TIL Mirror is now Paragraph. Now that it's 2026, it's an apt opportunity to review my first post on Mirror from '21, one last time.
The biggest divergence from this post is that demand for public blockchains has proven to be a negligible fraction of expectations. As a result, overall progress has been much slower than expected. Of course, I acknowledged that error as early as 2022, and examined why throughout 2022-23 with a focus on applications rather than infrastructure. What has continued to surprise me, and still shocks me to this very day, that the industry's sole focus is *still* infrastructure and supply expansion, almost no one cares about applications or stimulating demand. No wonder demand has fallen far, far short of expectations.
Not that the post was all wrong. Bear in mind that in 2021 the play was "the future is hundreds of L1s", so generally, things have played out as hoped, just at a negligible fraction the scale.
https://paragraph.com/@polynya/a-vision-of-ethereum-2025
I haven't talked about this for years now, but worth revisiting:
The real USP of strict global consensus applications is tokenless, governanceless, zero fee, permanent applications that target a focused niche
No idea what the context is, but it's pretty clear in the last few years the crypto industry has barrelled down the cascade of degeneracy. At this point, it's institutionalised and normalised, and even the biggest players are deep into the game.
To be clear, I use "gambling" in the broader sense, and yes, we're obviously seeing the same in AI stocks, Gold etc. Some of it is real, sure. The TAM for gambling is 1,000x larger than building actual products in the short-medium term, so I don't blame anyone, especially respect those who are clear they just want to extract.
It's all a shame because the few actual sustainable products are better than they have ever been, and if the crypto industry had gone down the productive path instead, I'd still be writing about those.
In 2017, Bitcoiners fought tooth and nail over block size, with one camp claiming Bitcoin will die without drastically increase throughput, while the other saying Bitcoin will die by increasing system requirements and compromising decentralization. In November 2017, Bitcoin Cash got to nearly 50% the market cap of Bitcoin under this premise, with a substantial minority convinced Bitcoin Cash would flippen Bitcoin
They were both wrong!
8 years later, Bitcoin despite its anemic block size is barely utilized, but it's become a multi-trillion dollar success story. Bitcoin Cash, meanwhile, is 0.5% the market cap of BTC, 1/100th of its peak
The lesson: the TAM for gambling is thousands of times greater than whatever "blockchain utility" you can think of
Re: "subsidy sustainability" - also doesn't matter, as by the time that happens, most BTC will be owned by ETF custodians, and they can effectively turn it into PoS or PoA or whatever they want. All that matters is belief and religious fervour in the ticker
Dan is right, but that aside;
Personally, I do care, but the crypto industry has failed to deliver it, and has chosen a path of plutocracy & corporatocracy. Which is fair enough, as institutional speculation has a 3-4 orders of magnitude larger TAM, and it's more important to pump bags than deliver maximally neutral technology.
On that note, there's a multi-trillion dollar cults built around BTC, ETH et al today, and it's definitely worth targeting that, even if technologically the solutions are suboptimal. Remember, you're invested in a ticker, not your blockchain or its tech, so in usecases where there's better & more appropriate tech than your blockchain available or waiting to be invented, you should embrace it under your banner, instead of dismissing it as "bearish". Blockchains are clearly transitory tech begging for obsolescence by more decentralised, more democratic, more efficient tech.
I've written about this before, but reminder:
A fully private CBDC is the ideal solution, and the only way to replicate the self-custody & anonymity of physical cash
You eliminate multiple unnecessary middlemen - stablecoin issuers, their banking partners, blockchain infrastructure and their node operators, at each step they add to your cost, but more importantly, they control your funds
Instead of Fed > Tether > Ethereum > Arbitrum > User, you go directly Fed > User
Of course, you need strict global consensus for products like LUSD or USDS, but over a decade the market has decisively spoken - USDT & USDC are the dominant types of stablecoin used
The privacy is critical. It can be granular using ZK proofs, so almost all transactions are fully private, and only certain ones marked automatically per legal parameters can be traced. USDT et all offer zero privacy.
I don't expect a major government savvy enough to pull this off any time soon, but I do know this is the right solution to the stated problem
Well, that didn't take long!
I've seen comments about "decentralisation is the goal" - it's not. I'll address it again, though it's just repeating my blog posts from 2023.
USDC is issued by Circle, a publicly owned company regulated by government agencies with a set of checks and balances, which are indirectly elected by hundreds of millions of US citizens.
By deploying on Ethereum, you make USDC less decentralized by adding additional intermediaries (chiefly a few thousand noderunners) which aren't necessary for a lot of usecases.
Ethereum is very centralised, it's just *differently centralised*
However, there'll always be usecases where that differential centralisation will be useful. There's also sticky network effects, as we've seen with USDT on Tron.
I only care about using the best technology for the job, and for most stablecoin payments, Ethereum or Tron are not it. But there are also certain usecases where they are, like jurisdictions that restrict USDC and access to Ethereum DeFi.
Given regulatory acceptance, technically, stablecoins no longer need the strict global consensus property of blockchains. The most efficient way would be for regulated issuers like Tether, Circle etc. to run their own infrastructure with blockchains as an option. It might even be prudent to directly transact with CEXs like Binance, as that's where a lot of consumer use of stablecoins is.
However, given Tron and Ethereum (+ L2s) have built network effects over the last decade, it's likely stablecoins will remain a big part of those networks for the foreseeable.
Of course, certain unregulated products like USDS or LUSD will continue to require Ethereum, but that's a much smaller niche.
In 2021, I wrote extensively on this topic. On paper, purely technically, it was a no-brainer - greater performance, interoperability, secure access to ethereum liquidity, security & decentralization, significantly lower inflation for economic sustainability, etc. etc.
Since then, all major new chains have been L2s - from DAOs (Unichain) to crypto corporations (Coinbase, Kraken) to tradfi (Robinhood), with the sole exception of Hyperliquid. There have been some L1 > L2 like Celo as mentioned. However, over 4 years, most legacy L1s have remained L1s.
Potential reasons:
- Application demand growth has been a fraction of expectations. Most L1s have barely any activity and have no need for the greater performance or interoperability of L2s
- L2 & blob dev have slowed down since '22 potentially due to the above, with performance much lower than expected in '25
- Much of their token value is based on being an anti-ethereum cult
- The industry has pivoted to degeneracy, decentralization/security matter less now
This is actually happening, potentially before ~2031 that I expected at the time (though still years away). ~Infinite chains, ~infinite composability, ~infinite scalability, all verified on your phone with a single infinitely recursive ZKP - this remains the only way to global scale
Of course, in 2025, this is not the bottleneck, this was going to happen anyway. The real focus must be on stimulating demand - Ethereum L2s are barely saturating 3 blobs for 15+ months, and for 3+ years the industry at large is severely oversupplying blockspace, and the gap is ever widening. Building an infinite ocean is useless if you have a swimming pool's water to fill it with. But hey, I've beaten that dead horse for 2-3 years, no one (or very few) cared.
https://polynya.medium.com/fanciful-endgame-5e760d427076
For my personal use, crypto is in a great place. Fees, speeds, wallet UX, applications are perfectly fine for my usecases, and they'll keep getting better. However, I'd like to see some things significantly improved, in order of importance:
- CRITICAL: Configurable withdrawal rate limiting in Safe & smart wallets and indeed, in all smart contracts, should be a standard. Without this, I significantly limit my usage of crypto
- Custodial standards for smart wallets
- Stage 2+ decentralisation for L2s, complete overhaul for Tron (normies receive USDT on Tron)
- EUR stables w/ liquidity
- Better frontends & standards
- More competition & options for minimal hardware wallets
- Range order management for Uniswap V4
- Significantly less volatile, reliable, long-term sustainable and high liquidity SoVs (BTC and ETH are far too erratic)
- Integration into widely adopted payment networks, both global and domestic
- Some inaccessible assets tokenised as RWAs
- Farcaster with bustling non-crypto content/discussions
Uniswap v4 feels like crypto's first complete product. Uniswap v3 was already nearly there, which is one reason why while previous versions took ~1 year, v4 took 4 years. I can see additional functionality (e.g. unified liquidity) and responding to infra changes (e.g. quantum-resistant zkVMs) that will justify a v5, but the core purpose of this application now feels complete for the foreseeable future. What's remaining is now on the governance & social end.
Aave v4 is the next candidate for a complete product. Interestingly, the applications are hitting final form years ahead of the infra they're built on, and in cases like Aave v4, being intertwined with infra itself.
(Side-note: of course, simple "products" like memecoins or USDT are "complete")
Blockspace demand/supply graph, the current situation and what's happening (S*)
What should happen: the industry must pivot completely to demand-side, instead of continuing to be all-in on the supply side
Checking in on my final blog post nearly one year on
My only regret is it was not alarmist enough, should have gone harder & more brutally honest
This was the most hate I've ever received, a little more wouldn't matter
Silver lining, the memecoin bubble burst before we got into truly violent territory and someone died (that we're aware of)
https://polynya.mirror.xyz/ptscXuh3J3KOj2uJAn0vrEanpn2nauwA7iytYZ4cM9U
I'd like to see more sophisticated rate-limiting features on Safe and competing multisig wallets. Is it possible via transaction guards? How can non-coder users access it?
So, for example:
- Can only send $10,000 via the Uniswap contract per 24 hours
- Can only send $1,000 to address X per 1 hour
Nothing else can be sent from the wallet. Changing these requires X/Y multisig permissions, and takes 7 days to execute.
Just an example, but the point is, the user can customise as per their needs. Rationale: User doesn't lose everything in one go if hacked - there's time and possibility to fix the situation.
Alas, in the 2+ years since, the vast majority in crypto have pivoted sharply towards full on greed, degeneracy, nihilism, and crime. Of course, there's a small minority that continues working on useful sustainability, and I have the utmost respect for them, but that keeps getting smaller by the day.
