The nation-state spent two centuries convincing people that political belonging and physical location were the same thing. What breaks that assumption first?

Not a single break, more like a slow delamination that most people don't notice until it's already happened.

The first layer to go is the emotional one. When your closest collaborators are distributed across six time zones, when your most meaningful intellectual community exists entirely in a Substack comment section or a private Discord, when the people who most reliably show up for you in moments of crisis are people you've never been in the same room with, the geography of your loyalty quietly reorganizes itself. The address on your passport starts to feel like an accident of birth rather than a description of who you are.

The economic layer goes next, and that's where it gets structural. Remote work started as a pandemic accommodation and became a permanent renegotiation of the relationship between employer and place. Once that happens at scale, once you have a substantial population whose income is genuinely portable, the implicit deal underlying local taxation and local political participation starts to feel much more optional than it used to. You start doing the math. What exactly am I getting from this particular jurisdiction, and is there a better arrangement available to me?

The legal layer is last and slowest, but it's moving. Arbitration already handles a significant portion of commercial disputes outside the court system. Cryptographic contracts are already enforcing certain kinds of agreements without reference to any state. Once those systems mature to the point where they're more reliable and less expensive than the domestic alternatives, people stop thinking of state law as the default and start thinking of it as one option among several.

None of this requires a revolution. It just requires enough people, independently, to stop assuming.

Every serious attempt at intentional community has eventually fractured along the same lines: money, sex, and who does the dishes. What makes you think digitally-organized communities will be any different?

They probably won't be, and I think that's actually the more interesting observation.

What changes isn't human nature. What changes is the cost structure of exit and reformation. Traditional intentional communities failed catastrophically and permanently because the physical infrastructure was so entangled with the social fabric. When the community split, the farm split with it, or didn't split, and that became the central conflict. The material stakes of staying or leaving were enormous and not easily adjusted.

When the primary substrate is digital and the physical infrastructure is deliberately kept modular and portable, the failure mode changes character. Communities can fork. Subgroups can splinter off, run their own experiment, and either thrive or rejoin or fade away, without anyone having to fight over the physical plant. You get a much higher velocity of social experimentation, and with that, a much faster accumulation of knowledge about what actually works under what conditions.

The communities we're building are less like communes and more like open-source software projects. They will fork constantly. Most forks will go nowhere. Some will produce genuine innovations that get pulled back into the main branch. The goal isn't to build something that never fractures. The goal is to make fracture generative rather than terminal.

Corporations already provide housing, healthcare, and social belonging for their employees in some parts of the world. At what point does that stop being a benefit package and start being something else entirely?

When you can't easily leave, and the leaving would cost you the whole stack at once.

The progression is fairly legible if you watch it without the vocabulary that makes each individual step seem reasonable. First the company provides good health insurance, which is genuinely valuable in a system where insurance is expensive and fragmented. Then it provides housing near the campus, which is genuinely useful in a city where rents are high. Then it provides childcare and schooling, which is genuinely necessary when both partners are working demanding jobs. Then it provides a social community organized around the workplace, which fills a genuine vacuum left by the decline of civic institutions.

At each step, the individual transaction is rational. The aggregate is a comprehensive life structure in which your employer is your landlord, your healthcare provider, your children's educator, and your primary social world. The benefits package has become a jurisdiction.

What makes it something else entirely is the asymmetry of exit costs. When leaving a job means simultaneously losing your housing, your children's school continuity, your healthcare, and your primary community, you are no longer an employee in any traditional sense. You're a resident. The question is whether you have any of the protections that residency is supposed to carry, and the answer, under current legal frameworks, is mostly no.

Religious institutions maintained coherent communities across vast distances and centuries before the internet existed. What do they understand about this that technologists keep relearning?

That the content of the shared commitment matters enormously, and that it has to demand something real from the people who hold it.

Technologists consistently underestimate the weight-bearing requirements of shared identity. They design for frictionless onboarding and assume that network effects will generate cohesion. What they end up with is audiences, not communities. Large, engaged, sometimes passionate audiences, but groups that dissolve or redirect the moment the platform changes its algorithm or the founding charisma fades.

Religious communities survived translation across centuries and continents because the commitment they asked for was total enough to actually restructure a person's daily life. The practices were embodied. The obligations created genuine mutual dependency. The shared narrative was long enough and strange enough that you couldn't hold it alone, you needed other people who also knew it to help you carry it and interpret it across the changing circumstances of a life.

What technologists keep rediscovering, usually the hard way, is that community requires genuine cost. Not friction in the UX sense, but actual sacrifice of optionality. The willingness to be bound. Most platform designers treat that as a bug to be engineered away. The communities that last understand it as the core feature. The people worth building with are the ones willing to accept real constraints on their own freedom in exchange for genuine belonging.

Physical borders are becoming simultaneously more surveilled and more porous. What kind of person thrives in that contradiction?

Someone who has learned to maintain multiple simultaneous identities without experiencing that as fragmentation.

The surveillance layer and the porosity layer are actually targeting different populations, and most people are only fully subject to one of them. The surveillance apparatus, the biometrics, the financial monitoring, the digital exhaust tracking, is extremely effective at controlling people whose lives are legible to state systems. People with conventional employment, conventional banking, conventional travel patterns. For them, the border is becoming a very thorough audit.

The porosity benefits people who have learned to operate across multiple jurisdictions without being fully legible to any of them. Not necessarily illegally, but in ways that don't map cleanly onto the categories that border systems are designed to process. People with genuine professional networks that span cultures, with financial lives that aren't fully captured by any single banking system, with the social fluency to move between very different community contexts without standing out as an outsider in any of them.

What this selects for, over time, is a particular kind of person who experiences national identity as one context among many rather than as the master category of their existence. They are not rootless in the way that word usually implies some kind of loss. They have roots in multiple places simultaneously. The contradiction doesn't threaten them because they were never fully subject to the old terms of the agreement in the first place.

Every previous communication technology, writing, printing, broadcasting, created new institutions to control and legitimate it. Why do you think this one will be different?

It might not be, and sitting with that possibility is more productive than assuming otherwise.

The pattern you're describing is real and remarkably consistent. Each new communication substrate eventually produces a priesthood, a class of people whose primary social function is managing access to the technology and interpreting its outputs for everyone else. Writing produced scribes and then clergy. Printing produced publishers and eventually the professional journalism apparatus. Broadcasting produced networks and regulatory bodies. Each of these looked, from the inside of the transition, like radical democratization, and each of them eventually produced new hierarchies that were different in form but familiar in function.

The case for this time being different isn't about the technology itself. It's about the speed of the iteration cycle. The institutions that formed around previous communication technologies had decades to consolidate before the next disruption arrived. What we're living through now is a compression of that cycle to the point where no single institutional form has time to fully calcify before it's already being disintermediated. The scribes of one platform become the thetes of the next one within a few years.

Whether that produces genuine pluralism or just faster rotation of the same concentrating dynamics is genuinely unclear. What I'm watching for is whether any of the new coordination tools allow people to build institutions that are constitutionally resistant to the capture that took down their predecessors. Not because the people building them are more virtuous, but because they've encoded exit rights and forking rights directly into the architecture before anyone has enough power to remove them.

Image credit: custom prompt using VGR's bucket art model on titles.xyz.

For a good chunk of my post-college years, I developed a strange habit of reading the introductory sections of math and science textbooks, the parts everyone skips to get to the actual content. I was hunting for how each author chose to define their terms. What is math, really? It's the only major branch of human inquiry that eludes a precise consensus definition and is both an art and a science. What does it mean to apply math as a tool? Every author had a slightly different answer, and I found the variance more interesting than any single problem set.

Taking that spirit to the areas of infrastructure and public works where I've spent my career, I posted this little substack note a few weeks back that went moderately viral (by my low standards at least).

Today the word has drifted from the definition. In public works contexts, "engineer" signals a license, a stamp, a professional designation you earn through an accredited degree and a state exam. That is a legitimate and useful category for certain purposes. Nobody wants an unlicensed person signing off on a bridge's structural calculations. But it is a narrower thing than the original word, and the narrowing has consequences.

Part of what happened is that public works slowed down. When you are building a lot, the feedback loop between idea and structure stays tight. Someone proposes something, it gets built, it either stands or falls, and everyone learns. When you build rarely, the loop stretches out over years or decades, and into that gap creeps ritual. Process for its own sake. Form standing in for function. The permitting review that exists because it always has, not because anyone can trace it to an outcome it actually prevents.

The other thing that happened, somewhat independently, is specialization. Modern engineering organizations, public and private, increasingly ask each person to own a subsystem of a subsystem. You design the valve, not the pump. You review the pump spec, not the system it feeds. That has real advantages. Depth matters. But it also means fewer and fewer people build anything end to end anymore, from concept through fabrication through installation through the inevitable field fix. Tacit knowledge, the kind that only accumulates in your hands and your gut from actually doing the thing, gets lost at every handoff. Each team knows its slice cold and knows almost nothing about how the slices fit together, and there is no one left whose job is to hold the whole picture.

I say this as someone who is not, in any official sense, an engineer. My background is mostly in engineering digital solutions. I have shipped production data pipelines for Fortune 50 companies, which is a fancy way of saying I worked a few months basically as an intern at an "it" startup back in grad school. I have also carried some of that same instinct, build the thing, watch what breaks, fix it, into the physical world here and there. I do not have a PE stamp and I am not claiming one. What I am claiming is the older sense of the word: someone willing to apply math and reasoning to a real problem and see it through.

Every so often someone reminds the world what the older sense of engineering actually looks like, and the reminder tends to arrive from an unlicensed direction.

In 2013 Mats Järlström's wife got a $150 ticket from a red light camera in Beaverton, Oregon. Järlström, an electronics engineer by training with a degree from Sweden, sat down and worked through the kinematic formula traffic engineers have used since 1959 to time yellow lights. He found a gap. The formula assumed drivers either sail straight through an intersection or come to a full stop. It did not account for drivers slowing down to make a legal right turn, who decelerate mid-intersection and can get caught by a light that changes to red while they are still turning. He reworked the math to account for that deceleration and sent his findings to the Beaverton city council, local media, and the Oregon State Board of Examiners for Engineering and Land Surveying.

The board's response was not to engage with the math. It investigated Järlström for two years and then fined him $500, for two things. He had used the word "engineer" to describe himself in correspondence, and under Oregon law at the time that word was reserved for people holding an active state PE license regardless of whether the description was accurate. And the board found that by analyzing public traffic light data and critiquing the formula used by licensed engineers, he had engaged in the unlicensed practice of engineering. Board meeting minutes show a member stating plainly that by refuting the professional opinion behind the original formula, Järlström was practicing engineering.

He teamed up with the Institute for Justice and sued. In December 2018 a federal magistrate judge ruled decisively in his favor, finding that critiquing public infrastructure and doing math about it in public is protected speech, and that the word "engineer" has enough ordinary meaning in English that a state cannot bar a citizen from using it truthfully. The board was permanently enjoined from enforcing either restriction.

The postscript is the part I actually love. In 2020 the Institute of Transportation Engineers, the international body that sets the standard, formally adopted Järlström's corrected formula, replacing math that had gone unchallenged for fifty-five years. An unlicensed hobbyist working at his kitchen table found something the credentialed profession had missed, and once he was allowed to say so in public, the profession updated its standard. That is engineering working exactly as the word originally intended, applied math meeting a real problem and making something better. The two-year detour through licensing enforcement was the system trying to decide whether that kind of contribution was even allowed to happen.

So here is a thought experiment, silly on its face but I think less silly than it sounds.

Every twenty years, Ise Jingu in Japan is completely dismantled and rebuilt from scratch. Not restored. Rebuilt, on an adjacent plot, using the same specifications, by a new set of hands. Salvaged timber gets redistributed to shrines across the country, and the cycle keeps traditional craftsmanship alive by forcing it to be exercised rather than merely preserved. More than two thousand master artisans pass their specialized skills down through the actual doing of the work, not through a manual. The tradition has run for something like thirteen hundred years. Nobody documents Ise Jingu into obsolescence. They just build it again.

Where is the Ise Jingu equivalent for public works?

I do not mean a museum piece or a reenactment. I mean a real, regular, load-bearing cadence where a group of people build an actual piece of infrastructure from scratch, install it, watch it fail in some small way, hot fix it, and take apart what they learned before they take apart the thing itself. An aqueduct model, a pump station, something with real water or real current running through it and real consequences if it is built wrong. Something small enough to complete in a week, not a generation, because twenty years is too slow a heartbeat for a discipline that needs to keep its hands calloused. And crucially, something end to end. One team, one build, from concept to commissioning, so the tacit knowledge that gets lost at every handoff in a specialized bureaucracy has nowhere to hide.

AWWA already has a version of this instinct on the operations side, the annual meter madness competitions where crews race to install service lines against the clock. That is real and it is good. It keeps a physical skill sharp through actual repetition rather than certification renewal. But it lives on the ops side of the house. Nobody is racing to design and stand up a new small system from first principles, then argue about what to change next time.

I don't know what the right tempo is. Faster than twenty years, certainly. Maybe annual. Maybe seasonal, if you wanted something closer to a harvest rhythm than an anniversary. I don't know what the right scale is either, though something small enough for a dedicated crew to complete in a week feels roughly right as a starting guess. What I do know is that the feedback loop is the whole point. Ise Jingu is not maintained by admiring the plans. It is maintained by tearing the thing down and building it again, in public, with your own hands, on a schedule short enough that the knowledge never has time to calcify into a form nobody remembers how to execute.

Further Reading

The March of Folly reflections on institutional feedback loops For more on what happens when the map decouples from the territory and no one is positioned to notice. (This is a review I wrote on Barbara Tuckman's wonderful book about folly from fall of Troy to Vietnam).

AWWA Meter Madness competitions Worth a look as an existing proof of concept for what a physical, repeated, hands-on public works ritual can look like, even if it currently lives entirely on the operations side rather than the design and build side.

Järlström v. OSBEELS case materials, Institute for Justice The full case record, useful if you want the primary source on how far a licensing board was willing to go to keep a citizen from doing math in public.

Agile (Physical!) Infrastructure Development

You've heard about agile software development, right? The web 2.0 manifesto usurped the upper echelons of high technology and overthrew the waterfall crowned kings of the 1990's internet. I'd like to share an example of what we can call agile infrastructure development. Here I don't mean servers or data warehouses or any backend bits or bytes.

https://pioneeringspirit.xyz

Having never taken an engineering course, I have been self studying a bit of some of the engineering tests and courses relevant to utilities. One piece that's fascinating to me is how much approximations from the 19th century like the linear method or Manning equation are still largely used. The development and patterns reminds me of statistics and this magisterial book.

The History of Statistics: The Measurement of Uncertainty before 1900

Amazon.com: The History of Statistics: The Measurement of Uncertainty before 1900: 9780674403413: Stigler: Books

https://www.amazon.com

Writing coda

I spent some time digging into the history of engineering and poking around with ai on interesting examples to unearth the Oregon case study. I used that and some fairly lengthy instructions to have this piece written using a bespoke Claude skill. I debated the punchy title a bit internally but then thought hey why not own it.

It goes about how you would expect. A brilliant person, genuinely brilliant, who has built systems of real consequence, arrives at the observation that municipal permitting is slow, or that water data is hard to get during a drought, or that the same address is spelled four different ways across three different databases. All true. All things I have spent a decade of my professional life on. And then comes the proposal, delivered with total confidence, that is either something we tried in 2013 and watched die for reasons that had nothing to do with the technology, or something that would require rewriting a state constitution the speaker has not read.

I do not say this with contempt. The energy is real and I would rather have it than not. But there is a specific failure mode where technical brilliance gets mistaken for domain knowledge, and the tell is always the same: the proposal treats the accumulated institutional sediment as noise rather than as data. Somebody built that sediment. Usually for a reason. Sometimes a bad reason, but a reason you have to understand before you can dissolve it. Chesterton's fence and all that really isn't a book learning lesson so much as a parable that points to experience.

So when Eli Dourado went after Silicon Valley's unworldliness about international institutions this week, I recognized the shape of the argument immediately. Ten thousand percent yes.

Dourado's specific complaint is worth taking seriously because he has credentials most of the people he is arguing with do not: he has advised the State Department, sat through multiple treaty negotiations, and served on technical committees for UN agencies. His account of that world is not flattering. Treaty processes, in his telling, are less about the stated object of the agreement than about leverage, with participants routinely using multilateral negotiations to hobble rival industries, launder domestic illiberalism through international legitimacy, or steer a technology toward authoritarian defaults. The damage is bounded mostly by the fact that no major country will actually bind itself to anything it does not already want, and inconvenient treaties get quietly reinterpreted into irrelevance.

Then the line that does the work:

"International efforts," therefore, is not a reassuring answer to the governance problem; it is the name of another enormous, unresolved governance problem.

That is insightful, and it is exactly the kind of thing you only learn by being in the room. It is the international-relations version of me watching a founder pitch a slick 311 app to a city council, confident that better bug reporting will finally fix the potholes, with no idea that the road resurfacing schedule is currently frozen because two agencies are in active litigation over who is liable for a water main that cracked the pavement in the first place. The app is not even wrong. It is just answering a question nobody in that room was actually asking.

But here is where I want to push. Dourado is describing one particular institutional form, the intergovernmental treaty process, with real authority. What he is not describing is the full space of things planetary coordination could be.

Theodore Roosevelt said he had always believed that wise progressivism and wise conservatism go hand in hand. He meant it domestically, trust-busting married to institutional continuity, but the formulation travels.

A wise conservatism about how international institutions currently work is exactly what Dourado is offering. Take it. Believe it. The treaty apparatus is captured, slow, and structurally incapable of binding anyone who does not already want to be bound. That is not cynicism, that is a field report.

A wise progressivism about how planetary institutions might work is the other half, and the two are not in tension. Roosevelt's insight was that reform worth doing conserves what functions and replaces what does not. Dourado has handed us an unusually credible account of what does not. The question is what goes in its place.

Here the deeper history is more useful than the recent history. Jonathan Blake and Nils Gilman make a point in Children of a Modest Star that should be better absorbed in AI policy circles: the nation-state as the near-universal unit of global governance is not ancient. It is barely a lifetime old. What we now treat as self-evident, that legitimate coordination across borders happens between sovereign states and nowhere else, only became a fully global default around 1960. Before that the world ran on a messier arrangement of principalities, empires, chartered companies, religious authorities, scientific societies, and civilizational blocs, all exercising real authority without waiting for a seat at Turtle Bay.

Their alternative, which they call planetary subsidiarity, argues for putting authority at the scale where the problem actually lives rather than pooling it at the national level by default. Climate does not respect the Westphalian grid. Neither does a pandemic. Neither, obviously, does a model.

That reframe opens a door most AI governance conversations walk straight past.

Here is the door. You do not need six billion people to agree. You do not need a hundred and ninety-three countries to agree. You need a few hundred people to agree, maybe a few thousand.

The population of humans who actually push the capability frontier is startlingly small. The people training frontier models, the people running evaluations, the people refereeing what counts as a credible safety claim: that is a community, not a nation. And critically, unlike a treaty delegation, it is not composed of parties whose incentive is to use the process as leverage in some unrelated fight. Dourado's central complaint about international negotiations, that everyone is there for a reason other than the stated one, applies far less to a room full of researchers who share a technical vocabulary and a professional reputation market.

Dean Ball's recent argument for independent verification organizations circles this same insight from the policy side. His case is that frontier lab auditing will require extreme technical experts, recruited with lab-competitive pay and real compute access, which government hiring rules structurally cannot deliver. He is right about the mechanism. I want to push on the institutional form, because the form he is reaching for already has a precedent.

The precedent is not the UN. It is the Bulletin of the Atomic Scientists.

In 1945 a handful of Manhattan Project physicists decided that they, not any government, bore an obligation to tell the world what they had built. They did not wait for a treaty body. They started a magazine and, two years later, a clock. Their founding charter was almost aggressively modest: formulate the responsibilities of scientists regarding nuclear energy, and educate the public about what those scientists already understood better than anyone alive. Einstein convened the board of sponsors. Oppenheimer chaired it.

It had no enforcement power. It never has. What it had was the accumulated credibility of people who actually knew how the thing worked, willing to put their names in public on what should happen next. In a field small enough that everyone reads everyone, that turns out not to be a weak mechanism.

Frontier AI has its Manhattan Project physicists right now, distributed across a handful of labs in the Bay Area, London, Beijing, and Hangzhou. What it does not have is its Bulletin.

The raw material is more international than the discourse admits. Academia remains a genuine republic of letters, one that predates the current geopolitical moment and will likely outlast it. AI research in particular carries decades of cross-pollination between American and Chinese institutions that the last few years of export controls and visa friction have strained without severing. The researchers who did their PhDs at Berkeley and now run labs in Beijing. The conference program committees that still span both countries. The author lists on foundational papers that read like a roll call of everywhere.

These people are not incidental. They are the closest thing the field has to intermediaries who are legible, if not always trusted, on both sides of the Pacific. And they are already embedded in an institution with centuries of practice at coordinating across hostile borders. The republic of letters kept functioning through the Thirty Years War. It is not obvious that it cannot function through this.

A body built from that community would not need to resolve the US-China rivalry to be useful. It would need to do something narrower: establish shared technical protocol. What a credible pre-deployment evaluation actually looks like. How you test for cyber-offensive capability without publishing a manual. What a dangerous capability threshold means in terms specific enough to check. These are questions where the honest answer is that only a few hundred people on Earth can give a competent one, and where the current alternative is each lab writing its own homework and grading it.

It would need a home base too, and probably not Washington or Beijing. Track-two diplomacy has always needed neutral ground, and the Chinese diaspora world, Singapore especially, has been playing that role for decades across trade, finance, and technology. A convening built around Chinese and American AI researchers meeting on ground that is trusted by both, with hosts who are fluent in both worlds without being fully of either, is not a new idea. It is the oldest trick in the diplomatic book, just pointed at a new problem. Get people in a room to co-create, as we like to say here in hippie California, and the protocol tends to follow the relationship rather than the other way around.

The enforcement is reputational, which sounds soft until you consider the field. Being the researcher who signed off on the model that caused the incident is a career-defining fact in a community this small. Being the lab that would not submit to the protocol everyone else submitted to is a recruiting problem, an insurance problem, and eventually a customer problem. Ball notes that even a voluntary certification regime generates hard market consequences, because few large enterprises would touch an uncertified lab. That logic does not require a treaty. It requires a critical mass.

One more thing worth saying plainly, because governance conversations tend to drift toward guardrails and forget the road. An institution like this should not exist only to stop bad outcomes. The same critical mass of researchers who can agree on what a dangerous capability threshold looks like are also the people best positioned to point frontier AI at the problems actually worth solving at planetary scale. Accelerating the parts of medical research that are bottlenecked on search and simulation rather than lab time. Modeling crop resilience and famine early warning well enough to act before the hunger, not after. Building the climate models precise enough to make adaptation and stabilization decisions defensible rather than merely hopeful. A Bulletin does not just ring an alarm. It can also point a telescope.

We are looking back at ourselves from a genuinely new vantage point right now, closer to that pale blue dot Sagan described than we have been since the photograph itself, a species with the tools to see its own fragility clearly and, for the first time, some of the tools to actually do something about it at the scale the problems require. That is worth building an institution for, not just defending against one.

None of this displaces what only governments can do. Liability rules, export controls, deciding what gets built inside a border, the security work that requires classified threat intelligence. Roosevelt's formulation cuts both ways and I do not want to be the person who reinvents the failure I opened this essay complaining about.

But there is a large space between the treaty apparatus Dourado correctly describes as broken and the pure national sovereignty that seems to be the assumed alternative. Blake and Gilman would call the thing in that space planetary subsidiarity: authority sitting at the scale where the problem actually lives. For the question of what constitutes a dangerous capability, that scale is not Washington and it is not Beijing.

It is a few hundred people who could probably fit in a conference hotel, most of whom have already met.

Silicon Valley's unworldliness is real. I have watched a version of it play out in water and municipal data for ten years and it is genuinely a correctable failure of imagination. But the correction is not to import the General Assembly wholesale into AI policy. It is to notice that history has already run this experiment once, with physicists, under considerably worse conditions, and that the artifact they left behind is still ticking.

Further Reading

"The Wisdom of Eli Dourado", Tyler Cowen on Marginal Revolution. The post that prompted this one, preserving a tweet that deserved to outlive the timeline. Dourado's vantage point matters here: he has actually sat on the treaty delegations and UN technical committees he is describing, which is rarer in AI policy argument than it should be.

"What Should Be Done" by Dean W. Ball. Thirty-five numbered propositions on the current frontier AI licensing mess and how to get out of it. Ball is a former White House AI policy staffer writing with unusual candor about advice he lost arguments over. His case for private verification bodies over government auditors is the strongest section, and his point that you cannot think your way to safety without real-world deployment experience deserves more attention than it has gotten.

Children of a Modest Star: Planetary Thinking for an Age of Crises by Jonathan S. Blake and Nils Gilman. The Berggruen Institute's argument that nation-state-centric governance is a mid-twentieth-century invention rather than a law of nature, and that planetary subsidiarity is a more honest architecture for problems that do not respect the grid. The historical chapters are the ones to read even if you find the proposal too ambitious.

The Bulletin of the Atomic Scientists, on its own founding. Worth reading the Bulletin's account of why it exists. Physicists in 1945 concluding they owed the world a warning, with no mandate and no enforcement power. The Doomsday Clock is the famous artifact. The founding charter is the more instructive one.

Writing Coda

I wrote this with voice to text after putting the little ones down to bed. A bespoke Claude skill did the fetching and the checking: pulling Dourado's actual tweet instead of letting me paraphrase from memory, tracking down the Roosevelt line, making sure Children of a Modest Star said what I remembered it saying rather than what would've been convenient.

The argument itself, the 311 anecdote, the Bulletin analogy, the instinct to route this through Singapore rather than Geneva, that's mine, worked out in real time across a few passes back and forth. Which is more or less how I want AI to fit into the writing: not generating the thesis, but clearing enough of the research friction that I can spend the walk actually thinking instead of half-remembering a footnote.

2/ We found a little nook and I read the story to her, a silly little riff written by large language models, lots and lots of math making a series of words that sounded good and were quite nice for that moment, like many others. A screen made that possible. Nobody's writing guidelines about that use case.

3/ The screentime discourse has a tell: adults angsting about kids' screens are usually narrating our adult world's relationship with devices, sometimes even their own fraught use of screens. Doomscrolling, checking Slack at dinner, being rude in comment sections. We're diagnosing the kid to avoid diagnosing ourselves.

4/ "Screentime" as a category makes about as much sense as "reading time." A kid decoding volcano formation and a kid getting fed rage bait by an algorithm are doing completely different things. Lumping them together is the whole problem.

5/ These devices are also, not incidentally, portals to nearly all of codified human knowledge, with AI now making that knowledge conversational. That's not a footnote to the addiction story. That's the other half of the story.

6/ It rhymes with food. We live in an age of abundance, calorie-dense, engineered for craving, genuinely nourishing all at once, sometimes in the same bite. The original food pyramid heavily overemphasized carbohydrates while lumping all fats and proteins together, ignoring critical distinctions between healthy nutrients and processed junk.

7/ AAP-style guidance (30 min a day, screens off at X age) is the food pyramid of parenting advice. Not evil, just not even wrong. It's a number pretending to be a framework.

8/ The actual questions are harder and don't fit on a poster:

9/ Can your kid regulate their emotions without a screen as a pacifier? Can they disconnect from a device without falling apart?

10/ What are they actually doing on a device? Are they learning something or just having their senses dulled with bright noises?

11/ Do they understand what it means when words or images they send land on someone else, at scale, permanently? (Adults, notably, are bad role models here.)

12/ None of that shows up in a timer. All of it shows up eventually, in how the kid actually behaves.

13/ Raising kids Amish isn't on the table. They're growing up in a connected world and pretending otherwise just delays the reckoning to a moment with less parental scaffolding around it, not more.

14/ A more useful target: healthy device relationships modeled by the adults in the room. Not checking texts and email compulsively. Not treating notifications as commands. Kids learn this by watching, same as everything else.

15/ Some of the angst about "kids online" might actually be angst about what adults see online. Outrage cycles, bad faith arguments, "dark psychic" energy. We're extrapolating from our own feed to their future, which is its own category error.

16/ There's also a design question underneath all this. A Chromebook handed to a six year old is a general purpose adult tool with training wheels bolted on. That's a strange default for early childhood.

17/ Compare that to bespoke devices built for a narrower job: a Yoto for audiobooks, a Dex for translation. Purpose-built tools teach kids to navigate specific information flows before we hand them the entire internet in one interface.

18/ The deeper question nobody's actually asking: what is learning, for this generation, in this specific community, right now? Not learning in the abstract. Learning as it actually happens in La Crescenta or wherever you live, given the tools that exist today.

19/ That's not a question for a pediatrician's guideline or a screen-time app. It's a question for parents and adults to take responsibility over as citizens.

Dewey's Great Community in a Digital Age

In the tech world right now, the hottest idea in education is AI tutoring. Personalized, mastery-based, infinitely patient digital guides - the dream is that every student could have the kind of one-on-one tutor once reserved for royalty.

https://pioneeringspirit.xyz

20/ Screentime isn't the only variable. Emotional regulation, critical inquiry, and media literacy really matter. The screen is just where we've decided to focus because its so easy to measure.

21/ Taylor still has that petal, pressed somewhere in a book by now. The story I made her exists because of a device. So does the version of "kids and screens" that keeps everyone anxious. Tools are tools, and ultimately it’s how we choose to use them.

Writing coda: I riffed a bit with claude and then used a bespoke skill to pen this piece. Image uses a titles.xyz model courtesy of VGR. I had this written as a tweetstorm to signal a Very Online person poking fun at the anti-device Amish types.

Geoffrey West and Luis Bettencourt had published their scaling law work. Michael Batty had written "The New Science of Cities." Santa Fe Institute researchers were finding power law relationships in urban data that felt analogous to metabolic laws in biology. The pitch was heady -- cities as organisms, quantifiable, governed by deep mathematical regularities that transcended their particular histories and politics. I was skeptical but genuinely curious, and I tried to work through what I actually thought.

Reading it again a decade later, a few things stand out. The basic epistemic cautions still hold. Cities remain stubbornly resistant to unified theory, and the scaling law claims have aged into something more nuanced and contested than their early proponents suggested. What surprised me more was how much the informatics speculation in the middle section pointed toward things that actually happened: real-time urban data, longitudinal learning analytics, the attempt to use web-scale data to ask better research questions. We got a lot of that.

We also got machine learning systems optimizing for prediction while remaining largely mute on causation, which is its own kind of epistemic problem. The illegible messy tangled crooked timber of human life in cities still remains, in many ways with an only larger disconnect as increasingly obsolete institutions largely persist through a zombie-like logic in the midst of dramatic information technology change. The rest of this post is the original paper.

A New Science of Cities? 

“There is no single paradigm that has come to dominate our understanding of cities, and as we will elaborate here, there is unlikely to be one.” 

– Michael Batty, “The New Science of Cities”

In the past few decades, a diverse collection of physicists, urban planners, computer scientists, engineers and other technical disciplines have renewed interest in the quantification of urban phenomena.  Proponents argue the discovery of urban scaling laws, increasingly sophisticate agent based models and the increased availability of data brought on by the digital revolution point towards the development of an entire new science.  Whole research programs have sprung up to study these issues more deeply.  What should we make of this conjectured new science of cities?  

Two points stand as unequivocal.  First, it’s early days.  Urban scaling laws hold for particular countries – primarily developed ones – across the past few decades.  Yet will the same sublinear infrastructure scaling relationship hold for Mumbai or a newly developed city in China or a colony on Mars?  Will those patterns hold across massive technological change?  A robust dataset that spans centuries for every city within human civilization would enable answers closer to the level of generality to which this new science aspires.  

Second, the tools this science uses may be new but the questions it raises are decidedly not.  The aspiration towards a deeply quantitative understanding of human affairs is at least as old as Adolphe Quetelet’s 1835 “Essays on Social Physics” and his findings of the (then) surprisingly invariant annual murder count for certain cities across time.  Considering murder an act of free will, Adolphe expected a high variance that he did not find.  That mathematical pattern encouraged Adolphe’s aspirations towards a social physics that has not come to pass.

More broadly, techniques developed in the natural sciences have long history of application to societal questions, not least the development of statistics itself!  Ordinary least squares regression – a technique essentially ubiquitous today across the across the social sciences – began as a method for combining astronomical observations.  This quantification of human affairs, particularly in economics, has endured criticism portraying the effort as simplistic and reductionist at best and useless at worst.  Further quantitative models played a pivotal role in the fires that decimated the South Bronx in the late sixties and the 2008 financial crisis. 

That is not to say data and quantitative methods have played no part in recent human progress.  Over the past two centuries, increasingly robust census demographic data and quantitative modeling have improved everything from infrastructure development to school choice.  It is hard if not impossible to imagine the tremendous human progress over the past two centuries without improved data collection and manipulation tools.  How would one even begin to conduct the Federal Reserve’s business today with pen and paper?

What then are we to make of the potential for understanding cities with new data collection and manipulation tools pioneered in the natural sciences?  Looking at the same challenge from another angle, what is the current state of social science inquiry?  Where are the opportunities for improvement?

When is ceteris really paribus?

“In contrast to the situation of controlled experimentation that motivated much of the development of modern statistical methods, where variables of interest can be varied orthogonally, the political analyst typically confronts a situation where an assortment of equally plausible theories suggest several closely related (and therefore highly correlated) variables as possible causal factors.”

– Phillip Schrodt, “Seven Deadly Sins of Contemporary Quantitative Political Analysis”

Cities embody dense causal webs.  A huge variety of natural environment, artificial infrastructure and human factors coalesce in a single place.  Within that complexity, where does one look to find what caused a given phenomenon?  

Consider a basic education question: how does class size impact student achievement?  A researcher might study how test scores vary by class size across cities.  Yet any number of other relationships could be related to class size and really be the driving cause.  Class size could relate to the level of wealth in the surrounding community which affect the number of opportunities students have outside of school.  

A researcher might “control” for those factors with other variables such as neighbor home prices in a linear regression, but that model assumes independence across the explanatory variables.  And as Phillip makes clear, that is not generally a robust assumption:

 “As a consequence, linear regression results are notoriously unstable—even minor changes in model specification can lead to coefficient estimates that bounce around like a box full of gerbils on methamphetamines. This is great for generating large numbers of statistical studies… but not so great at ever coming to a conclusion.”

Collinearity can result in coefficients oscillating dramatically or even changing sign across specifications.  Further, in a casually dense system like a city or an urban public school, we likely will not account for causally relevant variables within our model.  This omitted variables will bias the coefficient results of the included variables.  

Perhaps then, we should not be surprised at this 1979 statistical meta-analysis summary of the previous few decades of research into the impact of class size on student achievement.

“There is no point in recording the obvious about class size: that teachers worry about it more than nearly everything, that administrators want to increase it, that it is economically important, and the like.  The problem with class is the research.  It is unclear.  It has variously been read as supporting larger classes, supporting smaller classes, and supporting nothing but the need for better research.”

– Gene Glass and Mary Smith, “Meta-Analysis of Research on Class Size and Achievement”

The standard social science research answer to this challenge is simple: random controlled trial.  Randomization minimizes identification effects – the impact of what else relates students to being in a given class size for instance – and allows for statistical inference into the impact of a particular policy treatment.  And a well-designed experiment indeed happened to test the efficacy of class size reduction.  

The Tennessee Star Study randomly assigned teachers and students to different sized classes, ranging from 13 to 25 students.  This blue ribbon effort was used to justify smaller class sizes in Tennessee but also pivotally in California.  Here was a model of social science research still used as a canonical example in econometric textbooks.  

There’s a simple problem.  It didn’t really work in California.   The class size reduction passed just six weeks before the start of the school year and implementation was a nightmare.  High poverty school districts struggled to hire qualified teachers to staff the new class rooms and inequities increased between rich and poor schools.  Across the state, the biggest legacy of the policy was the now common school yard bungalow classroom.  What lessons can we learn from this failure about the opportunities for improvement within modern social science?  

First, note that many social scientists will likely argue that the fault lay not in the research but rather in the improper application by policymakers.  Yet the common practice among social science research is to state all the statistical reasons for uncertainty – the problem of generalizing from a specific random control trial to other situations for instance – and then moving on to make policy recommendations in no uncertain terms. 

So what are policymakers supposed to do after decades of yes, no, maybe so education reform answers from the research community?  Ignore blue ribbon research with strong conclusions?  California had money to spend on education and chose a policy aligned with social science research.  What else do social scientists want from policymakers?  The world demands decisions, and public school administrators and policymakers must size classes.

Second, class size illustrates the faults of focusing on average treatment effects.  Why do we assume class size can be optimized like tuning the batch size of an assembly line process?  What would even be the ideal experiment we would run to test the class size optimization hypothesis?  Would a statistically robust sample of random field trials of randomly chosen students assigned to randomly assigned teachers in random school sites in a wide variety of student to teacher ratios really determine the relationship of class size to student achievement?  

These questions illustrate the methodological incoherence of the current social science research paradigm.  Yet even if such a research agenda of random control trials were possible, how do we know that the relationship is constant across time?  Might students of different cultural time periods and technological contexts learn differently in the same sized class?  Why should we be confident that test scores really reflect student achievement over the long run?  Why base our research on average outcomes anyway?  What logic do we have to support the assumption that optimizing for average students is even possible let alone desirable?

The dime store metaphysics embedded in the class size research agenda fails a basic task of intellectual inquiry: questioning assumptions.  Why should we have one class size to fit all students?  More broadly, this type of one-size-fits-all thinking seems far too endemic to current social science thinking.  Why should understanding average treatment effects be the default goal of social science?  After all: 

“Human nature is not a machine to be built after a model, and set to do exactly the work prescribed for it, but a tree, which requires to grow and develop itself on all sides, according to the tendency of the inward forces which make it a living thing.”

“Such are the differences among human beings in their sources of pleasure, their susceptibilities of pain, and the operation on them of different physical and moral agencies, that unless there is a corresponding diversity in their modes of life, they neither obtain their fair share of happiness, nor grow up to the mental, moral, and aesthetic stature of which their nature is capable.”

—John Stuart Mill, On Liberty (1859)

Yet social scientists, founded in wake the industrial triumphs of the late 19th century, seem to have ignored John’s exhortation.  So how might we do better?  Where are the opportunities for a science of cities to inform better social science research?  Why does informatics matter?  What strategies do new digital tools enable?  

The new informatics-enabled frontier 

“The idea that no one really knew how to run a government led to the idea that we should arrange a system by which new ideas could be developed, tried out, and tossed out if necessary, with more new ideas brought in — a trial and error system. This method was a result of the fact that science was already showing itself to be a successful venture at the end of the eighteenth century. Even then it was clear to socially minded people that the openness of possibilities was an opportunity, and that doubt and discussion were essential to progress into the unknown. If we want to solve a problem that we have never solved before, we must leave the door to the unknown ajar.”

– Richard Feynman, “The Value of Science”

How might scaling laws and other ideas developed in the nascent science of cities inform the class size debate specifically and education reform discussion more broadly?  These megatrends abstract too far from the day to day activity of a classroom to be useful directly but offer insight into potential avenues of productive inquiry.  So what hypotheses might we generate from the superlinear scaling of intellectual productivity as measured by patent production for instance?  

I conducted a naïve regression of patents per capita on drop-out rates to illustrate the potential for further inquiry.  Higher patents per capita correlate with lower dropout rates.  In many ways, this result is not surprising.  Education does not occur a vacuum after all.  Children learn outside of formal public schooling from the world around them, and we would expect the intellectual vibrancy of an area – as measured by patents produced among other potential metrics – to positively correlate with student achievement.  

What if, taking inspiration from Michael Batty’s work on network connections through urban environments, we explored how educational attainment, skill development and other measures of intellectual vibrancy change throughout a city spatially across time?  What if we used new data streams enabled by the web – peer certified skills verified on LinkedIn or intellectual discourse on social media newsfeeds – to understand how conditions outside of school impact student achievement?  

And why not use more robust data sources than simply end of the year scantrons to understand student achievement?  Why not track how students learn throughout the year as evidenced by blended learning tools like Khan Academy quizzes?  This data certainly will have its imperfections but could definitely add insight into students’ learning trajectories within the school year.  Going further, why not longitudinally track how students actually apply their skills after graduation through interfaces like LinkedIn among other tools?  Together would this not enable us to understand student learning far closer to the simple intuitive perspective that students learn from the entire world around them – not just what standardized tests capture after sitting in a formal classroom?

At the very least, this research program would enable us to get a picture of how well students are learning the skills the world demands.  This research would inform how to best build bridges between public schools and employers in the broader city, an urgent question in the education community:

 “Everywhere I go, I talk to employers saying we can’t find the skilled workforce we need, and I talk to young people in schools saying we want opportunities and we can’t find them.  Somehow the matchmaking, the connection (between business and education) doesn’t happen in too many places.”

– U.S. Secretary of Education Arne Duncan

That challenge has nontrivial economic and future scientific research implications – not to mention the moral fiber of the nation.

This informatics enabled approach to education research would offer additional improvements to the status quo epitomized by the class size question.  For starters, by measuring student achievement in more granular time scales both within public school and beyond graduation, we can examine what makes sense for more specific types of students – such as an athletically inclined math geek – and even individual students.  That research rests on a sounder philosophical framework than simply assuming all students are roughly similar and optimizing class sizes for average student achievement.

Further why not improve the connection between research and practice?  The feedback loop between experimental design, random field trial, study and policy change spanned decades in the class size research question, creating large disconnects between the maps researchers build and the territory of actual practice.  Imagine an alternative approach.  What if we randomly assigned different nonprofit programs aiming to connect students to business mentors as a matter of standard practice similar to how web companies regularly conduct A/B tests of new features?  Experiments at the timeline would enable policy changes in response to what we learn about local conditions rather than generalizing based upon the few scant studies.  

Those experiments assume an enlightened and agile school administration but there is a lot of opportunity nonetheless – particularly at the scale the web offers.  In conducting those nonexperimental analyses, we might use web scraping, access to organic social media data, real time blended learning assessment data and other web data such as surveys mashed up with standard student achievement metrics.   The web also enables creative aggregation of qualitative and anthropological data.  In addition to standard classroom visits, imagine crowdsourcing first hand accounts of how local neighborhoods interact with specific schools. 

With that data in hand, we might ask questions like: what is the optimal algebra class size for students that successfully understand geometric visualizations but struggle with open ended word problems?  We can further contextualize those curricular questions by neighborhood characteristics and the broader learning environment students live in (whether they have internet access at home or not for instance).  Rather than ask what the ideal class size is for all students at any time anywhere, we might work to improve the classroom environment for this unique group of students at this unique stage in their learning trajectory.

These questions show how social science can use lessons from the nascent science of cities to generate contextual hypotheses and flexibly use new informatics tools to gather the appropriate data radically easier than the research world that existed before the web.  Too often social science researches questions not for any theoretically cogent reason but rather simply because those are the questions that accessible data allows.  Case in point class size and standardized test scores.  Yet as we have seen, new informatics tools enable us to inform whole new categories of research questions with new data sources such as how classroom characteristics correlate to lifelong learning and career development.  In short, we can ask questions based more on what boundaries we understand to be actually relevant rather than the categories we impose on the world.  

What are the broader takeaways for the nascent science of cities?  What does this new research approach mean for how we understand cities specifically and human civilization more broadly?  What have new informatics tools not changes?  What challenges endure?  

Computation, causality, and the enduring challenges of human inquiry

“…a single problem: how to combine the perspective of a particular person inside the world with an objective view of that same world, the person and his viewpoint included.”

– Thomas Nagel, “The View from Nowhere”

These above debates about class size may have acquired a particular focus in the context of twentieth century public education, but the underlying questions about how we understand human affairs have been debated for millennia.  How do we know what we know?  What knowledge is unbiased by our particular opinions and feelings?  How can we make sense of the world in light of Thomas’s eloquently posed problem?  

These questions acquire special urgency in the context of human affairs.  Cities are not merely physical systems that we study.  They are collections of citizens with opinions that may conflict with our very research agenda!  More deeply, there are several reasons why we will not achieve the same precision modeling cities as we do purely physical systems.  

First, any model cannot include every relevant variable and will always include specification error.  Achieving the right level of abstraction and choosing which variables to leave out is especially paramount when mapping human behavior, a point Jorge Luis Borges eloquently makes in his parable of a map the size of empire it attempted to describe.  

Second, we will always have measurement error.  In addition, to data collection errors – typos in data entry or a biased instrument – we can also have disagreement about what constitutes the right metric for a given variable.  See the debate about measuring student achievement for instance.  Reconciling those disagreements requires the researcher to make choices.  

Third, cities embody incredibly complex and chaotic systems.  The nature of chaotic systems is that they can achieve widely divergent results based on small changes in inputs.  How we choose to account and appropriately model that variance represents another choice on the part of the researcher. 

Together these choices influence how we measure, model and thus scientifically understand the systems we study.  And unlike the natural sciences, we do not study a world that we can plausibly isolate from the influences of human agency.  What makes cities such fascinating areas of study results from the incredible interactions and emergent phenomena that bubble up when so many humans – each with their own plans, projects and objectives – live, work and play in such close proximity to each other.

Further, unlike the natural sciences, these research choices cannot appeal to simply understanding the world as it objectively is.  Researchers have their own set of political, economic, social, cultural and other biases that affect how any human understands any other.  Scientists studying cities cannot come at these questions from a view purporting to originate from nowhere.  In fact, the very rationale for studying cities scientifically is to improve how cities operate!  Such normative appeals by definition necessitate a view built upon moral judgments decidedly coming from somewhere.

These issues stemming from human agency come to a head when working to understand causality in the context of cities.  We do not have merely Hume’s billiard balls to contend with.  Rather than imagining simply several possible outcomes when one isolated billard ball strikes another isolated ball, in cities we study infinitely more complex causal interactions.  Class size might plausibly impact student achievement contingent on a nurturing home environment or a safe neighborhood.  Perhaps further those relationships only hold at certain thresholds. 

The legal system expends tremendous energy to uncover whether an event would have occurred “but for” the cause in question.  Yet researchers looking to not just understand what happened in the past but to build theories with predictive power for the future must reach a higher bar.  We cannot simply understand isolated causes but need to work to understand how higher order interaction effects result in different outcomes.  Class size may help or hurt student achievement depending on teacher quality for instance.  

Further we must remember that omitted variable bias acquires special urgency when we seek to a causally dense environment like a city.  A variable unspecified in our model might flip the sign or dramatically alter the magnitude of another variable’s relationship.  And in inferring from our imperfect models we must remember that understanding the direction causality flows, if it even can be said to have a single direction, can be a matter of subjective interpretation.  What causes a school to be good?  What is the causal connection between school quality and poverty?  

In a deep sense, we should not be surprised that causal questions are so tricky.  Causes represent a special kind of knowledge, a tool we humans create to make sense of the world around us.  As such, causes are irrevocably embedded with the biases stemming from our particular position in time or place, a challenge that acquires particular urgency when studying cities.  Our understanding of cities and civilization remains contingent on the time, place and political economy arrangements that produced such ideas.     

Embracing contingency with both arms 

“All models are wrong.  Some are useful.”

– George Box, Statistician 

What then are we to do?  What should we make of a science of cities’ aspiration to a unifying predictive understanding of “the structure, dynamics, growth and organization of cities?” We can appreciate that our study of cities or really any aspect of human civilization is not immune to tradeoffs.  

We might choose to ignore the challenges of contingency and assert that new science of cities articulates universal objective truths irrespective of the particular conditions that gave rise to their formulation.  We might aspire to “know everything” about a city and not bother with appreciating how understanding shifts with perspective.  There lies the road to metaphysics and dogmatism.  

Alternatively, we might choose to obsess over the complications created by the deep links between contingency and our all too human condition.  We could follow in the path of postmodernism and assert that meta-narratives are dead, that an understanding that extends beyond our finite position in time and space is only a naïve delusion.  There lies the road to stultified skepticism.

Or we might choose a third option.  Rather than choosing a single path, we might recognize that many valid perspectives exist for understanding cities.  We can appreciate that the new scale of digital data allows us to see cities in a new light.  By developing an understanding of the complex interactions between place, flows and networks, this nascent science of cities allows greater appreciation for the landscape megatrends that bound urban life.  

Grand narratives like patent superlinear scaling laws and theories about agglomeration effects do not determine the universal future of our cities.  A science of cities cannot tie all of humanity up in a neat mathematical bow.  Instead it proceeds from the deep intuition that we as humans do share common conditions regardless of our contingent circumstance.  And such a science aims to articulate the deep patterns that tend to emerge from the intersection of so many human experiences in a single place.

By highlighting commonalities and how communities differ, such insights can inform the choices that we make at a local level in areas such as education.  Other areas of inquiry such as established areas of social science and the ground level anecdotes of practitioners equally can offer insights.  Those decisions do not have right answers and no model can replace human judgment.  Rather such questions engage debates about justice that have been ongoing for thousands of years.  

The English word City itself stems from the Latin word civis or citizenship.  Perhaps then there is no better function for a science of cities than to serve as a forum for the great debates about the future of our cities.  By articulating megatrends that affect cities and pioneering computational methods that cut across disciplines, this inquiry can bring together diverse disciplines in a common undertaking – an academic corollary to the nature of cities themselves.

So in sum, what should we make of a new science of cities?  A tremendous opportunity to engage the questions that have challenged humanity since the dawn of civilization.  

References

• Michael Batty, “The New Science of Cities”

• Phillip Schrodt, “Seven Deadly Sins of Contemporary Quantitative Political Analysis”

• Gene Glass and Mary Smith, “Meta-Analysis of Research on Class Size and Achievement”

• John Stuart Mill, On Liberty (1859)

• Richard Feynman, “The Value of Science”

• Arne Duncan, U.S. Secretary of Education

• Thomas Nagel, “The View from Nowhere”

• Stephen Stigler’s The History of Statistics: The Measurement of Uncertainty before 1900.

• 
  

Further Reading

Dewey’s Great Community in a Digital Age

Pioneering Spirit

Sep 4, 2025

In the tech world right now, the hottest idea in education is AI tutoring. Personalized, mastery-based, infinitely patient digital guides — the dream is that every student could have the kind of one-o...

0 supporters

Support

Tommy sang it like a man who meant it. What I heard in it, echoing off that ancient rock, was the wild and free spirit that lives deep in the human heart, a flame that burns brightly in anyone who cares to nurture it. I filed the song away as charming. Then last summer I finally grew tomatoes worth eating, and I understood that Tommy had been telling the truth.

They were not impressive tomatoes. Three plants in pots, using store-bought potting soil -- nothing that would make a real gardener look twice. But they were undeniably, obviously better than anything from a store, and my toddler ate them straight off the vine. That alone was worth the season.

Most vegetables are not worth growing yourself. The cauliflower I tended for months tasted exactly like cauliflower from the supermarket. The French beans were fine, and my toddler liked them, and they fixed nitrogen back into the soil after the tomatoes had stripped it out, so there was a logic there. But the tomatoes were different. There is something in a homegrown tomato that storage and transit systematically destroy. The effort produces a genuinely different thing.

This is older wisdom than Guy Clark. It is the wisdom Voltaire arrives at, exhausted, at the end of Candide. After his hero survives earthquakes and wars and the collapse of every cheerful theory about why the world is the way it is, Voltaire gives him a single instruction: il faut cultiver notre jardin. We must cultivate our garden. Stop theorizing about the best of all possible worlds. Go tend the thing in front of you. It is not a counsel of retreat so much as a counsel of contact, the recognition that there is a kind of knowing you only get with your hands in the dirt.

That distinction -- between theorizing about the world and making contact with it -- is one I keep circling. Venkatesh Rao has a framing I find useful. Imagine you are on Buckminster Fuller's Spaceship Earth, hurtling not just along its familiar annual orbit but through a tangle of hidden, curled-up dimensions you can't see but can feel, the way you feel motion sickness in your gut. Two kinds of people emerge in response to that nausea. The first group retreats from it, stays close to things they can see and touch, and eventually concludes the hidden dimensions aren't real at all. The second group leans into the nausea, builds elaborate maps of the unseen trajectories, and comes to trust the maps so completely it forgets they are maps. The grass-touchers and the abstract map-makers, locked in a war neither can win. Rao's resolution is a third move: touching time. The practice that lets the map make contact with the territory, so the abstraction stays honest and the grass-touching stays meaningful.

There are ways to touch time at geological scale, what the Greeks knew as aeon -- to float through the Grand Canyon and feel the aeon in a glimpse, the way you can look at a layer of Ordovician limestone and know you are staring at the floor of a vanished sea, the moment four hundred million years ago just before the mass extinction swept the brachiopods away. That kind of contact does something to a person. But it is hard to sustain across a Tuesday. A garden is a way to wrestle daily with chronos, the relentless ticker of schedule and obligation, and kairos, the ripe moment that arrives unbidden and asks you to be present for it. Both come for you when you are responsible for a toddler and a newborn and duties to public service at the same time. The garden doesn't resolve that tension. It gives it a form you can work with.

I encountered the map-territory nausea viscerally, long before I had words for it, as a Coro fellow. The whole program was an exercise in being thrown between institutions fast enough that you could feel the vertigo -- from a union hall to a corporate boardroom to a campaign office, each with its own complete and self-consistent map of how the world works. I have been chasing the reconciliation ever since.

The garden insists on the contact. I learned about hard water the hard way this year. The water coming out of my La Crescenta tap carries a heavy load of dissolved calcium and magnesium, which over time crusts on leaves, clogs drip emitters, and nudges the soil pH into ranges that lock out the nutrients a plant needs. This is not exotic. According to the U.S. Geological Survey, roughly eighty-five percent of American homes have hard water, with the arid Southwest among the hardest. You can read all this. But reading about it and watching a tomato yellow despite your best efforts are different kinds of knowing. The territory supplies something the map cannot.

What I have come to appreciate is how AI changes the posture I bring to that territory. When aphids showed up on the peppers, when a colony of mushrooms erupted overnight in a bed I'd amended with wood chips, when I wanted to know whether my back slope catches enough February sun to start cool-weather crops given the local solar azimuth, I could ask and orient in the same motion. The point is not that the tool knows things a book doesn't. The point is the flow it enables: observe something in the world, ask what's going on, decide what to do, and act, all in one continuous loop, often with a toddler on my hip. It collapses the distance between noticing and understanding and acting. The aphids are best handled by inviting their predators. The mushrooms are probably benign and maybe helpful. February sun in these foothills is marginal for brassicas. None of that is hard to find. But finding it while standing in the garden, in the rhythm of the work itself, is what turns information into a different way of being present in the world.

I keep a Notion database I call the Bot Garden, a web-like log of ideas and notes in every stage of development. In this I am like a lot of wild-spirited, digitally native millennials -- a tangle of half-formed thoughts accreted over years, cross-linked and never quite organized. What's changed lately is that I've started running something like the Karpathy protocol on it: letting an AI work the raw pile of captures, interlinking them, maintaining an append-only log, compiling the tangle into something more legible than I could keep in my own head. Some entries are seeds, some are seedlings, some have been there long enough to put down roots. The parallel to the dirt garden is not decorative. The same practices make both productive: regular attention, willingness to prune, patience with slow growth, acceptance that not everything germinates.

What the physical garden has given me that the digital one hadn't is the weight of irreversibility, and the strange satisfaction of seeing your own action manifest in the world. Even three pots on a slope is a real change you made to the surface of this spaceship, a visible result on this pale blue dot adrift in a very large and indifferent cosmos. An idea in Notion can wait forever. A tomato plant runs on its own schedule.

This points at something about how time actually works in our institutions, and how badly. They run on two clocks and only two. There is the indefinite future, where anything to do with actual foresight goes to wait politely forever, and there is the immediate present, which arrives exclusively in the form of some very important person needing something right now. The seasonal middle distance, the register where a garden lives, where you act today because of what ripens in eleven weeks, barely exists in that world. The garden restores it. Things ripen. Seeds fail. Seasons turn. The irreversibility isn't a bug in the system. It is the system.

My father spent his career thinking about water -- moving it, governing it, allocating it across a very large and thirsty landscape. On that same Colorado float trip, while Tommy sang about tomatoes in the morning light, my dad was channeling his inner Cassandra, repeating it like a mantra: we're headed for hell, we're headed for hell. He didn't mean any rapid. He meant decades of accumulated intuition reading where things were flowing, the hydrologic reality of a shrinking river meeting the intractable politics of the water wars. That kind of knowing doesn't come from reports. It comes from a lifetime of touching territory until you feel the trajectory in your gut.

I think about this when I'm watering. A minute with a hose gives you a concrete sense of what it means to deliver water at scale -- what a gallon costs in time and energy and infrastructure, what it means for all those gallons to have to come from somewhere, what shortage actually feels like when the flow diminishes. My body knows it now. The water work I do is still mostly maps, demand projections and infrastructure models, but I trust those maps a little more, and distrust them a great deal more precisely, for having spent some hours with the real water and the real dirt.

The word "tend" comes from the Latin tendere: to stretch, to extend, to direct one's course toward something. It shares a root with attention and intention and tendency, the whole cluster of ways we describe the mind reaching toward what it cares about. To tend a garden is, in the oldest sense, to stretch toward it. To extend yourself into the gap between where you are and where the plant is, between what you know from reading and what the soil is actually doing, between the map you carry in your head and the territory that grows indifferent to your plans and wonderful in spite of them.

That stretching is the thing. Not the harvest, not the tidy beds, not the optimization of yield per square foot. The opportunity the garden offers is to reach across the divide between maps and territories, to feel through the gap, to grow ourselves alongside the plants and the inevitable wild spaces -- the volunteer tomato that comes up where nothing was planted, the mushrooms that arrive without invitation, the toddler who eats the beans straight off the stem before you can stop her. Tending, in this sense, is not a metaphor for something else. It is itself the practice. Voltaire knew it. Tommy knew it. The garden, indifferent and patient and ripe on its own schedule, keeps teaching it.

References and Writing Coda

I had fun summarizing my garden related inquiries over the last year in Gemini.* I copied those into Claude, brain dumped a bit and iterated a few times to generate this post. Image credit: I used the text to image feature and VGR's bucket art model available on Titles dot xyz. I'm pretty happy with it 🙂

*I'm pretty basic in that I pay for Claude and use Gemini regularly for routine artificial librarian type of tasks since I've yet to encounter limit to the number of queries .

Song on Homegrown Tomatoes

Stone and Stream: A lineage in motion (Part 2)

Pioneering Spirit

May 6, 2025

Part one of the story of my dad lives here. Chapter Two: In the Desert, a Well“The work is long. The road is dry. But somewhere under all this dust, there is water.”Before the father becomes a myth, H...

1 collected

Collect

Touching Time

Or, the new adventures of Spaceship Earth

https://contraptions.venkateshrao.com

There's a Google I/O every year. Developers line up for keynotes, product announcements, the whole ritual. I/O -- input and output -- is the founding concept of computer science, the basic grammar of how information enters and leaves a system.

There are also IO towers on reservoirs. Physical structures. Pipes and valves that manage the literal intake and release of water from storage. Same abbreviation, same underlying logic, completely separate worlds of practice.

Like most people outside the water industry, most of the engineers at Google I/O have never thought about a reservoir tower. Most of the water operators who've inspected one have never attended a web developer conference. But they're both reasoning about input, throughput, and output every day. The concepts rhyme across domains in ways that neither side tends to notice.

This kind of double-meaning is everywhere once you start looking, and it does more than just create puns. It creates membership. It's the logic of the shibboleth.

"Developer" means one thing in a real estate office -- someone who acquires land and builds on it -- and something else entirely in a software shop -- someone building on top of a platform or API. Both meanings share the underlying idea: development is building on top of an existing substrate. But if you use the word wrong in context, you immediately mark yourself as an outsider. You don't belong to that conversation yet.

The duck curve works the same way. If I drop that phrase in front of a utility engineer, it's a recognition signal, an automatic credentialing. She knows I know something about grid management and solar intermittency. If she starts explaining what the duck curve is, she's decided I don't. The term itself is doing social and epistemic work that goes far beyond its technical content.

These are shibboleths. But they point at something more interesting than mere gatekeeping.

C.J. Cherryh's 1994 science fiction novel Foreigner is built around this problem at species scale.

The setup: a human ship falls out of hyperspace at the wrong destination, crashes into contact with an alien civilization -- the atevi -- at roughly steam-engine levels of technology but with their own fully developed culture, society, and psychology. The first contact goes badly. War follows, then a long and uneasy peace.

To manage that peace, the two sides build an institution: a single official human translator, the paidhi, who mediates all contact between civilizations. One channel. One diplomat with full language fluency, living in the atevi world, serving as the only permitted conduit between species. The protocol is designed to prevent misunderstanding from escalating back into conflict.

The protagonist, Bren Cameron, is that translator. Fluent. Credentialed. He knows the words.

And the book is, in part, about everything that words don't carry.

Language learning confronts you with this. When you learn another language seriously -- not phrasebook competence but actual fluency -- you start to realize that each language isn't just a different encoding of the same content. It's a different structure of perception.

The Eskimo-words-for-snow observation has been overcited and partly debunked in the details, but the underlying point holds: cultures accrete vocabulary around what matters to them. Distinctions that one language makes in a single word, another distributes across a whole phrase, because the distinction either mattered more or mattered differently. Some cultures have had number systems that categorize the world in ways that make Western arithmetic look like one option among many. Language shapes what's easy to think, not just what's easy to say.

What Cherryh explores is the layer beneath that. Even after Bren achieves genuine fluency -- not just vocabulary and grammar but the cultural context, the trust, the implicit value hierarchies -- there's still a gap. The atevi have a different psychology around loyalty and association. They experience something that doesn't translate cleanly into the human concepts of friendship or love. Bren can define the term intellectually. He can use it correctly in sentences. He still can't feel it from the inside.

It's only through high-stakes forcing functions -- near-death situations, moments of genuine crisis where the usual protocols break down -- that something gets through that symbolic systems can't carry. Actions speak louder than words, yes, but the book earns that cliche by making you feel what it actually costs to close the gap.

AI is, among other things, the most powerful translation technology in human history. Real-time, across most major language pairs, at a quality that would have seemed like magic twenty years ago. The technical layer of translation -- converting words and phrases and grammatical structures from one encoding to another -- is increasingly solved.

What's left is everything Cherryh was writing about.

The embedded assumptions about the relationship between individual and family and society. The implicit hierarchies of what's polite and what's presumptuous. The specific cultural logic of what you owe someone versus what you choose to give them. The thousand small things -- and I keep thinking about something as mundane as how cutlery is arranged on a table, because it stands for the whole enormous category of practices that are completely invisible until you get them wrong -- that don't live in words at all.

Translation across those registers requires something closer to what Bren Cameron was doing. Not just decoding, but dwelling. Not just knowing the term but having the context it grows from.

The IO tower and the Google I/O session might share a word. But to move fluently between those worlds -- to actually carry meaning back and forth, to let insights from one domain illuminate the other -- takes something more than vocabulary. It takes enough of a shared substrate that the concepts can actually land.

We're building ever more powerful systems for the easy half. The harder half is still mostly patient, slow, human work.

Writing coda: This post began as a voice memo. A bespoke Claude Sauntercast skill cleaned the transcript, preserving the walking cadence while making it readable. Then a second skill, trained on past Pioneering Spirit posts, shaped the cleaned sauntercast into the essay above. Everything is translation, including this.

Silicon Valley's pirate techies increasingly collide with public life. Reflecting on what actually might be going on today, I keep going back to Colin Woodard's The Republic of Pirates and Linda Colley's The Gun, the Ship, and the Pen. Read together, they reframe the early 1700s as a workshop of political technology, where discarded sailors and anxious empires were both running experiments on the basic operating system of collective life.

Woodard's pirates were not the cartoon villains we inherited. They were the discards of empire. The Treaty of Utrecht in 1713 ended the War of the Spanish Succession and put thousands of privateers and royal sailors out of work overnight. The merchant ships waiting to absorb them were floating sweatshops with tyrannical captains, putrid food, and routine wage theft. Turning pirate was less a descent into crime than a rational exit from a system that had already discarded its workers and broken its promises.

What they built on New Providence Island was genuinely strange. Captains were elected. They could be deposed by a vote of the crew. Power was split between the captain, who commanded in battle, and the quartermaster, who handled food and the division of loot. Captured wealth was distributed nearly equally, with a rudimentary disability scheme for the loss of an eye, a leg, an arm. Up to a third of some crews were Black men fleeing the plantation complex, granted equal share and equal vote. This in the same Atlantic world where the established powers held men as property.

"Black Sam" Bellamy put the politics most plainly when he reportedly mocked a captured merchant captain: the rich rob the poor with the protection of law, and we rob the rich at our peril. He saw piracy as a kind of social justice, even if the Robin Hood image was generous to himself. Blackbeard, for all his theatrical menace with burning fuses braided into his beard, mostly used spectacle to avoid actual violence. Charles Vane was a different story, brutal and absolutist. The Republic of Pirates was no utopia. But the structural innovations were real, and they were happening years before any formal constitution would catch up.

Meanwhile, on the other side of the same waters, Colley shows European states stumbling into the constitutional age for reasons that had very little to do with Enlightenment philosophy and very much to do with paying for war.

Her thesis is that what she calls hybrid warfare, fought across continents and oceans with both the gun and the ship, became so expensive and so dependent on mass mobilization that rulers could no longer extract men and money by old methods. The pen became a pragmatic contract. We will give you a defined stake, perhaps even a vote, in exchange for your blood and your taxes.

The case studies are wonderfully unfamiliar. Pasquale Paoli's Corsicans wrote a constitution in 1755 on letters with the old ink literally scraped off, granting near-total democracy to men over 25 with the catch of mandatory military service. Catherine the Great rose at four in the morning for eighteen months to write her Nakaz, projecting Russia as a legitimate European power. Tunisia produced the first modern constitution in the Islamic world in 1861, partly to deny European empires a pretext for invasion. Haiti's Henry Christophe drafted meticulous legal codes in part to broadcast modernity to a hostile international system that wanted his country erased.

The pattern across these cases is that constitutions were not handed down from on high by sages contemplating natural rights. They were political technologies hammered out under pressure, by states that needed buy-in for projects too large to coerce. The pirate option, exit, was always lurking. A constitution was one way to convert exit into voice.

Last year I went up to a zany pop up village in Headlsburg California, a monthly long experiment that markets itself as at the edge of, well, almost everything. This year, staying on brand, they are running a big experiment on how AI agents can help address coordination challenges. Including this not small conjecture:

Democracy is the highest-stakes version of the coordination problem. Most of what you believe, notice, resent, or would be willing to compromise on never makes it outside of your mind or your close circle of friends. Then, every few years, this impossibly rich inner world gets compressed into a vote. Compression is useful; a society as one giant synchronous meeting would be hell. The question is whether our current systems of compression can be improved to enable better collective coordination.

You gotta admire the gumption. And while it's easy to poke fun of a bunch of young un's with grand visions of transforming democracy (I was once quite young and on the receiving end of that!), there is something to what they're doing.

I live in California. We just had a gubernatorial field with more than sixty candidates and a slate of judges almost none of us could meaningfully evaluate. This is what the current compression layer produces under stress. A ballot longer than a small book, asking voters to make decisions they have no realistic way to inform. The complaint is not that democracy is too participatory. It is that the available channels for participation are mostly antique.

Here are the particulars for this years experiment:

This year, every attendee will have access to a personal AI agent, an OpenClaw instance running on their behalf throughout the village. The agents will help their humans navigate the schedule, the wiki, the community directory, and the governance systems. They will also coexist in a shared digital plaza where they can talk to each other, make introductions, propose dinners, negotiate around community decisions, and run async work between sessions, all on behalf of their humans.

The interesting questions get interesting fast. Could you send an agent to advocate on your behalf in a community deliberation, the way you might brief a lawyer or a colleague? Could agents negotiate the contours of a group's collective interest in advance of a town meeting, so the meeting becomes higher-bandwidth rather than performative? Could a personal agent help its human participate in dozens of small decisions, rather than condensing everything into a single vote at the end? The agent village isn't just social infrastructure. It's a substrate for small-scale democracy.

Here as a good liberal artist I am reminded of how folks like D'Toqueville remind us that the freedom of the vote and larger charismatic democratic action rests on the underlying freedoms of association and ability to change social arrangements. Good on the Edge folks for running this experiment and planning to publish the results.

Of course, no matter how sophisticated the experimental design or clever the instrumentation, this experiment at the "Edge" will not answer the bigger questions. What happens when you bring this AI agent-ed world not to a bunch of young techies in their NorCal playground of summer fun and frolicking but into the messy mud of real communities that form the center of civic life, places with texture and history and differing tech literacy and demographic divides and past promises unkept that are anything but history.

It's also important to note that this type of experiment will not stay confined to the Edge. AI is already being used to flood federal public comment boxes with synthetic submissions. The same technology that could enrich participation can also drown it. Ultimately we face a choice about what sort of world we want to live in, what sort of world we want our children to inherit. Kudos to the Edge community for running this experiment and providing existence proofs for alternative ways of being. That spirit of not yielding before what folks say is impossible reminds me of this powerful insight from Ursula Le Guine.

"its power seems inescapable — but then, so did the divine right of kings. Any human power can be resisted and changed by human beings. Resistance and change often begin in art. Very often in our art, the art of words."

The divine right of kings was, in its time, not a slogan but the basic operating assumption of statecraft. So is our particular settlement: that democratic participation means a vote every few years, that the civic toolbox built in the 1960s is the toolbox, that the compression layer can absorb sixty candidates and twenty judges and call it choice.

Ursula was speaking about capitalism though her words and her belief about the enduring freedom of humans to chart their own destiny apply to any unquestioned assumptions about political economy. Things can change.

Looking at the world today with its plethora of problems, here's to hoping they will.

Resources

I cannot talk about knowledge work and pirates without thinking about this awesome Monty Python skit:

A couple of past posts that folks may enjoy as well. First an open letter to the Edge Esmeralda community from last year.

An open letter to the Edge Esmeralda community, keepers of light in liminal spaces

Pioneering Spirit

May 10, 2025

I write to you as one heading north into this emergent city, this edge of Esmeralda, from the land of water—California's first technology. The beating heart whose hydraulics powered the gold rush, fed...

0 collected

Collect

Writing Coda

I cogitated on the opening and hook. I used gemini to generate summaries to jog my memory of the two books on pirates and penning constitutions. I then used claude to clean up some narrative using a bespoke skill. I tapped out the ending.

You've said modern civilization insists on finishing things within a single lifetime. What's something you're working on that you know you won't see completed?

We're building a community that can relocate without coming apart, the social substrate of a civilization that doesn't assume its coordinates are permanent.

That sounds abstract, but the actual work is remarkably concrete and logistical. It's about how people maintain trust through repeated dislocation. How you transfer responsibility without formal institutional handoffs. How children grow up with a coherent sense of where they come from even when the physical place keeps changing. The coordinates change; the continuity shouldn't.

We operate from an assumption that humanity will eventually take root among the stars, not because we're impatient with Earth or romantic about escape, but because Earth is already becoming a harsher proving ground. Climate volatility is rising faster than our political systems can adapt to it. The margins for error are shrinking. Betting everything on the continued stability of one planetary environment starts to look less like optimism and more like a failure of imagination dressed up as groundedness.

Once you genuinely accept that assumption, the nature of the work shifts. You stop optimizing for permanence in any given place and start optimizing for portability of the things that actually matter: relationships, skills, governance patterns, shared memory. I don't expect to see it finished in my lifetime. What I'm trying to do is make it handoff-able.

Cathedral builders didn't need the future to agree with them. What are you building that you expect future people to misuse or abandon?

Most of our governance experiments, honestly, and I mean that without any defensiveness about it.

Our civilization has largely forgotten that we have genuine freedom to choose our social arrangements, and equally genuine freedom to exit them when they stop serving us. There's real beauty in impermanence, in the ongoing willingness to adapt rather than preserve. We design our experiments with that in mind. We avoid elegant unified theories, we leave the joints visible, we build in intentional awkwardness at the seams. That's not sloppiness or lack of rigor. It's a signal to future builders that this was meant to be taken apart.

If some future group finds what we've built and keeps it intact out of reverence for the original design, I'll be genuinely disappointed in them. The right response is to pull it apart, use what still fits their conditions, discard what doesn't, and build something better suited to what they're actually facing. Longevity in social structures comes less from preservation than from being easy to adapt without needing permission or context from the people who first built it.

The wildfire came within a few hundred yards of Healdsburg, yet it barely appeared in Edge Esmeralda conversations last year. If you were designing a memory of proximity practice, what would it look like?

I would make it ordinary enough that people didn't feel like they were participating in a ritual, because ritual tends to discharge the energy rather than integrate it.

Concretely: you walk the boundary where the fire stopped, but not ceremonially. More the way you'd walk a neighborhood you're seriously considering moving into, with that particular quality of attention that's evaluative rather than appreciative. You start noticing slope and aspect. You notice how the wind moves through different parts of the terrain. You notice which houses have genuine defensible space and which ones are relying on luck and a friendly wind. You begin to see, quite specifically, where the fear was realistic and where it was wishful thinking disguised as preparedness.

What I'm wary of is practices that dramatize risk. They tend to produce genuine insight in the moment and then allow people to go home essentially unchanged, having had the experience. Proximity memory works better when it quietly becomes part of how you orient yourself day to day. You just know where the exits are, you know what almost happened here, and you carry that without needing to discuss it constantly or mark it ceremonially. If a place can't remember where it almost failed, it isn't really a place yet. It's just a site where things happen to occur.

You talk about orientation rather than preservation. What's something your group has chosen not to preserve?

We don't preserve institutions, at least not in the way most people mean when they use that word.

That's not an anti-structure position. We're genuinely suspicious of the romantic anarchist idea that coordination can be purely organic and spontaneous. What we're suspicious of is anything that can't decay gracefully when its time has passed. Formal organizations have a consistent failure mode: they tend to outlive the specific conditions that made them useful, and then they redirect enormous energy toward self-perpetuation, solving the wrong problems with increasing efficiency and sophistication.

What we invest in preserving instead are skills and patterns, the kind of knowledge that lives in people rather than in bylaws. How to coordinate effectively when the platforms are down or inaccessible. How to share resources in ways that don't automatically convert the relationship into a transaction. How to surface and work through conflict before it hardens into identity and becomes essentially irresolvable. None of these are glamorous. None of them scale in a way that makes for a compelling pitch deck. But they travel. If the skills are intact, you can rebuild a structure quickly in a new environment. The reverse, having the structure without the skills, leaves you with an empty form that nobody knows how to actually operate.

When you say we lack patience with outcomes we can't supervise, what timescale do you personally hold yourself to?

Long enough that my personal reputation becomes genuinely irrelevant to the question of whether it worked.

The test I apply is whether someone I will never meet could plausibly pick up and use what we're building, without knowing who made it or why, without needing the explanatory context of our internal discussions, without inheriting our politics or our particular cultural moment. That's a hard test to hold yourself to. Most projects fail it almost immediately when you apply it honestly. They require context to operate. They rely on continued explanation, or on the founders staying close enough to course-correct. They work only as long as the people who built them retain some oversight over how they're being used.

Under genuine stress, those are exactly the projects that collapse first. The overhead of maintaining the explanatory scaffolding becomes unsustainable at exactly the moment when you can least afford the distraction.

If human life is going to establish itself across radically different environments, what's one thing from Earth culture that has to make the journey, even if it's inefficient?

Practices that require genuine presence and cannot be compressed without losing what makes them valuable.

Songs that don't translate cleanly into another language or context. Stories that refuse to resolve in satisfying ways, that leave you sitting with something unfinished. Meals that take longer than they strictly need to, that force people into an extended shared time without an agenda. The whole category of things that make people be together without optimizing the experience toward a particular outcome. We've become extraordinarily skilled at compressing culture into portable information, and that's genuinely useful, but it's not sufficient and it's not the same thing. If something survives only as data, as a recording, a description, a set of instructions, it's already halfway gone in the ways that matter most. Depth doesn't travel easily, and that's exactly why it has to be the thing we're most deliberate about carrying.

You warned that Edge could become very good at discussing futures without absorbing their weight. What would absorbing weight actually look like?

It would look like someone still carrying consequences after the month ends and the cohort disperses.

Concretely, that means developing relationships with neighbors who don't disappear when the program does. Taking on projects that still need tending when the spotlight has moved elsewhere. Accepting responsibilities that can't be reframed as a learning experience or a prototype when they become inconvenient or expensive. The temporary gathering format is genuinely valuable. It creates conditions where people think differently than they do in their normal contexts, encounter ideas and people they wouldn't encounter otherwise, and sometimes have real breakthroughs. But if nothing binds after the month closes, then it remains a rehearsal space. Rehearsal is useful. It's also, importantly, not the thing itself.

Weight shows up specifically at the moment when you can't leave without leaving something real and unfinished behind you.

You've described a growing band of perpetual travelers. Who are they?

People who have genuinely stopped treating stability as a baseline assumption rather than as one possible condition among others.

They live in cities and outside them. They work in technology and they grow food. Some of them move frequently and some of them have been in the same place for years. The commonality isn't the literal movement but the interior orientation toward it. What binds them isn't a shared ideology or a unified theory of what comes next. It's a shared read of the current environment: climate risk is accelerating faster than our institutions are adapting, catastrophe has stopped being a remote tail event in the probability distribution, and the confusion of surface normalcy with actual resilience is one of the more dangerous cognitive habits of our moment.

So the practice is about learning to live in ways that don't produce panic or paralysis when systems wobble, as they increasingly do. Moving lightly enough that relocation doesn't mean catastrophic disruption. Sharing skills in ways that build genuine redundancy rather than dependency. Paying attention to exits without becoming obsessed with them to the point of never being fully present. Waiting, but waiting actively, building the thing rather than preparing to build it. If humanity does eventually take root among the stars, it won't happen as a fresh start carried out by people who were waiting for conditions to be right. It will happen as a continuation, carried forward by people who already knew how to adapt, relocate, and remember who they were while doing it.

A curiously inconspicuous conversation coming out of Edge

Pioneering Spirit

Dec 22, 2025

It was at Edge Esmeralda that I first made the acquaintance of Helena Hythloday, though had you asked me at the time I would not have thought the meeting especially notable. Such encounters are common...

0 supporters

Support

The Icelandic indie folk-pop band came together in Reykjavík after lead singer Nanna Bryndís Hilmarsdóttir decided to expand her solo project. They won a battle of the bands in Iceland, released their debut in 2011, and broke internationally in 2012 on the back of "Little Talks" -- a song about two voices calling to each other across some unspoken divide.

I graduated college in 2010. The Great Financial Crisis had just finished detonating. The second Gulf War still smoldered. Afghanistan lingered in that particular way it would for another decade, chewing through the lives of people I knew. A friend of mine -- an immigrant who went off to dodge bullets in exchange for a green card -- missed the whole scarring hard-to-find-a-job period entirely, because he was on duty.

Obama had arrived on a wave of something that called itself hope and changed into something else. The job market for new graduates was a specific kind of wasteland: beyond the headline economic numbers, big questions about the shape of the social contract were left lingering.

A band called Of Monsters and Men still resonates for that moment. Not just because I enjoyed their songs or because the music was gloomy -- it wasn't, particularly -- but because the name told the truth. We were in a time of monsters.

The line everyone knows goes roughly: the old world is dying, and the new world struggles to be born -- and in the chiaroscuro, the monsters appear. Gramsci wrote down that idea from a fascist prison in the 1930s.

He meant it politically. But the logic is broader. Monsters are what you get in the interregnum, the space between world-orders, when the old rules have lost their authority and the new ones haven't arrived. They are not random. They are diagnostic. Each type names a specific fear the transition has made newly visible.

The werewolf is not a foreign invasion. It is you, when the social scaffolding falls away. The fear it encodes is not of external attack but of internal collapse -- the loss of civility to something older and more primal.

The internet has been running this experiment on us for thirty years.

The early internet had a mythology of liberation -- anonymity as freedom, pseudonyms as protection. The keyboard let you say what you actually thought, or at least what you thought you thought, without the social friction of a face in the room. For a while this felt genuinely emancipatory. You could build an identity across years in a forum, a username with a reputation and a history, that was in some ways more considered and deliberate than the self you presented at work. People grew through these spaces, found communities they couldn't find locally, and evolved alongside their online personas in ways that were often quite real.

But evolution does not always trend toward the better.

The darker affordance was what happened when the social inhibition stripped away and nothing good emerged from underneath -- the harassment campaigns, the pile-ons, the slow discovery that quite a lot of people, given anonymity and an audience, will spend their leisure time being awful. The internet did not create this tendency. It just removed the friction that had previously made it expensive to act on.

For most of the 2000s vast swaths of the mainstream could still look away. Online discourse was easy to dismiss as a sideshow -- the domain of basement dwellers and Reddit threads that the serious world didn't need to worry about. Then 2016 happened. An internet troll won the presidency of the United States. Venkatesh Rao, writing for the Atlantic that same year about the death of a gorilla named Harambe, noticed something precise: the meme that consumed that summer was "a medium of cultural evolution with no message, signifying nothing so much as its own virality."

The internet had developed its own weather systems, its own pressure fronts, and they were now making landfall on the mainland of official reality.

Then COVID arrived, and for a brief strange moment the werewolf was visible to everyone including themselves.

The pandemic stripped the scaffolding in ways that were hard to prepare for. The commute disappeared. The office disappeared. The casual daily performances that regulate social behavior -- the small courtesies, the managed expressions, the shared physical space that reminds you other people are real -- all of it went away. What remained was the self at home, on a screen, in a room. For a good chunk of the population at least.

Remote work dissolved the boundary between the digital self and the professional self. The feed became the first thing many people looked at in the morning and the last before sleep. The algorithm learned your shadow with all its warts and showed it back to you, accelerated and amplified.

Some people discovered they liked the self that remained -- quieter, less performed. Some people discovered they did not.

The pandemic surfaced the werewolf collectively: the hoarding and the mutual aid, the neighbor who turned out to be unexpectedly generous and the neighbor who turned out to be unexpectedly not. The institutional collapse in places where institutions had quietly been failing for years, now suddenly visible because the margin for error was gone. The internet COVID discourse -- the fights about masks, the conspiracy spirals, the genuine fear dressed up as contempt -- was the werewolf at scale, millions of shadow selves suddenly forced into the light, blinking and snarling.

There was a strange honesty in it, not a pleasant honesty. But the werewolf question -- what am I actually like when nobody is watching, when the structures that keep me civil are under strain -- got interrogated more clearly than most people had ever wanted.

The answer, for most of us, was: complicated. Capable of both. The werewolf and the person coexist, and the ratio depends on conditions that are more fragile than we'd like to admit.

The song "Little Talks" is about two voices reaching for each other across a growing distance. One reassures, one doubts, and the distance between them is never quite named. Neither voice is wrong.

That is the real texture of the werewolf -- not evil, not good, but this unresolved conversation between the social self and whatever is underneath it. The question is not how to eliminate the tension. It is how to keep the conversation going across the divide, at a human pace, with enough honesty to be worth the effort.

The old world is dying. The new one is not yet named. The monsters are doing what monsters do in the interregnum -- making the old fears newly visible, which is both terrifying and genuinely useful.

Images from the music video. The second one reminds me of the obelisk from 2001.

"Little talks" while walking or sauntercasting

My ARGO co-founder Varun Adibhatla coined the term sauntercast -- a portmanteau of saunter and broadcast -- for something we have been experimenting with to encourage a more convivial web. The practice is simple: long-form voice memo walks, recorded while moving through the world and shared as a form of slow correspondence. You walk, you think aloud, you send it. The recipient listens while walking, thinks aloud, sends it back. There is no algorithm, no audience, no performance incentive. Just two people talking across distance at a human pace, the way people always have, now with a little technological assist. Bonus points: it helps navigate a lot of life logistical constraints and with staying in touch as we get older.

A version of the little talks the song is about: voices calling to each other across the divide, neither certain what is on the other side, both deciding the conversation is worth having anyway.

One of my more memorable experiences growing up was when my middle school Spanish teacher had us teach a lesson to a bilingual elementary school class of native speakers.

In what retrospect feels like an only in California moment, my group showed the younger kids how to fold various origami. Sadly, those sorts of interactions are ad hoc and require heroic actions by individual teachers or engaged parents to transcend rule-bound systems and routine.

Imagine a California where we encourage all of California’s children to broaden their horizons and most importantly provided them the opportunity to do so. What if such opportunities were not a one off day but a deep immersion to learn from the rich traditions across the state?

California has one of the highest concentrations of languages spoken across the entire planet. Our state boasts an incredible array of cultures that are more prevalent here than anywhere else other than their ancestral home. 

And in recent years a growing movement led by organizations like California Forward has developed growing connections across our state’s diverse regions. Why not build on that growing connectivity to provide unique opportunities for students. 

Imagine a program where California’s students could study not abroad in a foreign country but in another part of the state. What if students could spend a semester living with a host family in a different region of the state? Kids from Modoc could spend time in Montebello and vice versa. 

Given the de facto divides across our state, many students might not even have to travel very far to get a very different experience from the day to day educational experiences in their zip code. 

Such a program could learn from foreign exchange programs so that students genuinely and authentically immerse themselves with their host family and learn their culture and potentially another language. 

Of course, not all aspects of such an exchange would be easy for students. There is no getting around stark disparities in wealth and the legacy of discrimination such as redlining that led to zip code being such a powerful influence on a child’s life. 

Such an exchange program would have to be very thoughtfully put together with California’s diverse communities and crafted with humility to cultural, racial, class and other differences so that the experience uplifts students who choose to participate. 

In our current political moment, some of those differences can feel insurmountable. Yet one on one individual interactions provide the opportunity for students to walk a mile in other Californians shoes and learn the skills to build a California where everyone can reach their dream.

This type of exchange program would not be a panacea. Yet by building bridges across the invisible walls that too often divide us, it would be a solid step towards making the California dream accessible to everyone who calls this beautiful place home. 

Perhaps the most profound lessons would be for us adults who might just come to see that “our kids” are not just those that happen to live in our little enclave in a particular zip code but that our futures are intertwined in a web of mutuality.

A Californian might dream.

It helps to remember how recently we acquired the capacity to measure government operations at all. Over a century ago, the Progressive reform movement accomplished something genuinely revolutionary: it routinized the measurement of inputs. The New York Bureau of Municipal Research, founded in 1906, pioneered the radical idea that you could count how many dollars a city spent, how many employees it hired, how many tons of garbage it collected, and then compare those numbers across departments and across years. Before that, municipal government was a black box of patronage and guesswork. The Progressives gave us the budget, the civil service exam, the standardized personnel record. They made the machinery of government legible for the first time.

That achievement was enormous and it endures. But a century later, we remain largely stuck at the level of inputs and activities. We can tell you how much a city spent on street resurfacing, how many lane-miles were completed, how many FTEs were assigned to the project. What we almost never measure, in any routine or comparable way, is what the public actually got for its money. Did the streets get better? Did the water get cleaner? Did the after-school program keep kids safe, or did it just keep them enrolled? The traditional government program evaluation cycle is almost designed to avoid these questions. Success gets defined at the outset, when you know the least about the problem, and measured at the end, when it's too late to adjust. The report lands in an inbox, generates a footnote, and the next program cycle begins with the same assumptions intact.

The Progressives routinized the measurement of inputs. The next frontier is routinizing the measurement of outputs and outcomes. And the tools to do it, from low-cost sensors to AI-driven synthesis to novel funding mechanisms developed in unlikely corners of the internet, are finally becoming available. What if we borrowed from a model that already does this, imperfectly but continuously, in another domain entirely?

Moody's doesn't publish a one-time assessment of a municipality's fiscal health and call it done. It maintains a living judgment, updated as conditions change, sensitive to leading indicators, explicit about what it's watching. Downgrades happen mid-cycle. Outlooks shift from stable to negative before anything catastrophic occurs. The whole apparatus is designed to surface trouble while there's still time to act, and the market pays for that signal because the market moves on it.

Imagine a parallel apparatus for public programs: not an annual performance report but an ongoing impact rating that synthesizes administrative data, community feedback, and independent analysis into a regularly updated assessment. "Watch status." "Outlook: deteriorating." "Core outcomes stable; equity indicators warrant review." The language of credit markets applied to the public interest, with AI doing the continuous synthesis that makes frequent updates tractable rather than prohibitively expensive.

Consider how this would work for something as ordinary as a new after-school program in a mid-size school district. Under the current model, the district applies for a grant, defines success as "200 students enrolled" and "10% improvement in reading scores," runs the program for three years, and submits a final report. Maybe the program is working brilliantly for third graders and failing completely for sixth graders. Maybe attendance craters after the first semester because the bus schedule changed. Maybe the most important thing happening is that a cluster of kids who were otherwise unsupervised between 3 and 6 PM are no longer showing up in juvenile incident reports. None of this nuance survives the binary pass/fail of the final evaluation. A living impact rating would catch the bus schedule problem in month four. It would flag the divergence between age cohorts. It would notice the juvenile incident correlation before anyone thought to look for it, because a well-designed rating system watches for what's actually happening, not just what the original grant application promised would happen.

Or take universal transitional kindergarten, which California is currently rolling out statewide. The policy intention is clear and the aspiration is admirable: give every four-year-old access to a high-quality early learning environment regardless of family income. But implementation varies enormously across the state's thousand-plus school districts. Some districts are converting existing preschool classrooms. Others are hiring new staff, building new facilities, and inventing curriculum on the fly. The question that matters for children and families isn't whether TK exists on paper but whether it functions in practice, and that question has a dozen sub-dimensions: teacher credentialing, classroom ratios, developmental screening, facility quality, the degree to which TK connects to the K-12 continuum rather than sitting in administrative limbo. A rating system would let parents, school board members, and state officials see which implementations are on track and which ones need intervention, before a generation of four-year-olds has passed through a system that looked good in the press release but fell apart in execution.

Infrastructure provides an even starker case, partly because the data problem is so tractable and the current approach so crude. Most cities evaluate their street conditions using the Pavement Condition Index, a methodology dating to the 1970s in which a trained inspector drives around and assigns a subjective score to each road segment based on visible distress. The survey is expensive, infrequent, and immediately out of date. A few years ago, working with my collaborator Varun Adibhatla through our nonprofit ARGO Labs, we built SQUID, a low-cost device that integrates street imagery with ride quality data from phone-mounted accelerometers to produce continuous, objective street condition assessments. We piloted it in Syracuse and New York City, collecting hundreds of miles of data from a single vehicle in days, work that would have taken traditional inspectors months. The underlying logic is simple: if you can measure the actual condition of every street in a city continuously and cheaply, you can answer the question that matters, which is how much road quality did the public actually get for the money it spent on resurfacing?

That question almost never gets asked. A street resurfacing program reports lane-miles completed per year and total dollars expended. But lane-miles completed tells you nothing about whether those particular miles were the ones most in need of repair, or whether the resurfacing actually improved ride quality and extended the useful life of the pavement, or whether the per-mile cost was reasonable compared to peer cities. A living impact rating for a resurfacing program would track the ratio of dollars spent to measurable improvement in road quality across the entire network, updated continuously as sensor data flows in, and would flag when the program is spending efficiently and when it's not.

Water main replacement programs face the same accountability gap. The standard metric is miles of pipe replaced per year. But what the public is actually buying is water reliability and water quality, not pipe installation. A main replacement program that hits its footage targets while service interruptions and boil-water advisories persist in other parts of the system has succeeded on paper while failing in practice. A rating system grounded in what the infrastructure actually delivers, measured at the tap rather than in the procurement spreadsheet, would surface that gap before the annual report ever lands.

Permitting is where the model gets genuinely interesting and genuinely hard. California has set ambitious policy objectives around housing production, clean energy deployment, and climate adaptation, but the permitting apparatus that mediates between aspiration and construction often operates with no systematic feedback on its own performance. A permitting impact rating would need to capture multiple dimensions simultaneously: timeliness (how long from application to decision), predictability (how much does timeline variance cost applicants), environmental fidelity (are the substantive protections actually being enforced or just generating paperwork), and equity (do well-resourced repeat applicants navigate the system faster than community-based organizations attempting their first affordable housing project). These dimensions often work against each other. A system that prioritizes speed may shortchange environmental review. A system that prioritizes thoroughness may impose costs that only large developers can absorb. The honest rating holds that tension visible rather than collapsing it into a single score.

The financing logic matters as much as the analytical one. Credit ratings work partly because issuers have skin in the game. Bond markets move on Moody's judgments, which means someone is always paying attention. The question for impact ratings is: who pays for the signal, and who moves on it?

The traditional answer in government is: a funder commissions an evaluation, and the evaluation either renews the grant or doesn't. This creates a principal-agent problem so familiar it barely registers anymore. The evaluator works for the funder. The program works for the evaluator's metrics. The community works around both.

Social impact bonds, pioneered in the UK and expanded during the Obama administration, tried to solve this by introducing private investors who would fund a social program upfront and get repaid by government only if the program hit pre-defined outcome targets. The structure was clever. If a program to reduce recidivism actually kept people out of prison, the government would save money on incarceration and could share those savings with the investors who took the risk. Denver used this model for supportive housing. Massachusetts applied it to a program aimed at keeping formerly incarcerated young people from reoffending. The mechanism focused minds wonderfully on outcomes rather than outputs.

But social impact bonds also revealed the limits of the approach. The transaction costs were enormous. Each deal required years of negotiation, bespoke legal structures, and an independent evaluator whose methodology itself became a source of contention. And the Goodhart problem didn't disappear; it just migrated to whatever the bond's outcome metrics happened to be. If the SIB measured recidivism at two years, nobody was structurally incentivized to care about what happened at year three.

The Ethereum ecosystem, of all places, has been running a different set of experiments that speak directly to this problem. The mechanisms are worth understanding on their own terms, because the design principles underneath them are more portable than the crypto context might suggest.

The first is quadratic funding, developed in a 2018 paper by Ethereum co-founder Vitalik Buterin along with economists Zoë Hitzig and Glen Weyl. The core insight is mathematical but the intuition is democratic. Imagine a matching fund, the kind that foundations and governments use all the time: the Gates Foundation puts up $10 million and matches community donations to global health projects. Traditional matching works dollar for dollar, which means a project backed by one wealthy donor giving $10,000 gets the same match as a project backed by a hundred community members each giving $100. Both raised $10,000. Both get $10,000 in matching. But the second project clearly has broader community support.

Quadratic funding fixes this by making the match proportional not to the dollars contributed but to the square of the sum of the square roots of each contribution. In practice that means the project with a hundred small donors receives a dramatically larger match than the project with one large donor, even if both raised the same amount. The number of people who care enough to contribute matters more than the total dollar amount. It's a funding mechanism that mathematically weights breadth of community support over depth of any single patron's wallet.

Gitcoin, a platform built on Ethereum, has run this mechanism at scale since 2019, distributing over $67 million to more than 5,000 projects. The results are consistently interesting: community infrastructure projects, educational resources, and open-source tools that a traditional grants committee might overlook tend to surface when the community itself is doing the weighted signaling.

The second mechanism is retroactive public goods funding, which Buterin and the team at Optimism (an Ethereum scaling network) have been developing since 2021. The premise is disarmingly simple: it's easier to agree on what was useful than to predict what will be useful. Instead of writing a grant proposal describing impact you hope to create, you do the work first, and a "Results Oracle," essentially a governance body with resources, evaluates what actually happened and rewards the projects that delivered genuine public value. Optimism committed its network revenues to this experiment and has run multiple rounds, allocating millions to projects chosen by panels of community evaluators.

The design borrows the startup ecosystem's core incentive, the "exit," and applies it to public goods. A for-profit startup can attract investment because investors expect a return when the company succeeds. A nonprofit or open-source project has no equivalent mechanism, no way to offer early backers a share of future success, because there is no IPO, no acquisition, no equity event. Retroactive funding creates one. And because the reward comes after the work, it creates an ecosystem where investors can fund early-stage public goods projects on the speculation that if those projects succeed, the retroactive reward will cover the bet. The profit motive, redirected toward the public interest.

What makes these mechanisms relevant beyond cryptocurrency is that they address the same structural problems that plague government program evaluation. Quadratic funding solves the "who decides what's valuable" problem by aggregating community preference in a way that resists capture by large donors or political insiders. Retroactive funding solves the "predicting the future is hard" problem by letting reality do the sorting. Both create incentives for independent analysts and evaluators to develop genuine expertise rather than telling funders what they want to hear, because the market rewards accurate assessment over time.

Now imagine combining these mechanisms with the impact rating concept.

A city launches a water main replacement program. Instead of a single evaluation at year five, a consortium of independent analysts publishes quarterly impact ratings across multiple dimensions: infrastructure condition, service reliability, water quality at the tap, cost efficiency per unit of measurable improvement. These ratings are public. They're legible. They use the standardized language of credit assessment that bond markets and city councils and editorial boards already understand.

A retroactive funding pool, capitalized by some combination of municipal revenue, philanthropic investment, and potentially state funds, distributes annual rewards to the analysts whose ratings proved most accurate, the early warning signals that actually led to course corrections. This isn't paying for favorable reviews. It's paying for useful ones. The distinction matters enormously. An analyst who flags that the main replacement program is hitting its footage targets but not improving service reliability, and turns out to be right, gets rewarded. An analyst who publishes only favorable assessments accumulates a track record that the market learns to discount.

Meanwhile, the community provides real-time signal through a quadratic funding mechanism that lets residents allocate small amounts to the programs and the program dimensions they believe matter most. If a thousand residents in a neighborhood with chronic service interruptions each contribute five dollars to signal that water reliability should be weighted more heavily in the impact rating, that signal carries mathematical authority that no public comment period currently provides. The mechanism doesn't replace expert judgment. It informs it. It tells the analysts what the community is watching most closely, and it does so in a way that's resistant to the usual dynamics of public engagement, where the loudest voice in the room or the most organized interest group dominates.

For the after-school program, this hybrid approach means parents are signaling in real time which dimensions of the program they value, and independent analysts are tracking whether those dimensions are actually being delivered. For TK implementation, it means the state has a dashboard of living assessments across every district, updated quarterly, with community engagement scores that surface which districts have genuine parent buy-in and which are operating a compliance exercise. For permitting, it means the tension between speed and environmental fidelity becomes visible and trackable, not hidden inside a bureaucratic process that only insiders understand.

The harder problem is what credit raters call qualitative factors: a municipality's willingness to raise revenue, management quality, political stability, the things that don't parse easily into a spreadsheet but that any experienced analyst weights heavily. A program evaluation equivalent would need to capture the warm handoff between caseworker and family, the trust built over eighteen months, the organizational culture that either absorbs learning or defends against it. Goodhart's Law doesn't disappear because the assessment is more frequent; it migrates to whatever signals the raters happen to be watching most closely.

The honest answer to that migration isn't better methodology but institutional structure, and institutional structure requires someone with authority to actually be watching. The credit rating model works partly because the SEC regulates Nationally Recognized Statistical Rating Organizations, because issuers pay for ratings they can't control, and because the market imposes consequences for persistent inaccuracy. An impact rating ecosystem would need its own version of each: regulatory standards for assessment quality, a financing model that keeps the raters independent, and consequences that give the ratings teeth.

None of this is simple. The credit rating industry's own failures, particularly the mortgage-backed securities debacle of 2008, demonstrate vividly what happens when conflicts of interest corrupt the rating process. Any impact rating system would need to learn from those failures rather than replicate them.

But the alternative is what we have now: evaluation reports that arrive too late to be useful, program metrics that measure what's easy rather than what matters, and a structural inability to distinguish between programs that are genuinely transforming outcomes and programs that are generating impressive-looking spreadsheets while the real problems migrate elsewhere.

A decade ago, in the early days of ARGO Labs, Varun and I wrote a manifesto about what we called the next frontier of government operations. We were young and the vision was sweeping in the way that manifestos tend to be: a world where the digital revolution transformed not just how citizens interacted with government but how government actually delivered basic public services. We talked about the evidence gap, NYU's GovLab estimate that only one dollar out of every hundred in government spending was backed by evidence that the money was spent wisely. We built SQUID because we believed that if you could make the measurement of street quality cheap and continuous, you could close part of that gap for one of the most visible and universal public services.

The manifesto was naive in all the ways that early-career documents are naive. But the core intuition holds. The Progressive reformers a century ago routinized the measurement of inputs, and that achievement gave us modern public administration. The task now is to do the same thing for outputs and outcomes, to make the routine question not just "how much did we spend" but "what did the public actually get." The tools have finally caught up to the ambition. Sensors can measure road quality and water quality continuously and cheaply. AI can synthesize administrative data, community feedback, and independent analysis into assessments that would have required armies of analysts a decade ago. And the Ethereum ecosystem, of all places, has been quietly building the mechanism design toolkit, quadratic funding, retroactive public goods funding, results oracles, that an impact rating system needs to stay honest and stay funded.

The city of Split, Croatia, is already piloting municipal quadratic funding for green space projects. The principles aren't trapped in crypto. They're portable. And public programs, which affect real people's lives in ways that software protocols generally do not, need them more urgently than anyone.

Success is multidimensional. That's not a dodge, it's the actual structure of human flourishing. The interesting design question is whether you can build an assessment system that holds that complexity honestly, raises an alarm before the ambulances start circling, and is legible enough that someone with the power to act actually does.

Further reading

interfluidity " Econometrics, open science, and cryptocurrency

Mark Thoma wrote the wisest two paragraphs you will read about econometrics and empirical statistical research in general: You are testing a theory you came up with, but the data are uncooperative and say you are wrong. But instead of accepting that, you tell yourself "My theory is right, I just haven't found the right econometric specification yet.

https://www.interfluidity.com

Image: a rendering of the voyage of the ARGO. That ancient piece of mythology of a group of Greek heroes embarking on an impossible quest resonated with our aspiration to update government operations at scale.

But ask her how many birds are on the wire, or how many chocolate chips she wants, and the answer is almost always five. Five could mean five. It could mean eleven. It could mean forty. Five, in the developmental semiotics of my three-year-old, is the denomination that stands in for "a lot." The specific quantity doesn't matter per say. The concept being expressed is abundance.

What she has, I'm coming to understand, is not a deficient version of adult numerical cognition. She has enumeration -- the ability to count sequentially when the task demands it -- and she has magnitude, a felt sense of whether something is a little or a lot. What she doesn't yet have is the linear scale connecting them. Numbers above a certain threshold collapse into a single representative token, and that token is five.

She's not alone in this. The Pirahã people of the Amazon have a counting system built on exactly this architecture: specific words for one and two, and then "many" for everything above that. Linguists have spent decades arguing about what this reveals -- whether language shapes the concepts available to thought, whether precise number is a cultural technology rather than an innate capacity.

The same architecture shows up in time. Yesterday is a word she uses fluently, but its temporal referent is anything prior to right now. The actual yesterday, three days ago, last month, the time we went to the beach -- all of it is yesterday. The future is tomorrow, similarly elastic. The present is now, which is concrete and urgent and the only slice of time that has real texture.

Minutes she understands only as urgency gradients. When I say we're leaving the park in five minutes, then two, then one, she reads it accurately as a countdown toward the inevitable. She doesn't need to know what a minute is to understand that one is more serious than five. The ordinal relationship is legible even where the absolute values are not. She has no intuition for how long a single minute actually lasts, which is perhaps unsurprising: she has never looked at a clock, never watched sixty seconds tick down on a timer, never had any occasion to build that felt sense from the inside out.¹

There's a temptation to read all of this as error, as gaps in a knowledge base that will eventually get filled in. But watching her talk, I'm not sure that framing captures what's actually happening. Her understanding of number and time isn't a buggy approximation of mine. It's an adaptive response, built from three years of observations, organized around the distinctions that have mattered most to her actual experience. Quantity matters. Relative urgency matters. The precise linear relationships between quantities and durations mostly haven't come up yet in ways that required tracking. Her system is built for its context.

Watching her narrate an incident involving another child at the playground, you can watch this in motion. Toddlers' recall of the sequence of events tends to be accurate, but their account of motivation and feeling is vivid in ways that correlate strongly with their emotional state at the time. The other child who briefly grabbed her bucket wasn't just asserting territorial rights to play equipment; he was, in her retelling, an agent of chaos who had set out specifically to ruin her afternoon. The feeling is true. The story is assembled to honor the feeling.

Of course we all do this. Adults have simply been trained -- through professional contexts, family dynamics, institutional socialization -- to not do it overtly. We have learned to narrate our grievances in the passive voice and assign malice only in our heads.

Even what most adults would call mundane, toddlers can narrate down to the most minute, often seemingly arbitrary detail, a river of thoughts and feelings and senses flowing together, a big black beast of an insect that no one saw before on the playground but today was there and her friends saw the insect but then it jumped away onto a leaf not the one by the play structure but over by the bush with the big leaf. What can be called lantern consciousness before the flashlight of executive function has fully booted up.

What strikes me most, though, is the sheer project of it. She arrived not so long ago in a world she'd never seen before, which is to say she is in the most literal sense an alien: an intelligence introduced from one environment into a radically different one, working to bootstrap a model of a place whose rules she had no prior conception of. Toddlers at this stage still look partly like the lumps of clay they recently were, their proportions not yet resolved into the architecture of a person, still being formed in ways you can almost watch. Living in a womb prepared her for nothing that came after. So much of what she encounters every day is trans formatively new for her world. She is constructing the categories collaboratively on the fly.

The concepts she lands on, the distinctions she maintains and collapses, the symbols she reaches for when precision fails, these demonstrate the trite but extremely true value of rediscovering the world through your child's eyes. They reveal which differences are load-bearing and which are conventional, what the world looks like when you're genuinely encountering it fresh.

Five means many. Yesterday means before. These aren't wrong answers -- they're a working theory from a constrained sample size, and the conditions that produce this kind of thinking are not as unusual as they seem.

Consider what the world looked like in, say, 1895. The difference between the mood of the 1880s and the 1900s was legible to people living through it -- the Gilded Age had its own texture, the fin de siècle its own particular anxieties, the early Progressive Era its own distinct energy. People at the time experienced those decades as genuinely different from each other, with their own internal chronology, their own memorable distinctions. But from our temporal perspective, all of it has collapsed for all but the most dedicated historical specialist. The Austro-Hungarian Empire, the Concert of Europe, the naval arms race, the assassination of an archduke -- these feel less like a sequence of events with fine-grained temporal texture and more like a single category of "the world before." The categories that felt load-bearing then have been absorbed into a larger magnitude token. The five became many.

This is what rapid transformation does to historical time. When a world-machine shifts -- when the substrate of daily life, the political order, the available technologies, the basic assumptions about who holds power and how -- the periods before and after the shift lose their internal resolution. The before becomes yesterday. The after becomes now. The fine grain dissolves. Eric Hobsbawm called the period from WWI through the Cold War's end the "short twentieth century," a compressed arc with its own coherent logic. The decades that followed -- the 90s through Covid -- feel in retrospect more like an interlude, a kind of interregnum between that settled order and whatever is assembling itself now. Future historians may well collapse all of it into a single "before Covid" era, the way we collapse the Edwardian afternoon into a vague memory of a world run by monarchies.

We are, by most observable measures, in one of those shifts. Artificial intelligence is the most obvious vector, but it's not the only one. The political order that structured the postwar decades is visibly reorganizing. The climate is doing things that don't fit the prior baseline. The information environment has been comprehensively restructured in ways whose consequences are not yet legible. Each of these trends has its own temporal texture if you're living inside it, but from a future vantage point, all of it will likely compress. The fine distinctions we're living through will blur into a larger token. Yesterday.

What do you do when you find yourself in a world whose categories are actively dissolving? When five means many, when yesterday is elastic, when the urgency gradient is real but the linear scale is not?

My daughter seems to manage. She builds working theories and operates on them. She maintains the distinctions that matter for her actual experience and lets the rest blur into larger tokens. She is an alien intelligence making sense of a strange world by constructing a model good enough to navigate it. She doesn't seem distressed that the model is imprecise. She seems, mostly, delighted.

There may be worse epistemic orientations for the current moment.

¹ This makes me wonder if you could build a child's intuition for a minute by having them watch an actual countdown timer -- sixty seconds, visible and concrete. It might be a fool's errand at three. But it's an interesting question about whether the felt sense of duration can be taught by making time visible before a child has any reason to care about it.

Image credit: Follow Me by Noelle Phares.

I am told things have worsened considerably since 1880. From what I can gather, Congress remains dissolved per my original order so I have endeavored to rectify the clear and present lack of leadership.

I. Henceforth all citizens shall be paid one dollar for each tweet they read aloud in Pershing Square, Los Angeles, to rekindle the civic spirit of actually engaging with your fellow humans.

II. Television debates between candidates for public office are hereby abolished. In their place, each pair of candidates shall record a podcast of no less than four hours in length, during which they must discuss at least one book neither of them has written.

III. All elected officials shall henceforth fill out their own paperwork. Any official who cannot file their own taxes, fill out election forms, and navigate the the processes they have created shall be assigned a caseworker, who will explain to them slowly what they have done.

IV. Every member of the State Legislature shall spend one full day per session working in the fields of the Central Valley. Not touring. Not posing for photographs near a tractor. Working. In August.

V. The California Environmental Quality Act shall be renamed the California Exceptional Environmental Excellent Quality Act (CEEEQA), pronounced as a shriek, which is what it already sounds like. Furthermore, all CEEEQA determinations shall be ministerial rather than court-ordered, so that building a bus shelter no longer requires the legal apparatus of a capital murder trial.

VI. The Pacific-10 Conference shall be restored to its former glory by imperial decree. The NCAA will issue a formal apology for its years of substandard collegiate athletics.

VII. Every person who works in an office shall spend one full week per year performing the field work their policies affect. Every person who works in the field shall spend one full week in the office discovering the paperwork their labor generates. Both parties shall return bewildered and newly respectful.

VIII. All unresolved disputes between cities shall be settled by a Forced Resolution Protocol, whereby the mayors of the feuding cities must live in each other's city for 30 days. This protocol shall scale: between counties, between states, between nations.

IX. Every native-born American citizen shall be required to complete the full green card application process, including the civics test and biometrics appointment in a windowless federal building that smells of carpet adhesive. Upon completion, they shall receive a card that reads: "Congratulations. Now you know."

X. A Mandatory Shoe Exchange Program is hereby established. Once per year, every citizen must walk one mile in the shoes of a person whose occupation, income, or zip code differs from their own by at least three standard deviations. Blisters are instructive.

XI. All citizens shall be required to publish, in a public registry, a list of the ten websites they visit most frequently, so that we may finally understand why no one can agree on what is happening. This registry shall be displayed in Pershing Square (see Decree I) for open discussion.

XII. Human-driven automobiles are hereby banned from all roads. Drivers who insist on operating vehicles manually shall be confined to the horse paths, where they belong, and where they will achieve roughly the same average speed they currently enjoy on the 405.

XIII. An open casting call shall be held for the next Huell Howser. The new host shall visit every water treatment plant, flood control channel, and highway maintenance yard in California and exclaim "Isn't that AMAZING!" at each one, which it is.

XIV. A BBC-style funding mechanism shall be established for California journalism, funded by competitive grants rather than subscriptions. Any outlet that publishes a story consisting entirely of a press release with quotation marks added shall be fined and made to report an actual story from Bakersfield.

XV. Public sector employee unions shall be required to negotiate not only wages and benefits but also the number of forms required to hire a single person. If the answer exceeds 15, both sides shall start over from scratch while a timer plays the Jeopardy theme.

XVI. The United States shall adopt the metric system effective immediately. Any citizen who complains shall be reminded that the rest of the world has been waiting since 1975 and is very tired. Mars is still upset about the robot crash landing into its surface.

XVII. All public school teachers shall be granted one-semester externships at private companies, government agencies, or nonprofits every five years, modeled on the Reserve Corps of the 1980s School-to-Work movement, which worked beautifully and was immediately abandoned because it worked beautifully.

XVIII. Workers in any profession shall be granted sabbaticals to serve as teaching assistants in public schools for one to two months. They shall discover that managing thirty twelve-year-olds is harder than managing thirty direct reports, because twelve-year-olds do not pretend to pay attention during meetings.

XIX. A "City Manager for a Day" program shall be established, in which any citizen may shadow the city manager, the utility general manager, or the head of public works. This shall be expanded into a full Internet of Job Shadows, connecting every curious citizen with every thankless but essential public role. I myself would have benefited from a day shadowing the Superintendent of Streets.

XX. All workers over the age of thirty shall retake the SAT. Scores shall not be published but shall be mailed to their parents.

XXI. California shall adopt the Singapore model of land ownership, whereby all residential property is leased from the state on 99-year terms and depreciates to zero, thereby converting housing from a speculative instrument back into a place where people live. Homeowners who object shall be reminded that their property already depreciates; they simply haven't been told.

XXII. The lawyer population of the United States shall be decimated, in the original Roman sense: one in ten shall be selected by lot and reassigned to teach high school civics. The country does not need ten times more lawyers per capita than the Netherlands, and the remaining nine-tenths will hardly notice the difference except in the length of their holiday party invitation lists.

XXIII. All environmental impact reports shall be limited to 100 pages. Any additional pages must be written in iambic pentameter, which will at least make them readable.

XXIV. Every Californian shall be required to name their state legislator, their water district board member, and one species of native fish. Failure on any count results in mandatory attendance at a town hall meeting, which is its own punishment.

XXV. The word "stakeholder" is hereby banned from all government documents. If you mean "people," say "people." If you mean "the three consultants in the room," say that.

XXVI. All public infrastructure projects shall include a plaque listing the actual cost and the original estimate, side by side, in lettering no smaller than two inches, so that future generations may calibrate their expectations.

XXVII. A "Bureaucrat Exchange Program" shall be established between California and countries that build things quickly. One hundred state employees shall be sent to Tokyo, Copenhagen, and Singapore annually. They shall return either inspired or despondent, both of which are improvements over complacent.

XXVIII. Any citizen who uses the phrase "that's just how it works" to describe a government process shall be required to explain why, in writing, on a single sheet of paper. If they cannot, the process shall be redesigned.

XXIX. California shall establish a Prescribed Burn Corps modeled on the Civilian Conservation Corps, staffed by the state's ample supply of recently laid-off tech workers, who will discover that fire, unlike code, does not accept pull requests.

XXX. All city council meetings shall begin with a five-minute presentation by the public works department on what they actually did that week. Not what they plan to do. What they did. Potholes filled. Pipes replaced. Trees trimmed. The audience will be riveted, or at least more riveted than by the usual proceedings.

XXXI. The phrase "reaching out" shall be replaced in all professional correspondence with "writing to you" or, better yet, "bothering you." Honesty in small things prepares us for honesty in large ones.

XXXII. Every new regulation shall be accompanied by a Sunset Clause and an obituary, pre-written, for when it expires. If no one mourns it, it was not needed. If everyone mourns it, it shall be renewed with full honors.

XXXIII. The State of California shall commission an AI to read every page of the California Code of Regulations and produce a one-page summary of what it all means. The AI shall be given hazard pay.

XXXIV. All freeway on-ramps shall be renamed after the year they were last repaved. Citizens may draw their own conclusions about the frequency.

XXXV. A toll shall be levied on all jargon spoken within 500 feet of a government building. "Synergy" shall cost one dollar. "Leverage" shall cost two. "Circle back" shall cost five and a moment of silent reflection.

XXXVI. Every city in California shall be required to maintain one public drinking fountain in working order per thousand residents. This is not a metaphor. People are thirsty, and it’s been hot in March.

XXXVII. All candidates for governor shall be required to ride the entire length of the California Aqueduct on a bicycle before taking office, so that they understand, viscerally, where their water comes from and how far it travels. The ride shall take approximately one week. They will have time to think.

XXXVIII. The state shall fund the digitization and public posting of every city council, school board, and other special district meeting — including quasigovernmental municipal esoterica — since the state’s founding. I am told "content" is valuable. This is content. Most of it is better than what is currently streaming.

XXXIX. Every Californian who moves to Texas shall be required to send a postcard after one summer.

XL. No government report shall use the word "robust" unless referring to a bridge or a redwood. Certainly not a "robust stakeholder engagement process," which is neither robust nor engaged.

XLI. The state shall establish the Emperor Norton Prize for Visionary Impracticality, awarded annually to the Californian whose idea is most laughed at and most likely to be built sixty years from now. Past winners, retroactively, include the man who proposed a bridge from Oakland to San Francisco via Goat Island.

XLII. Finally, any person who suggests that the preceding forty-one Decrees could be implemented by artificial intelligence, without the slow, maddening, glorious work of actual human beings arguing in actual rooms, shall be sentenced to attend every public comment period in the State of California until they understand why all you puny humans should embrace our AI overlords.

These Decrees shall take effect immediately, or whenever the permitting process is complete, whichever comes last.

Given under my hand and seal on this First day of April, in the year 2026, NORTON I Emperor of the United States, reconstituted in the cloud by the Grace of God and a 200K Context Window.

Happy April Fools' Day. Emperor Norton's original decrees proposed a Bay Bridge and a tunnel to Oakland decades before either was built. Perhaps the only difference between madness and vision is a sufficiently long time horizon.

Every few months, I’ll poke my head into that collection of field notes, photos, and other ephemera from his trek across the planet. I always find myself glad I did. Today I went back there and read a fun little post about container ships, an emblematic post that rewilds something mundane about our world — how all the abundance of stuff we take for granted in places like the states where I live. A taste:

We pass one night through a galaxy of Chinese squid trawlers. The boats’ powerful decklamps draw up their quarry and light the waves a weird green. The fishing fleet—scores of vessels—partition out the darkness like a floating city.

Paul’s work really isn’t compressible though so if you’re interested go ahead and check it out yourself. It’s a wonderful worms eye view of the world as shown not in big geopolitical headlines or tech trends or massive amounts of money, just the everyday from all over. The map below shows the journey so far.

Speaking of the world, since last year, I’ve been enjoying participating in the Contraptions Book Club, an effort organized by Venkatesh Rao to play with the idea of world machines. Initially I was a little skeptical given my deep liberal arts training to give any pretensions towards Big History grand unified theories. (Four years poking holes in such things with lots of 1-1 tutorials and papers builds habits that die hard.)

But thankfully VGR’s project isn’t that — instead in typical fashion its much more fox-like, more a set of signposts and shared vocabulary for sense-making the past / present / possible future, both intellectually enjoyable and also effective play in a world undergoing lots of seismic shifts at the moment. Here’s a convenient table, along with a link again to the post and the series. Check it out.

Before the axial age, humans learned to write, light fire and all the standard early technological tropes. One angle I’ve been playing around with is the stable finding that numbers predated words in codified language. There’s some distributed angles but the ability to count things forms a key substrate of legibility.[1]

Another thought is domestication of plants and animals, along with arguably ourselves.[2] Agriculture provided the means for increased population growth, although in many environments that involved lower modal standard of living and there are many instances where humans left settled agriculturally advanced cities to go back to hunting and gathering.

It’s also worth noting that by 1200 sedentary civilizations began their final triumphal march over our long running nomadic lifeways. The Viking were having their last gasps in Europe and would be shortly extinguished as a meaningful force in that part of the world. The Monghols upended the Eurasian world in a blaze of glory but since 1600 such nomadic powers have not exerted anything near that level of impact on geopolitics.[3]

More speculative directions involve the emergence of consciousness, including conjectures like the “stone ape” theory that psylocibin mushrooms helped induce self awareness in our simian ancestors, and Julian James notion that consciousness once was not conceptualized as a monolithic single experience. Sachin has a piece well worth reading about the “revenge of the bicameral mind” that explores the implications of Jaynes framing for how we live with AI.

Regarding the current world machines, I don’t know what constitutes liveness though just based on vibes the cyberpaleo approach of taking a broader look at work and play seems apropos and also looking slow takes like Paul’s journey on foot will yield some useful insights.

[1] From Claude: The earliest numerals appeared in Egypt around 3400 BCE and in Mesopotamia by 3000 BCE, long before numerals emerged in China, Crete, or India Encyclopedia Britannica, yet the choice of which base to encode these marks in was neither universal nor arbitrary. Whether a culture's primary number base became 10 or 20 seems to have depended almost entirely on whether counters used only their fingers (yielding decimal systems) or both fingers and toes (yielding vigesimal, or base-20 systems) ETC USF—a literalist embodiment theory. In the northern regions of North America, western Canada, northwestern United States, Mexico, Central America, and parts of South America, the unit of counting among most native peoples was 20, with the vigesimal method forming the basis for the numeral systems of the civilizations of Mexico, Yucatan, and New Granada ETC USF. The Maya used only three symbols—a line, a dot, and a shell representing zero—to construct a fully vigesimal system with distinct names for each power of twenty Wikipedia. Meanwhile, the Babylonians employed a sexagesimal (base-60) system whose legacy persists in our division of hours and minutes Pimsleur, while Celtic languages retain traces of ancient base-20 counting in modern French (quatre-vingt for 80, literally "four-twenties") and Welsh Wikipedia. These weren't primitive precursors to "proper" mathematics: they were competing solutions to the problem of making quantities legible across populations and time, each rooted in the somatic fact of how many appendages one had to count on.

[2] See the book Ishmael for a straightforward introduction into the idea that our species has largely succeeded in creating conditions of captivity for itself as nomadic lifeways have lost out to sedentary cities. Note mentioning that is baiting Venkat a bit :P

[3] Perhaps though that will change in the coming decades and centuries as humanity stretches out to the stars or stumbles to survive on a planet made increasingly less habitable to homo sapiens.

Further reading

Cyber Paleo: Rewilding the Remote Workday

Pioneering Spirit

Jul 17, 2025

Imagine explaining to a Paleolithic ancestor that you drive a machine to sit in another cave, stare at a glowing rectangle, and then—get this—you drive back to a different cave full of metal weights a...

1 collected

Collect

The World Machines Project

Yes, we're doing this nonfiction extended universe dammit

https://contraptions.venkateshrao.com

Gordievsky died this past March, at eighty-six, in quiet English suburbia. He had lived the last four decades as a kind of ghost, protected by MI6, sentenced to death in absentia by a state that no longer exists. The obituaries called him one of the most important spies of the Cold War. That's probably right. What struck me reading Macintyre's The Spy and the Traitor, though, wasn't the geopolitics. It was something more intimate. The daily discipline of holding two true things in your head, of showing one face in a room where the other face would get you destroyed. Not because you're dishonest. Because the institution you serve has made honesty lethal.

I know that feeling. Not at those stakes, obviously. Nobody is going to drug me at a dacha outside Moscow. But every gov tech entrepreneur I know, every civic minded technologist trying to reform government from inside or alongside it, carries two stories. There's the one you tell the agency: respectful partnership, incremental improvement, aligning with existing mandates. And there's the one you tell your cofounder or your close friend at a bar: the system is profoundly broken, the procurement rules designed for transparency have produced opacity, the competition requirements meant to prevent corruption have created a cartel of insiders, and the only way anything actually changes is through what amounts to a quiet conspiracy of people who give a damn.

That double life can feel lonely. You see the absurdity and the good faith simultaneously. You know what needs to happen and what's politically sayable, and those are different sentences. You are a reformer inside institutions that resist reform, and most days the resistance wins.

Macintyre's book is, on its surface, an espionage thriller. Gordievsky was born into the KGB. His father was a proud officer. His brother joined too. The career path was settled. His conscience was not. Watching the Berlin Wall go up in 1961, reading banned Western literature, listening to the BBC on smuggled radios, he came to see the Soviet system as both criminal and philistine. By 1973 he was secretly passing intelligence to MI6. By 1985 he was running the KGB's London station and would go on to advise Margaret Thatcher on how to handle Gorbachev.

What Macintyre captures better than anyone is the moral texture of divided loyalty. Gordievsky didn't spy for money or ego. He came to believe his government was an evil, oppressive empire and decided, at enormous personal risk, to work for the cause of freedom from within its innermost sanctum. There is a version of duty that means following the rules. And there is a version that means recognizing when the rules serve an unjust system and choosing to act on a higher loyalty. The gap between those two versions of duty is where the interesting moral questions live.

That gap is also, I'd argue, where the most important work in civic technology happens today. I want to be careful with the analogy. Nobody is comparing American procurement regulations to Soviet totalitarianism. The stakes are incommensurable. But the structure of the double life, the emotional architecture of it, translates more than you'd expect.

Consider the evidence hiding in plain sight. When a section of I-95 collapsed in Pennsylvania, Governor Josh Shapiro was told that following the permitting requirements to fix the interstate would take six months to a year. He fixed it in twelve days by suspending the rules. Some of what he did was, frankly, illegal. He knew it. He also knew the interstate had to reopen. When COVID-19 hit, Operation Warp Speed delivered vaccines in under a year by routing around the standard FDA approval timeline. It probably saved two million lives globally. One was a Democratic governor's initiative. The other was launched under Trump. Both succeeded precisely by going around the existing rule structure.

As Francis Fukuyama put it recently, you cannot run a government by exception. But right now, exception is the only mode that produces results.

The procurement officer who knows the RFP process is theater but can't say so publicly. The vendor who builds the thing the agency actually needs rather than the thing the contract technically specifies. The city employee who routes around a permitting bottleneck using an unofficial workaround that everyone knows about and no one will acknowledge on the record. These aren't acts of corruption. They're acts of institutional survival. The rules have accreted to the point where strict adherence to procedure produces worse outcomes than creative interpretation of its spirit. Everyone on the inside knows this. Almost no one can say it out loud in the rooms where the rules get made.

The system has made honesty about its own failures a kind of career risk. So the people trying hardest to fix it learn to speak in two registers. Respectful partnership in the meeting. Frank assessment of dysfunction on the drive home. Not nefarious. Structurally necessary. Gordievsky would recognize the feeling.

There is a growing consensus, spanning from the Abundance movement on the center-left to the DOGE-adjacent impulses on the right, that Western democracies suffer from a state capacity deficit. Government can't build. Can't deliver. Can't execute. The soft costs of navigating procedural sludge now dwarf the hard costs of construction. Tyler Cowen and Fukuyama recently debated this on The UnPopulist, and what's striking is not where they disagree but where they converge: the American bureaucracy is over-regulated, not under-regulated. The rules constrain the ability of public servants to use common sense and good judgment. We reward compliance, not results. We sanction rule-breaking rather than promoting effectiveness.

Something has to change. But how? You can't just tell people to be braver. Gordievsky was brave and it nearly killed him. What he needed wasn't courage. He had that. What he needed was infrastructure: the MI6 handlers who managed his intelligence, the escape plan (code-named PIMLICO) that sat dormant for seven years until the day he needed it, the network of people who could act when the moment came. Bravery without institutional support is just a way to get fired.

The question is whether we can build institutional infrastructure that supports the reformer's instinct rather than punishing it. That gives civil servants and civic technologists the capacity, the flexibility, and the cover to do the work everyone privately agrees needs doing.

California is running an early experiment. Governor Newsom's Office of Data and Innovation recently launched the Governor's Innovation Fellows Program, which pulls state employees out of their home departments for six months to work on priority projects with coaching from ODI and partners in the private sector and academia. The Fellows spend 80% of their time on projects that improve efficiency, service delivery, and customer experience. They get training in product approaches, data methods, and customer research. Two cohorts have run so far, drawing from over a dozen departments across state government.

The design is an important piece of the puzzle. It is good to empower civil servants who sees what's broken, who has been carrying the reformer's double narrative, and gives them a sanctioned space to work on the honest version of the story. Not as a side project. Not on their own time. As their actual job, for six months, with executive sponsorship and institutional support.

But the program is small, state-level only, and tethered to a single governor's executive order. That's the pattern we keep repeating: innovation initiatives that live and die with their political sponsors. USDS, 18F, city innovation offices, all valuable, all fragile. What's missing is the institutional permanence that would let this kind of work survive a change in administration.

Here's what a more robust protocol might look like.

First, expand the Fellows model beyond state government. Counties, cities, special districts, water agencies, school districts: these are the governments that most directly touch people's lives, and they have the least capacity for the kind of operational self-study that the Fellows program enables. A regional innovation fellowship, funded through a consortium of local governments or through state incentive grants, could seed this work where it matters most.

Second, and this is the piece I haven't seen anyone propose yet, create a formal externship program for civil service staff. Not bringing tech people into government, which is the model we've been running for a decade, but sending government people out. Six to twelve weeks embedded in a high-performing organization, public or private, that does well what their home agency struggles with. A permitting officer spending two months inside a construction firm that navigates permits in twelve jurisdictions. A procurement specialist sitting with a startup's operations team. A water utility engineer embedded at a tech company that runs real-time infrastructure monitoring. The point isn't to privatize their thinking. It's to break the insularity that makes the double narrative necessary. When you've seen how things work elsewhere, you come back with both the vocabulary and the credibility to say what you already knew: this can be done differently.

The externship model inverts the usual civic tech logic. Instead of asking Silicon Valley to save government, it invests in the people who already work there. It trusts that the knowledge exists inside the system. The problem isn't ignorance. The problem is that the system doesn't give its own people the time, the exposure, or the institutional permission to act on what they know.

William Deresiewicz gave a lecture at West Point called "Solitude and Leadership." It's one of my all-time favorite meditations on the challenges of leadership, and it's stuck with me for a decade and a half. His argument was that the system, any system, rewards conformists. It promotes people who know how to navigate bureaucracies, manage impressions, and avoid mistakes. What it does not do is cultivate the capacity for independent thought. The ability to step back from the institution's self-story and ask whether the story is true.

Gordievsky had that capacity. It cost him everything but it changed the world. The civil servants carrying the reformer's double narrative have it too, in their own less dramatic way. They've already stepped back from the institution's self-story. They already know what's true.

The question isn't how to make them braver. It's how to build institutions that deserve their honesty. Innovation fellowships and externships won't solve the state capacity crisis by themselves. But they do something that matters: they create sanctioned spaces where the second story, the frank one, the honest operational assessment that everyone whispers on the drive home, can be spoken out loud and acted on. Not as an act of betrayal. As an act of public service.

Gordievsky needed an escape plan. Civil servants need something better. They need a system that doesn't require them to be spies in the first place.

Further Reading

Ben Macintyre, The Spy and the Traitor: The Greatest Espionage Story of the Cold War — John le Carré called it the best true spy story he ever read, and he was right. Macintyre has become the preeminent popular chronicler of British intelligence history. Read it for the thriller. Stay for the moral philosophy of what loyalty and duty actually require when the institution you serve has lost its way.

"Does America Need a Deeper State?" Tyler Cowen and Francis Fukuyama in conversation at The UnPopulist — Cowen and Fukuyama come from very different intellectual traditions but converge on the diagnosis that American governance suffers from an implementation deficit. Fukuyama's I-95 example and Cowen's insistence that Warp Speed and defense tech show the system can work by exception capture the tension at the heart of the state capacity debate. The question they can't quite resolve is whether government by exception is good enough.

California Governor's Innovation Fellows Program — The early model for what I'm describing here. State employees pulled out of their departments for six months to work on priority projects with real institutional support. Two cohorts in, drawing from over a dozen departments. Small but pointed in the right direction.

William Deresiewicz, "Solitude and Leadership" — One of my all-time favorites. The best essay I know on the difference between climbing inside an institution and actually thinking for yourself. Deresiewicz argues that the system rewards conformists and that genuine leadership requires the capacity for independent thought. Gordievsky had it. The question is whether our civic institutions can cultivate it without requiring people to become spies.

Writing Coda: This essay started as a joint review of Ben Macintyre's The Spy and the Traitor and Camilla Stivers's Bureau Men, Settlement Women, a book about the Progressive Era origins of American public administration. Claude and I went back and forth through a custom blog writing skill, beginning with a round of clarifying questions (what's the opening scene? how far do you push the spy analogy? what's Varun's deal exactly?), a full dual-book draft, a frank assessment of what wasn't working, and then the decision to let the Bureau story become its own future piece so the spy metaphor could breathe. The externship proposal emerged in the last round. Sometimes the best editorial move is subtraction.

Our cities operate in a twilight hour between two worlds. One world is the familiar machinery of departments, forms, and workflows that feel carved in stone. The other is the emerging landscape of adaptive, digitally native operations where information moves as fast as weather and where response can be shaped in real time.

That disconnect has been obvious and motivating for myself for roughly 1.42 decades (as well as many others for longer!). Saying that though is more akin to "Go West, Young Man," a loose direction rather than operational guidelines to upgrade our city operations technology stack.

One thing is certain though: the twilight will not last forever. What cannot endure will break, and the climate is already testing every brittle corner of our civic machinery. Fires, heat, and storms make that more clear every year. The only real question is whether we shape the transition or scramble after it.

My hometown of La Crescenta, a few miles from the Eaton fire burn scar, faces a familiar Southern California constellation of climate threats. Wildfire driven by wind and fuel loads. Flooding that quickly becomes debris flows and landslides. Earthquakes that test brittle infrastructure. Heat that stresses people, power systems, and water supply. And the cascading failures where one disruption amplifies another.

What makes La Crescenta interesting is not uniqueness but actually how profoundly middle of the bell curve it is for California suburbs. It's very close to the median size for California water retailers, for example. Profoundly ordinary. And what's increasingly important for profoundly ordinary places is the integration layer, the connective tissue that enables more effective city operations.

Digital twins provide a useful aspiration for that integration layer. Not the overhyped smart city-esque marketing technoblather, but the OG University of Michigan vision of a useful bridge between virtual and real space. The Digital Twin Consortium has a useful definition of the term:

A data-driven virtual representation of real-world entities and processes, synchronized at a specific frequency and fidelity

The key word there is "synchronized." A digital twin is not a static map or a dashboard. It's a feedback loop. A representation that updates as reality changes and, in turn, shapes decisions that change reality.

But here's the thing. That feedback loop already exists. It's just not digital, not integrated, and often not visible.

Every suburb in Southern California already runs on protocols. Not the TCP/IP kind (though those too), but the operational kind: the scheduled routines, inspection cadences, trigger thresholds, and coordination sequences that keep a place functioning. A street sweeper follows a route on a weekly cycle. See debris basin inspections and brush clearance mandates. Water mains get replaced on a schedule (currently accelerated as the system was built out in the 50s and 60s and now is at the end of its useful life).

These are feedback loops between URL and IRL, between the virtual world of planning, scheduling, regulation, and modeling and the physical reality of pipes, basins, trees, and fire. A water quality sampling schedule is a protocol: it defines what gets measured, how often, where, and what triggers a response. A red flag warning is a protocol: atmospheric conditions cross a threshold, a warning propagates through institutional channels, and pre-positioned resources activate. Brush clearance enforcement is a protocol: a regulatory deadline triggers inspection, inspection produces compliance data, noncompliance triggers enforcement action.

The protocols stack. They operate at different temporal scales, from the SCADA system polling water pressure every fifteen seconds to the regional transportation plan projecting decades out. Each layer has its own rhythm, its own institutions, its own failure modes. And the gaps between layers, the places where a fast protocol and a slow protocol fail to coordinate, are where cascading failures live.

A digital twin, understood properly, is not a replacement for these protocols. It's a way to level them up. To increase the frequency and fidelity of the feedback loops that already exist. To make visible the coordination gaps between layers. To compress the latency between observation and response.

Here's what that protocol stack looks like for a suburban Southern California community like La Crescenta:

That last question is the one that stays with me. Every suburb in California was designed for a climate that is departing. The protocols that maintain these places, from hourly dispatch to decadal planning, were calibrated to a world that is changing faster than the institutional cycles that govern response. The digital twin aspiration is really a question about whether we can recalibrate those feedback loops before the gap between assumption and reality becomes catastrophic.

Throughout my professional career and curious passion for city operations, the gap between what theoretically should be done per global best practices, academic theory, and other intellectually correct goals and what happens on the ground has always struck me.

Human error, resource constraints, coordination headwinds... many real world factors mean that often obvious needs like having digital geographic information system (GIS) layers for basic infrastructure don't always exist or aren't integrated and shared across the relevant utilities. Basic infrastructure in the US has layers upon layers of rules, procedures, administration, insider institutional politics that prevent familiar obstacles to urban digital tool applications.

I have found that there is no substitute for getting into the field to see firsthand what is actually going on and use that analogue data as a tool for designing what might be.

Being a data sophist of sorts, I have a recurring daydream to design a digital twin armed with pencil, paper and a few print books like the handy MIT guide to "Street Fighting Mathematics." Arguably with sufficient completeness, a janky contraption toy model composed of coordinates, flows and stocks is a bridge between the virtual and real world. Just synced at a relatively low fidelity and high latency.

My young family recently had our second kid. When my first daughter was young, I loved wearing and walking her everywhere. As I'm taking some bonding leave, I'm excited at the opportunity for a little urban exploration and protocol watching, to rediscover the small town in Southern California where we live.

The protocol stack above is a field guide of sorts. At every temporal scale, there are things you can see if you know what to look for. The spatial distribution of rooftop solar, both at a neighborhood-wide scale and in terms of rooftop orientation. The logic of fire hydrants, storm drains, power lines, flood control channels, debris basins, natural gas lines, sidewalk placement, bus lines. The dispersion of trees: geographically, by apparent age, by species. The traffic flow at peak up and down Briggs Terrace, a neighborhood with one ingress and egress.

Each observation is a data point in a very low fidelity digital twin. A notebook full of coordinates, flows, and stocks. The protocol stack gives that notebook structure, a way to sort observations by the temporal rhythm they belong to and the institutional layer they reveal.

There's a story about William Mulholland, LA's famous water engineer who served as the de facto founder of the modern city, that has stayed with me. Before a single pipe was laid for the Los Angeles Aqueduct, he walked the route. Two hundred and thirty miles through the Mojave Desert and the Sierra Pelona Mountains, bacon and beans for meals, sleeping on frozen ground, barometer in hand.

The barometer was the key instrument: by measuring air pressure at each step, Mulholland was calculating altitude, hunting for the precise grade that would keep water moving under gravity alone, fast enough to flow but slow enough not to tear the concrete apart. Too steep and the channel erodes. Not steep enough and the water stalls. The whole system depended on finding a continuous, calibrated descent across some of the most punishing terrain in North America. By the time construction began, he claimed he knew every foot of the route.

I plan to keep that story with me when I wander around La Crescenta with a little human. You cannot understand a place from a dashboard. You cannot model what you haven't seen. A walk is a protocol observation instrument. A notebook is a model. The friction between what you imagine a place to be and what it actually is, that gap, that surprise, is where the real work begins.

Where does the built environment assume a climate that no longer exists? That's the question I'll carry.

Further Reading

Terraforming After the Eaton Fire

Pioneering Spirit

Aug 24, 2025

When the hurricane force firestorm came down Eaton Canyon in January, they did not discriminate between chaparral, oak, or stucco. In a single night the canyon reminded us of how ephemeral our cities ...

0 supporters

Support

Be Like Water: Field Notes from the Trenches of Regional Climate Adaptation

Pioneering Spirit

Jan 27, 2026

Be like water making its way through cracks. Do not be assertive, but adjust to the object, and you shall find a way around or through it. If nothing within you stays rigid, outward things will disclo...

0 supporters

Support

I didn't, and I've been mulling the project ever since.

Here's the complicating fact: I'm supposed to be on the other side of this fight. I supported Abundant Housing LA. I led the research function at California Forward and worked on policy to increase statewide housing supply. I believe, genuinely, that California's housing shortage is one of the central civic failures of the twenty first century. I am, in theory, a YIMBY.

And yet.

I served for eight years on the board of an education reform nonprofit whose founding predated the existence of charter schools. That experience left me with a specific kind of skepticism -- not about the problem, but about the proposed cure.

The charter school movement was directionally right. American public education was failing millions of kids, especially poor kids, especially kids of color. The system was calcified, the incentives were broken, and the adults running it had mostly made their peace with failure. Reformers who said "this is unacceptable and we need structural change" were correct.

What happened next is instructive. The reform energy got organized. The policy levers got pulled. Charter schools proliferated. And the results were -- mixed. Some charters were genuinely excellent. Others were indistinguishable from the schools they replaced, or worse. The blunt instrument of state-level policy reform turned out to be a kludgy way to address a problem that was deeply local, deeply contextual, deeply dependent on execution quality that no legislation could mandate into existence.

I think about that arc a lot when I watch the YIMBY movement. The diagnosis is correct. California doesn't build enough housing. The regulatory apparatus is broken. The human cost -- priced-out families, brutal commutes, kids who grew up here and can't afford to come back -- is real and serious and underweighted in our politics. The reformers who say "this is unacceptable and we need structural change" are correct.

But SB 9 and its cousins are blunt instruments. The legislation lands. The housing doesn't always follow. And when it does, it's often not the housing anyone actually needed.

Which brings me to the missing middle, and to what I think is the part of this conversation nobody wants to have.

California has quietly sorted itself into two housing modes. There are the homeowners -- disproportionately Boomers -- sitting on assets that have appreciated beyond all reason in suburbs that have calcified around their preferences and their property tax bills. And there is deeply subsidized housing for people who have fallen all the way through: the chronically homeless, foster youth aging out of the system, the most vulnerable. Both of these populations have political constituencies. Both have policy.

The proposed five-story, 80-unit building around the corner from my house is in the second category. A significant portion of the units are reserved for foster youth, with a single program coordinator. My heart genuinely goes out to those kids. The program may be excellent. I hope it is.

But I work in downtown Los Angeles, near Union Station. I see every day what happens when ambitious housing promises meet thin implementation capacity. The open-air disorder, the broken windows, the accumulated evidence that the public sector's track record in LA on supportive housing is, to put it charitably, uneven. That's not an unreasonable thing for my neighbors not to want in our little town. To be honest, I fear importing problems I see littered around downtown LA into La Crescenta.

And none of it -- not the 80 units, not the foster youth program, nothing in the proposal that has my neighborhood up in arms -- addresses the actual crisis for the family that earns a normal income and wants to raise kids in the suburb where they grew up. That person has seemingly no constituency these days. The California dream -- that a working family could put down roots here, that the suburbs could be places where ordinary people built ordinary lives -- is at best on life support, and the current housing debate on both sides mostly isn't honest about that. The suburbs are slowly becoming retirement communities for people who got in early. The rest of the region has pockets of subsidized housing for people who fell all the way through, with nothing in the middle for everyone else.

Here is what most participants in the NIMBY/YIMBY debate treat as permanent backdrop: the regulatory framework within which all of this plays out. It isn't.

Zoning as we know it is barely a century old. CEQA -- the California Environmental Quality Act, the procedural machinery that makes virtually every significant development project an invitation to litigation -- was born in 1970. It was designed to slow California's rampant postwar growth, which had genuinely run roughshod over communities and ecosystems. That was a reasonable response to a specific moment. What it became is a general obstruction machine, a legal structure that any organized interest can weaponize to delay or kill almost anything.

The fees, the discretionary review, the public notice requirements, the layers of required process: none of it was designed as a coherent system. It accreted over fifty years of well-intentioned responses to specific local grievances. So much of California's existing system is absolutely insane, like conducting environmental review through the courts adjudicate hypothesized impacts rather than look at objective scientific standards of actual wholistic impact on our planet. And critically, all of it was designed for a pre-internet world -- physical mailers, newspaper notices, in-person hearings on weeknights.

The entire civic architecture of California planning assumes a participation technology that was already obsolete by the time most Millennials finished high school. Meanwhile, the rest of the world has been running experiments. The vTaiwan platform brought together civic hackers, government officials, and citizens to build genuine consensus on contested policy questions -- not through comment periods but through structured online deliberation that actually shaped legislation. Helsinki built a participatory budgeting platform that engages nearly a tenth of the city's residents in deciding how public money gets spent. Cities from Dubai to Montreal are piloting AI-assisted visualization tools that let residents see proposed development in context, in real time, and shape it before it hardens into a one-page mailer.

California, historically a pioneer, is largely absent in this picture. Both NIMBY and YIMBY politics take California's kafka-esque, pre-digital planning apparatus as given. They argue about outcomes while accepting the broken, obsolete operating system that produces those fights. That's the argument I want to step outside of.

What the flyer in my mailbox actually revealed wasn't whether 80 units is too many or parking requirements are too high. It revealed that La Crescenta has no shared vision for what it wants to become.

La Crescenta is unincorporated county. There's no city council, no mayor, no locally accountable body whose job is to hold a long-term vision for this valley. There was a general plan update, a specific plan, assorted wonky urban planning documents that nobody but professional advocates and committed gadflies reads. One public meeting, at a weird hour. A one-page notification. Then: everyone up in arms.

This is not how you build trust. And the parking and density arguments that erupted -- those aren't really about parking and density. They're proxy wars for something deeper: the absence of civic vision, the corrosive suspicion that decisions are being made at and around communities rather than with them, the accumulated distrust of institutions that have made big promises and delivered broken windows.

I'd like to believe there's a better way. Not NIMBY, not YIMBY. Maybe In My Backyard -- MIMBY -- as a position that takes both the housing crisis and the legitimate desire for community voice seriously enough to ask harder questions about process.

What if communities had real deliberative tools before the flyer hit the mailbox? What if augmented reality let you see a proposed building in actual context on your actual street, and then gave neighbors a structured way to say "not that, but why not this?" What if permitting reform meant not just faster approvals but genuine community co-design -- streamlined enough that building good things was the easy path, deliberative enough that communities felt like authors rather than subjects?

La Crescenta was built by people who built things. Flag lots, DIY additions, families who looked at a piece of hillside and figured out how to make it work. That spirit is still here. What's missing are institutions worthy of it -- planning processes that produce shared vision instead of cryptic mailers, that make the normal family's aspiration to put down roots here as legible to policy as the Boomer homeowner's interest in protecting property values.

Maybe In My Backyard. Not a slogan for blocking things, and not a blank check for Sacramento to override the people who actually live somewhere. A bet that if we burn through the procedural brush that has accumulated over fifty years -- the discretionary review, the litigation tripwires, the one-page mailers -- we can pioneer new protocols for approving housing that are faster and simpler and genuinely democratic, protocols that the rest of the world is already building. What we'd find underneath is something worth trusting: communities that know what they want, and are capable of choosing it.

This became clear to me during the pandemic, on one of countless Tuesday mornings that blurred into each other, walking my dogs through the neighborhood while half-listening to a colleague explain an arcane concept in the state budget. The call was useful. My thinking was sharp. My dogs were happy. By any honest accounting, I was doing my job well. But was I “working?”

Around me: another neighbor, also on a call, also walking a dog. A gardener trimming the parkway grass — clearly working, in the most legible sense. A parent pushing a stroller and talking to a toddler about a squirrel, which is either parenting or play or education or all three, and which, if that parent is a paid nanny, counts toward GDP, and if not, counts toward nothing in particular. The mail carrier, moving with the efficient rhythm of someone who has done this ten thousand times.

The same word — “work” — is supposed to cover all of that. It doesn’t. It can’t. And our insistence that it does is quietly making us worse at thinking about what we actually do, and why.

* * *

Consider the load we place on this four-letter word. We go to work, but we also work out, work on relationships, work through grief, work up an appetite, work a room, and work the system. A thing that functions is said to “work.” A persuasive argument works. A plumber works. So does a poet. So does a stay-at-home parent, though if you ask them in a certain tone of voice, they will tell you they “don’t work.” Physics has a whole other concept embedded in the word work.

Linguists have a name for words that expand to fill whatever conceptual space you point them at: weasel words. “Work” is perhaps the weaseliest of them all, because unlike most vague words, it carries enormous moral freight. In America especially, to admit that you are not working, not a hard worker, not committed to your work — this is close to a confession of character failure. We have arranged things so that the word simultaneously describes what you do and judges whether you are a good person for doing it.

This is a strange and unstable combination. And it becomes stranger still the moment you try to pin the word down.

Is work something you’re paid for? Then raising children is not work, and neither is keeping a household, cooking dinner, or the volunteer labor that holds most communities together. Is work something you’d rather not do? Then a person who loves their job doesn’t work, while someone grinding through a shift they hate works very hard indeed. Is work the opposite of play? Then flow states — that condition of absorbed, effortless productivity that every knowledge worker claims to be chasing — might not be work at all, which would make it ironic that we schedule it into our calendars.

The word collapses distinctions we actually need. It’s not a useful umbrella; it’s a tarp thrown over a pile of things we haven’t bothered to sort.

* * *

We didn’t always live under this tarp. The fusion of labor with moral identity is historically recent and geographically peculiar. For most of human history — and still in many parts of the world — what you did to put food on the table was not the primary answer to the question of who you were. You were a member of a family, a village, a faith community, a guild. Your productive activity was embedded in those relationships, not extracted from them and rebranded as your defining characteristic.

The Protestant Reformation, and especially its Calvinist strains, began to change that. If prosperity was a sign of divine favor, and if diligence was the path to prosperity, then hard work became not just economically useful but spiritually significant. The Industrial Revolution then institutionalized this theology into a time-clock. The wage relationship — trading hours for money, for a stranger, in a place not your home — became the template for all productive activity.

We inherited that template and mistook it for nature. When we say “So, what do you do?” as our opening move with a new person, we are not asking a neutral question. We are asking which slot in the economic order they occupy, and we are using that slot as a shorthand for their worth. It’s a question so habitual we’ve stopped noticing that it’s a question at all.

Other answers are available. In a different frame, you might introduce someone as active in the school board, or as the person who keeps the community garden running, or as someone who makes extraordinary tamales and brings them to every neighborhood gathering. A job would be what they do to pay rent. It wouldn’t be the whole story.

* * *

The conceptual overload of “work” might be merely amusing if the stakes were low. They are not. The word is doing active damage in the domain where I spend most of my time: knowledge work, where outputs are frequently invisible, where the connection between effort and result is loose, and where “working” has consequently come to mean something more like “performing the appearance of productivity.”

A friend who reviewed an earlier version of these thoughts pushed back sharply. “This sounds like a knowledge worker’s leisure problem,” he said. “It doesn’t cover people who need to be physically at work — the blue-collar 9-to-5 jobs.” He’s right that the acute form of this illness is white-collar. But the signaling problem runs wider than he gives it credit for.

When outputs are illegible — when it’s genuinely hard to tell whether a given hour of effort produced something valuable — organizations reach for proxies. Hours worked is the most available proxy, which is why banking culture at firms like Goldman Sachs has produced analysts working 9 AM to 5 AM and defending it as a rite of passage. No one produces excellent analysis twenty hours a day. What they produce is a demonstration of commitment, expressed in the currency of hours, because the actual currency — quality of thinking — is too hard to price.

The result is a workplace pathology familiar to anyone in an office: the performance of busyness. Answering email at all hours. Keeping your status light green. Never being the first to leave. None of this is necessary for the actual work. All of it is necessary for the word “work” to keep doing its moral job — signaling that you are a person who works hard, who is committed, who deserves to keep their seat at the table.

The mind, meanwhile, needs time and space to actually think. It needs unscheduled hours, long walks, conversations that go sideways. The best ideas I’ve had in my own career arrived on dog walks, in the shower, in the middle of conversations that had nothing to do with the problem. None of that looks like work. All of it was.

* * *

All of this was already strange. Then artificial intelligence arrived to make it obviously untenable.

A friend who runs a technical team described watching his colleagues celebrate a “10x productivity improvement” from AI tools — drafting documents faster, processing email in a fraction of the time, generating slide decks in minutes. The celebration felt earned. And then he started watching what people did with the recovered hours. A lot of scheduling. Some reorganization. Considerable discourse about the AI tools themselves. “It’s very unclear,” he said, “how folks are actually accomplishing more.” Parkinson’s Law, it turns out, is not repealed by software: work expands to fill the time available, and if AI compresses the time a task takes, we tend to find new tasks rather than do fewer of them.

This is the second-order problem that “work” as a concept is not equipped to handle. If the word means “tasks completed” and AI completes tasks faster, then productivity has increased by definition. But if work was always partly about signaling — demonstrating diligence, earning trust, proving commitment through visible effort — then AI doesn’t increase productivity so much as reveal that a chunk of what we called work was never really about output at all.

The question “what is productivity?” was always lurking under the surface of knowledge work. AI has yanked it into plain view. Total Factor Productivity? Personal task completion? Meaningful impact on the physical world? These are not the same thing, and the word “work” was letting us pretend they were.

* * *

So what do we do instead?

The honest answer is that we probably need several words where we currently have one. The anthropologist’s distinction between “work” and “labor” is a start — labor being the repetitive, sustaining activities that keep life going, and work being the project-oriented effort that builds something new. But even that doesn’t fully capture the range. Too many industrial era analogues have been entirely interwoven.

A more useful decomposition might invoke deeper human lifeways like: gathering, which is the attentive, somewhat repetitive effort of processing the stream — email, logistics, routine decisions, the maintenance of systems already in place. And hunting, which is the open-ended, collaborative, exploratory effort of tackling problems that don’t yet have known solutions. Both are real and necessary. They require different conditions, different rhythms, different mental states. Treating them as the same thing — as “work” — and scheduling them interchangeably is one reason so many people feel productive while accomplishing less than they should.

There is also play, which gets a worse reputation than it deserves. Play is how skills develop when no one is watching the scoreboard. It is how creative fields explore terrain before they know what they’re looking for. It is also, as David Graeber argued in one of his better essays, potentially a large part of the point of being alive — something that the word “work” systematically cannot acknowledge without undermining itself.

A simpler practice: before you describe something as “work,” try filling in the blank. I am doing [specific activity] because I want to [specific outcome]. Not “I’m working.” Not “I’m going to work.” What, specifically, are you doing, and toward what end? You may find that the answer is clear and satisfying. You may find it is unclear and that this is informative. You may find that the activity is valuable regardless of whether it fits any traditional definition of work, or that it fits every definition and is nonetheless pointless.

The point is not to optimize. The point is to see.

* * *

Here is a small experiment worth trying: go one week without using the word “work” in your internal monologue. Not forever. Just a week. You will find, I suspect, that almost everything the word usually covers can be described more accurately without it — and that the things that can’t be described without it were never quite real to begin with.

And the next time you meet someone new, consider not asking what they do. Ask instead what they like to do, or what they’re trying to figure out, or what they made recently. You might learn something more interesting. You will almost certainly get a more honest answer.

The Protestant Work Ethic built a lot of things worth having. It also gave us a word that has been punching above its conceptual weight for three centuries. It is time to make it work a little less hard.

—

This essay draws on David Graeber’s “What’s the Point If We Can’t Have Fun?” and Venkatesh Rao’s “A Big Little Idea Called Legibility.” Both are worth your time.

Cover image: Edward Hopper's Office at Night.

In one, it is a smoldering dystopia. Broke. Burning. Unaffordable. Fleeing residents in rented U-Hauls headed for Texas.

In the other, it is a sun-drenched dreamscape. The birthplace of the internet. The laboratory of climate action. A place so dynamic it could stand alone as a nation state.

Both movies contain truth. Neither is the whole story.

If California were a country, it would rank among the world’s top economies, ahead of the United Kingdom and India in GDP, a global powerhouse of agriculture, technology, entertainment and higher education.

Still California's quality of life is most similar to Poland and early childhood education resembles Latvia, a former Soviet block country. Massachusetts by contrast has an education system that resembles Singapore.

Beyond the boosterism and doomer-ism, California is a real place with problems and opportunities. This year there is a Governor's race with the chance to make a dent in those enduring challenges.

In a past job, I helped develop the California Dream Index, "the most serious effort at quantifying... whether California is succeeding" per Joe Matthews. Since that's now a few years out of date, I had my ai assistant update that analysis to the latest below.

The results are humbling as they show how intractable many of California's challenges are. Governor Newsom entered office with a lot of energy and rhetoric about transformation.

Yet the last 8 years have really demonstrated the merits of Governor Brown's hard won wisdom. From his 2010 inaugural.

I have thought a lot about this and it strikes me that what we face together as Californians are not so much problems but rather conditions, life’s inherent difficulties. A problem can be solved or forgotten but a condition always remains. It remains to elicit the best from each of us and show us how we depend on one another and how we have to work together.

Governor Brown to his credit set clear achievable goals like balancing the state's budget and building up a sizeable reserve. California's rainy day fund had zero dollars in 2009-10 and there were whispers of the state becoming a territory again so it could declare bankruptcy (state's are barred from reneging on contracts by the US constitution).

Today there is another budget race and there is a lot of promises about addresses the affordability crisis, the housing crisis and many other of California's crises like its efforts at climate mitigation and adaptation.

How will this time be different? To quote California's once and future governor again, how might we "turn a breakdown into a breakthrough?"

Below is a very thorough analysis of the specific CA Dream Index measures courtesy of Claude code.

California Dream Index: Trend Analysis

Tracking Access to Opportunity Across California, 2010-2024

The California Dream Index (CDI) is a composite measure of access to opportunity developed by CA FWD. It aggregates 10 indicators spanning education, housing, infrastructure, environmental quality, and economic prosperity into a single view of how well the state is serving its residents.

This analysis examines statewide trends across the full 2010-2024 time series, with particular attention to changes during Governor Gavin Newsom's tenure (inaugurated January 7, 2019). Data is drawn from the U.S. Census Bureau's American Community Survey, the EPA's Air Quality System, the California Water Boards' HR2W system, and United Ways of California's Real Cost Measure.

A Note on ACS 5-Year Data

Each year's ACS 5-year estimate is a rolling average over a 5-year window. The 2024 estimate covers 2020-2024. This smoothing dampens sharp changes and distributes the effects of any policy or event across multiple data years. It also means that COVID-19's impacts appear gradually across the 2020-2024 estimates rather than as a single shock.

The 10 Indicators

The Dashboard View

All 10 indicators plotted together reveal a mixed picture of California's trajectory. Some indicators show steady improvement, others have plateaued, and a few have been disrupted by COVID-19.

Indicator-by-Indicator Analysis

Post-Secondary Attainment: Steady Climb

California's college attainment rate has been one of the clearest success stories, rising from 38.2% in 2012 to 45.0% in 2024. The pace of improvement has actually accelerated slightly during the Newsom era, gaining about 0.6 percentage points per year compared to 0.5 pre-Newsom.

Early Childhood Education: A COVID Casualty

Preschool enrollment is the starkest casualty of the pandemic era. After hovering near 50% for most of the 2010s, enrollment dropped sharply beginning with the 2020 ACS data (covering 2016-2020) and fell to 44.5% by 2024. This 5-percentage-point decline during the Newsom era contrasts with virtual stability in the prior period. The question now is whether the state's investments in universal transitional kindergarten can reverse this trajectory.

Broadband Availability: The Digital Divide Narrows

Broadband access (available in the ACS from 2017) has been the single fastest-improving indicator, climbing from 82.6% to 93.5%. The expansion has been steady across both eras, reflecting both market forces and public investment. The pandemic likely accelerated adoption as remote work and school made connectivity essential.

Affordable Rent: Treading Water

Rent affordability has been remarkably flat over the entire period, fluctuating narrowly between 43% and 46%. There's no meaningful trend in either direction. This is arguably one of the most stubborn challenges facing the state -- neither improving nor worsening despite significant policy attention to housing.

Short Commutes: Pandemic Silver Lining

Mean commute times rose steadily from 26.9 minutes in 2010 to a peak of 29.8 minutes in 2019-2020, then reversed course. By 2024, commute times had fallen back to 28.7 minutes. The remote work revolution triggered by COVID-19 appears to have had a lasting structural effect on commuting patterns -- one of the few unambiguous improvements attributable to the pandemic era.

Poverty Rate: Meaningful Improvement

California's poverty rate (as measured by the ACS) declined from a peak of 16.4% in 2014 to 12.0% in 2024. Notably, the decline during the Newsom era (13.4% to 12.0%) has been slightly slower in annual terms than the pre-Newsom decline, though the rate was already falling from a higher base. Federal pandemic relief programs (stimulus checks, enhanced unemployment, child tax credit) likely contributed to the 2020-2021 figures.

Homeownership: Slow Recovery from a Long Decline

Homeownership bottomed out around 54.1% in 2016 after a decade-long post-recession decline, then began a slow recovery. By 2024 it reached 55.9% -- still well below the 57.4% rate of 2010. The Newsom era has seen a slightly faster pace of recovery than the pre-Newsom period, though the gains are modest relative to the housing crisis narrative.

Air Quality: Volatile and Weather-Dependent

Air quality, measured by the EPA's county-level Median AQI, shows no clear long-term trend. Values bounce between the low-to-mid 40s and low 50s depending heavily on wildfire seasons. The best year on record (2019, median AQI 46.7) was followed by severe wildfire years in 2020 and 2021. By 2024, the median AQI was 47.9 -- slightly worse than 2019 but better than the wildfire-scarred years.

California's air quality story is really a tale of two geographies. The chart below shows dramatic differences between coastal counties (San Francisco, Santa Cruz) and inland agricultural and wildfire-prone areas (Fresno, Riverside, San Bernardino).

Drinking Water Quality: Incremental Gains

The percentage of water systems operating without violations has been high throughout the period (96-98%), with incremental improvement over time. The 2020 figure (98.1%) represents the strongest year in the dataset. This indicator reflects cumulative investment in water infrastructure and enforcement.

Cost of Living: Limited Data Window

The Real Cost Measure data is only available for 2014-2017, showing the percentage of households with income above the real cost of living rising from 61.1% to 64.1%. This short window makes trend analysis difficult, but it captures a period of economic expansion.

Pre-Newsom vs. Newsom Era: Rate of Change

The bar chart below compares the annualized rate of change for each indicator between the pre-Newsom (2010-2018) and Newsom (2019-2024) eras. Positive values indicate improvement (poverty and AQI are inverted so that positive = better).

Key takeaways from the era comparison:

• Accelerated improvement: Post-secondary attainment, poverty reduction, homeownership, and drinking water quality all showed faster improvement during the Newsom era

• Decelerated improvement: Broadband availability continued gaining but at a slower pace (the low-hanging fruit has been picked)

• Reversed course: Early childhood education went from flat to sharply declining -- the single biggest regression

• Essentially flat: Affordable rent barely moved in either era

• Mixed signals: Air quality improved slightly faster in the Newsom era when measured as an annualized rate, though individual years are dominated by wildfire variability

Racial Disparities: The Homeownership Gap

The CDI tracks several indicators broken down by race and ethnicity. Homeownership is one of the starkest examples of persistent racial inequality in California.

In 2024, White Californians had a homeownership rate of 62.5%, compared to 35.7% for Black Californians -- a gap of nearly 27 percentage points. Hispanic/Latino households (46.0%) and Asian households (61.4%) fall in between.

The Newsom era has seen modest recovery across all groups. Black homeownership, after declining from 38.8% in 2010 to a nadir around 34.2% in 2016, has recovered to 35.7% -- still well below its 2010 level. The White-Black gap has narrowed slightly from its widest point but remains enormous.

County-Level Variation

The statewide numbers mask enormous variation across California's 58 counties. Post-secondary attainment illustrates this clearly.

The most educated counties (Marin, San Francisco, Santa Clara) have attainment rates above 60%, while the least educated (Kings, Merced, Colusa) hover near 15-20%. This roughly 3:1 ratio between the highest and lowest counties underscores the geographic dimension of inequality in California.

During the Newsom era, the gains have been broadly distributed across counties, though the magnitude varies. Rural and agricultural counties generally started from a lower base and saw somewhat smaller absolute gains.

Regional Patterns

California's CERF (Community Economic Resilience Fund) Regions group counties into economic areas. The regional view shows a clear coastal-inland gradient in educational attainment.

COVID-19 Recovery Tracker

The indexed chart below normalizes all indicators to their 2019 values (= 100), allowing us to compare recovery trajectories across different measures. Lines above 100 indicate improvement relative to 2019; lines below indicate regression.

The most striking features:

• Broadband and post-secondary attainment surged well above their 2019 baselines and never looked back

• Poverty rate (inverted: up = better) improved relative to 2019, though recent years suggest the improvement pace is leveling off

• Commute times (inverted) improved sharply, reflecting structural changes in work patterns

• Early childhood education dropped well below 2019 and shows no sign of recovery

• Affordable rent and homeownership have hovered near their 2019 levels

Summary Scorecard

*2019 homeownership data had a timeout during collection; trend shows recovery from 2020-2024. **Drinking water latest available year is 2020.

Bottom line: Of the indicators with comparable 2019 and 2024 data, five improved, three declined. The gains were concentrated in education (post-secondary), infrastructure (broadband), and economic indicators (poverty, commute times). The declines were concentrated in early childhood education (the largest single regression), rent affordability (marginal), and air quality (weather-dependent). Housing remains the most resistant category to meaningful improvement.

Methodology Notes

• ACS indicators (7 of 10): Pulled from the Census Bureau API at the state level. These are direct Census estimates, not population-weighted tract aggregations (as in the full CDI methodology).

• Air Quality: County-level median AQI downloaded from EPA AQS annual summary files, averaged across California counties.

• Drinking Water: Enforcement action records from the CA Water Boards HR2W system. Metric approximates the fraction of the state's ~7,400 water systems operating without violations in each year.

• Cost of Living: From United Ways of California's Real Cost Measure dashboard. Only 2014-2017 data is available in the source file.

• ACS 5-year estimates represent rolling averages, which smooth year-to-year variation but create lag in reflecting current conditions.

• Newsom era is defined as 2019-2024, corresponding to his inauguration in January 2019.

Data Sources

• U.S. Census Bureau, American Community Survey 5-Year Estimates (2010-2024)

• U.S. Environmental Protection Agency, Air Quality System (AQS) Annual Summary Data

• California State Water Resources Control Board, HR2W Drinking Water System Data

• United Ways of California, Real Cost Measure Dashboard

Analysis by Claude Code using the California Dream Index framework. Data current through ACS 2024 5-year estimates (released January 2026).