this:

• Half the world speaks just 6 languages with English scaling to 1.5 billion speakers

• We are seeing people learn English at an accelerated rate, with a big pickup since 2017

• The frontier of technology is operating in a conversational modality. We are past the era of code, math and numbers at the user interface level.

• While hard science and technology development will remain in the realm of engineering and mathematics in the short term, access to those advances is being radically democratized via chat interfaces into artificial intelligence.

• The world continues to get flatter, more connected and operates at an accelerated pace. Language is able to adapt to these changes by concentrating speakers into fewer languages, communicating at higher frequencies making it easier to create and disseminate updates to its corpus in a way that pushes other fields forward.

Because language is so big and so basic, I think it's elevation in importance is being overlooked. There's some blindness at play here as well, we've always mentally separated the world into industrialized and developing regions. The value add is not coming yet from turning Madagascar into a chip manufacturing powerhouse, it's coming from integrating a broad distributed base of very online people into an emerging digital economy which is traditionally underreported. No one writes articles about all the labor gig workers contribute, they write about the creator who outsourced and assembled that effort.

This will change in time. It will change when the economic activity of more weird productless digital ideas' reach tipping points where capital stands up and takes notice. Today, venture capitalists and hedge funds are dipping their toes into memecoins because the returns are there. We are not far off from that extending into media. Influencers and streamers have already shown they are capable of using the power of language to assemble audiences which rival broadcast networks, they just are doing it under the umbrella of tech platforms like YouTube, Instragram, TikTok and Twitch. There will come a time when independently produced entertainment will be distributed over permissionless networks at a scale that forces Hollywood to stand up and take notice.

Our larger story already has a scripted ending. It's the arc promised by globalization, where economic development (AKA labor arbitrage) was an investment in the future of developing nations. If we reach a point where that optimism plays out and there are flat screen TVs and Hondas in every house in the world, then analysis will point to free trade agreements and western liberal democracy as the catalyst for transformation. It won't focus on the actions of each person who adapted their lives. It won't point to our horrific treatment of migrants, discouraging them from moving to another country. It will not point to language, nor to the self-organizing effects of the internet.

Under the surface, there will be another story. It will be one of scale, connection and language. More people communicating in common tongues on a shared public good, using more powerful tools to create new categories, scenes and even industries. This is where the opportunity is. It will be strange, seem niche, and move at uncomfortably fast speeds. Language will lead it, because we cannot have new things without words for them. Soon enough it will be words that produce objects, code and systems.

It remains to be seen how impactful AI is or what its end interfaces will look like, but the more it mirrors how we communicate today the more we will integrate it into our lives. If current trends hold, we could see a quarter of the world speaking the same language. Combine those two things together and the potential for growth is explosive.

Maybe in hindsight, the case for language will have been obvious. Only time will tell.

Whether you've ever gotten the Led out or not, this question sounds very familiar. You've probably heard it recently, just asked a different way by very online Gen X-ers and Millennials.

Do you remember when the internet was fun?

There was a time when the internet was in fact fun. I know it's hard to believe. but it is true. I'm as old as Robert Plant's laughter quip and I can remember when the internet was a place of discovery and serendipity.

Most people use Tumblr as their touchstone for a kinder, weirder, funner internet, but honestly that's a Zoomer structuralist response. It's an answer given by people who only know of doomscrolling, of content being sandwiched into blocks and served up in factory-templated feeds. Yes Tumblr was weird and it was beloved, but it carried all the hallmarks of what has ruined the internet today. Aggregation, structural conformity, reactions and comments...in short, herding.

It's nearly impossible in this day and age to think of a website whose job it is to send you somewhere else (even search engines keep you contained as much as possible), but before Tumblr there was another beloved website / browser extension which did just that, StumbleUpon.

StumbleUpon operated on a premise of discovery. The internet was wide open, unmapped and full of surprise, only we often had no idea where to go or what to do. You'd use StumbleUpon by telling the site what subjects you are interested in, and then it'd launch you out into random (but topically related) websites. Here's a quick, super old YouTube about it:

A big part of the appeal of StumbleUpon was that we as users weren't entirely sure what the internet was capable of. Yeah, we had Yahoo and Google, but they didn't own mindshare like FAANG does today. We were more open to the possibility that the internet was infinite and it was worth seeking out all the nooks and crannies it contained, just to see what was there.

Honestly until generative AI and LLMs came along, I had forgotten what it was like to use networked tech just to see what was there, to see what we as a hivemind are capable of creating before being yoked down and herded by entrenched interests.

I know I titled this essay 20 Minutes into the Future, but we are going to do one last bit of time travel, and go way way back to ancient Greece to talk about the tension between the genos and the demos, along with the best solution they had for it in 750 BC. It's a good setup for when we do flash forward.

The genos is what you inherit when you are born into the world. It's literally a family bloodline and a term for the noble houses of classical Greece, but metaphorically it was more than that. The genos defined culture, tradition, and laws. It was the weight of history that guided how people moved through the world.

The demos were defined by their otherness to the genos. They were the common people, the ones who had to live with the weight of being outside the genos. When the population of Greece picked up after the Bronze Age Collapse, good land and resources became scarce. Tension between traditional rights and immediate needs flared up.

Did the genos give up their ancestral holdings to make room for newcomers? No. The answer was to go elsewhere and open up undiscovered country.

Flash forward to the present. Apple is currently defending, stalling and even disregarding attempts to open up its app store and payment monopoly. The demos is demanding better rights and treatment, and the genos is doing all it can to keep that from happening. Same as it ever was...or better said, same same but different.

When Cornith got too full, its demos and disinherited genos hauled their butts off to Sicily and started Syracuse. That same process is happening today. Crypto can be viewed as one early and pivotal exodus from mainstream net culture, and the next one is emerging with generative AI and LLMs.

Which brings us to Websim where we can finally step 20 minutes into the future.

Websim is named for what it does. It simulates the internet, treating the address bar in its virtual browser as a chat interface where users can prompt the platform to render virtual websites dynamically.

The project isn't even a month old, and was not an intentional creation. Its founders, Sean Lee and Rob Haisfield, met up at a hackathon where each was interested in another AI project engaged in simulation. Serendipity took hold and in a weekend, they built a tool that simulated internet pages.

Using Websim is quite easy. Lets walk through it. For starters, you need to login with a Google account and then you'll see a virtual browser within your browser:

You then click into the address bar of your virtual browser and use it to navigate a simulated internet which is created by an LLM (currently Anthropic's Claude) under the hood.

The address bar takes either a URL that you make up or you can just use it like a chat window and type text into there which the AI uses to create virtual web pages with. I find both work equally well, but enjoy constructing fake URLs more just to see how it interprets the query strings and params to create a visual output from it.

To be very meta in this explainer, I took all the text from this essay so far and just pasted it into the address bar:

And Websim then produced a page that looks a lot like the one you are reading this essay on now:

Which is simple enough, but what if I told you I also made this very glitchy cube using it?

Wild huh? Once I discovered this unlock, it was deep down the rabbit hole to see just what it was capable of rendering.

And now the possibility of Websim starts to come into focus. This isn't just a modern day website builder, some piece of SaaS with an AI strapped onto it. No, this is the ability to summon and conjure up vector space. It is the ability to travel into the depths of a large language model and transform the information within it into net new experiences rendered in networked address space.

We have not seen anything like this before and it is very exciting. The community in their Discord is doing astounding work, creating games, virtual spaces, art, simulating models and really only being limited by imagination. It's truly inspiring to see how people are exploring this new functionality.

In my last essay I introduced the concept of Complex Adaptive Media Systems (CAMS), which is an organizing model for how to think about future media once we have the ability to scale our creations past objects and websites and into worlds and simulations. Thanks to generative AI lowering production barriers and massive scaling in crypto's throughput and capabilities, we are on the cusp of a revolution where media will become programmatic, heavily networked and operate on orders of magnitude larger than it does today. Think of how your parent's landline telephone was used in 1990, now think of how you use your iPhone today. That sea change in scale, scope and use cases is coming to media over the next decade.

A critical aspect of that change will be in how we interact with media. Rather than passive, scripted media we snack on from our couch, these virtual worlds and simulations will be hands on and exploratory. We'll need tools and new modalities for summoning what is inside them in ways that work for us. In short, we'll need AI to render user interfaces on the fly based on input from us. We will tell the AI where and what we want to investigate, and it will dynamically create liminal spaces that transform all the objects, information and relationships within a virtual world into an interaction layer for us to explore.

This is the new world that awaits us, and Websim is the first time I've interacted with a project that approximates what is coming. It is a proto-CAMS, a glimpse 20 minutes into the future.

The internet is fun again thanks to Websim, give it a try and see for yourself.

I had a lot of fun playing with Websim over the last week and wanted to document what it was capable of in its first month of existence. My ten favorite outputs are assembled into a NFT collection on Zora as editions of 12 (120 piece set total). Each piece is a video capture of a Websim simulation, and metadata includes a link to render the piece in browser on Websim.

Wallets can mint up to 10 pieces in case anyone wants to collect the full set. Enjoy!

Editor's Note: I am an idiot and messed up the release date. The drop will go live Monday, April 29th at 3:45 ET.