The Cypherpunk Endgame & Real World Ethereum

Date

November 7, 2024

Author

DC Posch

Thanks to Liam Horne, Ansgar Dietrichs and David Ma for ideas and edits.

I vividly remember over a decade ago, in a coffee shop in Berlin, my first time using Bitcoin in the real world. The girl behind the counter cracked a smile, "oh..." and left to get someone, probably whoever put the BITCOIN ACCEPTED HERE chalk art on the wall. Apparently, I was the first person to ever try this on her shift.

The minute later, a guy comes out, thrilled. Someone’s actually doing it! It doesn’t work. He locks into debugging… five minutes later, I have a QR code to scan. I use a very rudimentary app (might’ve been blockchain.info, simpler times…) to send payment. Even after all that, it was an honor system transaction: with ten-minute blocks, I walked away with zero confirmations and a black iced coffee.

Bitcoin doesn’t actually work for this purpose. Ironically, the original and most basic purpose of blockchain—“peer-to-peer electronic cash”—turns out to be at the tip of a substantially large tech tree which we are just now reaching.

The core ideas of crypto are still just as captivating as they were 10 years ago: permissionless, unstoppable, global. The tech is unrecognizably better, to the point where the endgame is in sight. We have one-cent, one-second transactions, rapidly maturing rollups, smart contracts, stablecoins, much better consensus algorithms, scalable data availability. The pieces are all there.

1. This time is different

Crypto adoption happens in cycles. For better and worse, each cycle so far has been driven by a speculative bubble in a new kind of asset. There is a core of technologists pushing the underlying tech forward continuously. But the majority of people engaging with crypto are onboarded via these cycles.

The first in 2011-2013 centered around altcoins, meaning clones or near-clones of Bitcoin. The next in 2016-2017 centered on ICOs, many of them inspired by Ethereum with the token being an ERC-20. The most recent cycle in 2020-2022 started with defi summer and finished with NFTs.

This story is often visualized via a zoomed out price chart, where it’s easy to tell a sunny story about numbers going up and Bitcoin halvings and so on:

But this obscures actual usage, which is a less rosy chart. Getting an accurate census of real human crypto users is hard, so we’ll use people searching to install apps as a proxy:

There’s a lot going on here. The ICO cycle looks like a classic asset bubble, with a spike of new users followed by a fast drop back to zero. The most recent cycle is more complex. We see two spikes, one for defi summer and another for NFTs. Unlike defi project tokens, NFTs generally can’t be traded or held custodially, so Metamask installs spiked only for the second.

But then something really interesting happens. Unlike the other two apps, Binance is widely used in the developing world to hold and transact in stablecoins, especially USDT. We see that after the NFT bubble disappears, interest in Binance alone stays high. Comparing search volume for “binance usdt” confirms what’s going on here.

Our bet is that the next big chapter of Ethereum will center around stablecoins and real world use. This doesn’t mean people buying coffee with crypto. Instead, it will mostly be in higher-value areas where local cash doesn’t work, such as global USD savings, payroll, institutional adoption for cross-border fund transfers, international online checkout, and dollar-based crypto apps like Polymarket.

By "real world use", we mean anything that is non-self-referential to crypto. So Polymarket meets this definition, while using a stablecoin as a settlement asset for trading memecoins does not.

What's exciting is that this gives this next chapter a completely different shape than previous cycles. Speculative bubbles have an expiration date; you can see this in the usage charts below. Stables and real world use will feel more like a normal tech adoption S-curve.

A bit of snark about asset cycles.

The classic tech adoption S-curve.

This future has a lot of promise. It means Ethereum can grow toward its final form as a global settlement layer. It means more revenue-generating businesses and fewer token launches. It means we can attract great technologists to work on the application layer, not just the core protocol. A real adoption curve with sticky, non-self-referential use cases is a very attractive proposition and will drive a new level of investment and development.

It also sets the stage for Ethereum's biggest challenge...

2. Tradfi is coming

Liam Horne wrote about the growth of stablecoins a year ago. At the heart was the idea that financial apps with lots of users will want their own stablecoin, and potentially also their own chain to go with it.

Since then, we've seen many examples of this type of adoption. Revolut joined PayPal in announcing their own coin. Stripe re-enabled crypto payments and bought Bridge.xyz, a stablecoin on- and offramp company. And earlier this week, Robinhood, Kraken and others launched a joint coin called USDG, "Global Dollar".

This is driven a two key factors:

• Stablecoin yield. The market leaders, Circle and Tether, each earn around 5% on a combined $150b in stablecoin market cap. Tradfi sees this →
• Geographic and regulatory advantages. For example, stablecoins may let those companies onboard users in countries where they have no fiat presence, or offer their existing userbase new products and better interoperability.

These advantages, combined with the new fertile soil of post-4844 L2 Ethereum, will make these companies onboard many millions of users to Ethereum. There are big open questions around how much freedom those users will have. What can we learn from history?

3. What happens to cypherpunk ideas when they scale?

Ethereum is primed for sustainable adoption. Big, capable companies are investing serious effort into building polished products. So what happens next?

There are clear trends for what happens to liberatory technology at scale, meaning technology that enables people to do things they weren’t previously allowed to do. To be clear, these is not what we think “should” happen, but rather what empirically has happened.

The broad strokes are invention → excited early adopters → legal conflict with incumbent interests → fitful growth as these conflicts are waged and sorted out → and finally, mass adoption. This last stage often happens via a product quite different from the initial invention, built for maximum convenience and embodying some kind of stable compromise to settle the conflict.

Three examples:

1. End-to-end encrypted messaging. The original dream of secure and private conversations over the internet was pioneered by PGP, an encrypted email standard. In the 1990s, encryption was still considered a military technology, and the American government tried to prosecute the authors of PGP under arms-export laws. The authors responded by publishing the source code as a printed book, then successfully beat the court case on First Amendment grounds.


Later, Signal built a much more user-friendly encrypted chat app. It onboarded orders of magnitude more users than PGP. Saw significant growth after the Snowden leaks increased awareness of mass surveillance, but topped out in the tens of millions of users.

Finally, the end-stage product that brought end-to-end encryption to billions of users was Whatsapp, followed by iMessage. Users like the idea of privacy but they really don’t like it when they lose a phone and all their old messages are gone. As a result, both of those apps enable message backup by default! This is end-to-end-to-server encryption, defeating the main idea of the original concept. Even if you personally opt out of backup each time you’re prompted, most people you talk to are not doing that, so your messages are still uploaded by them. E2E is reduced to a security engineering upgrade in the vein of “SSL added and removed here :)”

The silver lining is that Signal is still quite popular and available for whoever needs it. The less-silver lining is that none of these are interoperable, they are each their own closed network.

2. File sharing and online music. Did you know that Daniel Ek, the founder of Spotify, was previously the CEO of uTorrent? Online music is a crisp example of the trend. The original invention was Napster, which inspired the idea that you can listen to any song, on demand, over the internet—a revolutionary idea at the time. It was shut down. A decentralized successor, Bittorrent, couldn’t be shut down. The messy conflict that followed saw many related centralized services like torrent search engines targeted, VPNs popularized, individual users sued, and so on. The grand compromise was Spotify and similar streaming services: convenience maximized, user choice limited, incumbents partially paid, and permissioned. A beautiful walled garden.
3. Uber and ride share. This is further removed from crypto, but still a fascinating example of an app that created a new and useful freedom, and then fought a protracted conflict against incumbent interests. Here the rider side of the app survived largely intact: the basic action of opening the app and dropping a pin works just as it did at launch. The driver side has been transformed, from a P2P “ride share” concept, giving someone a ride on your way home, to something professionalized and specially regulated. In some jurisdictions like Germany, Uber lost the main regulatory battle, and the app is simply a UI for calling a taxi. In many others, like New York, Uber defeated the taxi cartel by cutting a compromise with the state involving licensing and taxation.
4. Peer-to-peer electronic cash. Early attempts were centralized and received the Napster outcome. Bitcoin was the first successful decentralized implementation and couldn’t be shut down; this expanded the Overton window of what was permissible. However, Bitcoin largely failed as a method of payment because you’re doing a slow, expensive transaction in a volatile asset. Now with stablecoins on rollups, we have fast cheap transactions in a stable asset. The big compromises so far are centralized stablecoins (USDC beats RAI on adoption by ~1000x TVL, because it’s profitable while RAI carries costs), lack of privacy, and restrictive on- and offramp. However, the conflicts are just beginning. What the final mass-market products here look like is an open question.

Empirically, the most popular so far are custodial and focus on making on- and offramp as easy as is locally possible. In Argentina, for example, the “Spotify of Ethereum” so far is Lemon. You can hold aaveUSDC for +6% yield and offramp it directly into purchases by tapping your Lemon card. In more restrictive countries with worse banking, the winner so far is Binance with BinanceP2P for unblockable on- and offramp with minimal dependence on tradfi.

4. What are we trying to preserve?

Looking at all of those examples… if history is any guide, the product of billion-person Ethereum is likely to be radically different from where we started.

First, a testable prediction (+ tough pill for some in Ethereum to swallow): the majority of those users will be on custodial apps. How do we know? We talked to a hundreds of crypto and stablecoin users across many countries. We originally started Daimo with the vision of making self-custody really easy to use, only to find that outside niche scenarios, it’s not something people want. “Custody” is a verb, and no matter how smooth you make it it’s still work and responsibility that most people do not desire! To quote a great thinker, “the definition of sanity is dealing with reality on reality’s terms”.

Self-custody is still a critical capability even if only a minority of users use it, just like free speech is a important right even if only a minority of people are publishing contrarian ideas. The fact that Ethereum is an open ecosystem where users can easily move to other custodial apps, or self-custody if needed, is a key ingredient in promoting competition and preserving our values.

Furthermore, most users will think in dollars or local currency, which will move frictionlessly and globally as stablecoins. So we have a world secured by Ethereum, but where most users are on custodial apps built on stablecoin rails.

In this world, what does it look like for Ethereum to win?

One principle is that we want end users to benefit from Ethereum’s values and permissionless nature, not just PayPal and Revolut. The average user may be happy to live inside the garden, but anyone who wants to open the gate should be able to do that. Interoperability is key. It’s good for users, and it’s critical for app builders: we need a vibrant ecosystem where you can ship a new idea on Ethereum and anyone else can come and use it.

The Ethereum community put a lot of effort into securing the base layer, rollups, contacts, and other infrastructure. Now we need to secure the application layer, too. Here are some freedoms we want to preserve for Ethereum users:

For custodial wallets: can you escape the garden?

Big question is, to what extent will all these new users be able to do things on Ethereum outside of the app that brought them?

With a custodial app, you inherently don’t have privacy from the custodian, but it’s actually easier to achieve privacy from the block-explorer-viewing public. Custodial apps also carry the potential for excessive restrictions and selective non-interopability; Ethereum cannot prevent this, but we can expose it to confer brand + legitimacy advantages to apps that respect user freedom.

With intent addresses, as long as an app lets you send assets to an address, we can accomplish arbitrary on-chain actions. Ethereum is sufficiently general that “transfers are all you need”.

But even that, we don’t always get. For example, Venmo has a crypto tab, but all you can do is buy and sell. (Venmo does theoretically let you withdraw to an address, but it’s L1 only and asks for a second layer of KYC so invasive that it’s clear they don’t want users doing this. We wonder if they added the capability just for legal reasons.)

The gold standard is an app that lets you send assets to any address on at least one rollup, instantly. From there, we can efficiently accomplish anything on L2 ethereum.

It’s easy to take this for granted because crypto-native custodial apps all let you do this: Coinbase, Binance, Lemon, and so on. When told to curtail basic capabilities like swaps or asset transfers, these companies are willing to fight hard because it’s existential for them.

Tradfi-on-Ethereum has different incentives. Companies like Robinhood (let alone banks) have large existing businesses with significant regulatory exposure. If faced with pressure to nerf their new crypto offerings, they might well give in. One open question for the community is, “what are the strongest incentives we can put in the other direction, to ensure user choice?”

For assets: are they permissionless, or at least permissionless-by-default?

There are three levels:

There’s a very important design space for stable assets beyond RWA-backed, blacklistable stablecoins. Can we get sustainable yield-bearing coins? (This could happen via just a regulatory change, allowing a coin with the economics of USDY and the permissionless interoperability of USDC.)

More importantly, can we get stable assets without a central point of failure?

One tailwind is that ETH itself is getting less volatile. The 2017 cycle ended in a >90% price drop; the latest, “only” 80%. If future growth is driven by real world adoption, it might get much less volatile still. This makes constructions like RAI/LUSD more viable, since the amount of risk you have to hedge away is lower, and also makes ETH more directly useful as money. “The ticker is ETH.”

For chains: is it a real rollup with permissionless exit?

This one has already been covered in depth, Stage 2 or bust.

Danny Ryan has a nice line about rollups that I think generalizes to the rest of Real World Ethereum as well:

It's not obvious that L2s will by default retain Ethereum's brand, value, and soul.

The Ethereum community today holds immense power in setting the norms that will define how this L2 movement plays out for decades to come.

We must ensure that L2s not only inherit Ethereum’s security but also its legitimacy.

— from Reflections 2023

—

If we’re going to have a wave of apps + stablecoins + chains onboarding tons of new users onto Ethereum, then we need high standards for all three.

The net result is a beautiful progressive onboarding pipeline. We get millions benefitting from instant global settlement, often without even knowing they’re using crypto—but if they want to go deeper, they have an always-on asset they can send to any address, where they have full access to the Ethereum ecosystem and unbeatable security guarantees.

5. Some risks.

Here are a few of the “Black Mirror” bad scenarios.

“Tradfi embrace / extend / extinguish.”

Fintech, from Stripe to Robinhood to Vanguard, increasingly runs on Ethereum. Initially, this is celebrated and everyone is happy. Over time, regulatory pressure (perhaps driven by regulatory capture / old banks trying to protect their moat) forces the apps, coins and chains involves to become increasingly permissioned.

Maybe withdrawing to an address from your Robinhood requires attesting that you own that address, or if not, providing KYC info for who does. Maybe Circle is forced to use the USDC blacklist a lot more often than they do today.

On the far end, maybe we get rollups with an address-freezing feature or even a KYC-driven address whitelist. On the milder end, we get a professionalized, regulated version of what's already happening at scale with USDT+Tron: digital dollars on a centralized chain.

“Just enough decentralization for regulatory arb, not one bit more.”

New tradfi-adjacent chains stay at Stage 0 indefinitely, and the new friendlier regulatory environment (in the US, anyway) removes a sense of urgency for them to actually decentralize. This exposes the ecosystem to tail risks, both regulatory and otherwise. We get situations like this one from earlier this week, but at much larger scale.

The new centralized rollups carry layers of mostly hidden risk—technical (in the story above, 54 hours of downtime?), counterparty (what if the RaaS operator goes out of business?), and regulatory (if you can do it, they can make you do it). All is fine for a while.

One day deep into the Great Stablecoin Upgrade of international finance, we get an FTX moment but rather than an exchange, it’s a chain.

—

How do we track these kinds of tail risks? L2Beat is a great start. We may need something more publicly legible to mark unsafe rollups explicitly as Not Ethereum before risks materialize.

"Total government victory"

Governments, in particular the US and maybe China, end up balkanizing and controlling crypto. This could happen via front door (CBDCs) or back door—start with "tradfi embrace/extend/extinguish", then simply exercise greater control over crypto fintech companies until “USDC=CBDC”, KYC is effectively mandatory everywhere, and so on.

GM. Yesterday, an official with the US Treasury Office of Tax Policy announced at the American Bar Association Virtual Fall Tax Meeting that IRS will finalize this year additional regulations requiring tax reporting from developers of noncustodial software who would under these regs be deemed brokers. […] — @BillHughesDC

…we don’t know how close this particular proposal was to becoming law, but it’s a great example of the type of policy we’d see under a “total government victory” failure mode.

Given the recent election, this issue may be reduced in the short term… but it might be an even bigger tail risk in the long term, since crypto now has an explicit partisan valence in the most impactful regulatory jurisdiction.

Meanwhile all over the world, governments are doing early stage experiments with CBDCs and “govchains”, like this one in Turkey.

—

The best mitigation might simply be speed. If we ship the right way and grow fast enough, we can set irrevocable precedent, just like cities with Uber and Lyft are never going back to taxis.

6. Conclusion + what needs to happen today?

We have an incredible opportunity with the stablecoin wave. It’s going to be completely different from past asset bubble cycles, and it will feel more like the standard technology adoption curve.

What do we do today? Well one thing, of course, is to build whatever products are complementary. If 100 million people are about to get fingertip access to stablecoins via an app already in their pocket, what net-new things can they do with this? Our mission is to make things people want.

This an opportunity to serve a big market that not many people see coming, and to advance the plot of Ethereum. We are working on Daimo Pay for exactly this reason.

Beyond that, here are some possible priorities:

1. Write down our principles and espouse them often. It's a political movement: self-custody, privacy, sovereignty, and a permissionless ecosystem where anyone can ship to the world.
2. The industry needs to be serious about self-policing. L2Beat and Bluechip are amazing and we need more of them.

For one, we need an L2Beat for self-custody. There are a lot of wallets claiming to be self-custodial when they are simply not. There are important nuances to separate, but that’s for a future post. There are popular embedded wallets today that claim self-custody but also can recover a user’s account with just SMS auth…this is clearly custodial.

Allowing fake self custody to slide risks both user funds and Ethereum’s brand.

We also need an L2Beat for custodial apps. This should measure interoperability and important safety concerns like proof of reserves.

3. Help regulators regulate. Write sane policy for lobbyists, based on (1) and (2), and get it included in major jurisdictions.

The US is about to become much crypto friendlier. This is a great opportunity to engage, pass thoughtful regulation, and set good precedents.

4. Make it profitable to join us. Don't fight fintech and banks, bring them into the fold as long as they’re reasonably aligned, interoperable and respect user choice.

We are very excited for Real World Ethereum. It’s going to accelerate because the fundamentals of ethereum and stablecoins are just excellent—the stablecoins generate sustainable yield, the chains confer worldwide reach, the apps allow permissionless innovation on top of instant and borderless assets.

The risk of capture is real. But if we play our cards right, we can turn this into a massive growth opportunity for the whole open ecosystem as well as a liberatory event where a lot of the world gains access to the things that make Ethereum special.