Author: Justin Bons , founder of Cyber Capital, compiled by Followin
Most L2s will always remain centralized because their incentives are perverse!
Competing L1 and L2 are cannibalizing ETH’s user base, while Ethereum leadership is promoting and celebrating ETH’s downfall. This is a sad state of affairs, as it betrays the founding principles they once claimed to hold dear. Promoting centralized solutions while empowering corporations that are forced to comply with government censorship.
Privacy has been one of the cornerstones of the cypherpunk movement, as cryptography brought the promise of widespread use of privacy-enhancing technologies. Despite this tradition, ETH is pushing most users toward L2, which can monitor, freeze, steal, and censor your funds; this is clearly far from the ideals of the old cypherpunks. Going down the same self-destructive path as BTC, abandoning on-chain scaling in favor of L2. History is actually repeating itself.
L2 Centralization
The reality today is that all major L2s are completely centralized and can censor and steal user funds. Admin keys controlled via multisigs can change contract rules (including theft), and centralized orderers can now censor anything as well.
More important, however, are the potential paths to change. This is where things start to get really bad, as all proposed solutions to L2 centralization are overly optimistic and require for-profit companies to give up a lot of their current revenues…
This completely ignores human nature and history, a common mistake made by senior engineers and computer scientists. This is why studying blockchains must be multidisciplinary, including the humanities. This is precisely because the criticisms of ETH's proposed solutions are not technical, but point to the difficult social coordination problems inherent in these proposed solutions.
Decentralization requires powerful parties to give up their power. Historically, this has rarely happened because it goes against their incentives. Sometimes, outstanding people do the right thing. But on average, especially when looking at large groups of people, we should always bet on incentives because that is more predictive of the masses.
This is why I don’t think most L2s will be decentralized. Since the incentives are clearly pointing towards L2s staying centralized, it’s not enough to just say “trust me, bro”, especially when we should verify rather than trust.
Drake's rebuttal
Changing which part of the system collects these revenues is also not a suitable solution, as Ethereum Foundation Core Researcher @drakefjustin recently tried to do by placing Base’s revenue in the execution layer instead of the sorter layer. This is because for Base to be truly “decentralized”, all revenue must be sacrificed; as Drake implies here, keeping execution centralized is not a suitable solution at all.
The harsh truth is that Coinbase will likely never be decentralized, and that’s exactly what the “L2 scaling” roadmap is all about! Users capitulating to centralized and inherently custodial solutions has crushed the original vision under the weight of KYC, AML, and institutional-grade scrutiny.
L2 Interoperability
L2 will always fight against universal interoperability protocols, trying to get everyone to adopt their own solutions, even if it hurts their long-term success. This is similar to the tragedy of the commons problem in political science. Twenty or so attempts to establish a unified interoperability protocol is equivalent to no unified interoperability protocol at all!
L2s compete with each other and with L1s themselves, creating competing ecosystems instead of a single ecosystem, unlike L1 scaling. The free market will continue to create a variety of competing L2s. This represents various power groups that don’t always get along. This dynamic is good in most cases, but for blockchain scaling, it can only bring massive fragmentation that ruins the user experience. The idea that everyone will use those seamless interoperability protocols. And the custodians will pack up and move on to more advanced technologies… is a fantasy and does not represent how the free market actually works, because in that environment there will always be custodians and centralized L2s.
Ironically, as ETH Core pushes towards a fixed/based L1 collator, L2s are pushing for their own “shared collators” like Arbitrum’s Superchain, Polygon’s Agglayer, etc. The only way a “shared collator” would work is if we all used the same collator, which is not feasible. It is unrealistic to expect these major L2s to abandon their efforts to “solve interoperability”. The same goes for Eigenlayer and other restaking platforms, as they also implement collator-like functionality. This all makes true shared collators completely unusable to start with, as they are mostly greedy fantasies. Driven by the idea that if everyone uses the same L2, this will solve the UX problem! Technically true, but practically wrong.
This is why fragmentation and composability breaks between L2s will never be solved. Interoperability between L1s remains unsolved to this day for the same reason. However, at least under this paradigm, L1s are not artificially bottlenecked by this toxic L2 narrative. This is why my problem is not with L2s per se, but with the lack of expansion of L1s, even if that is a result of L2 lobbying.
Economic security
The result of this shift away from actually using ETH is decline and death, as cryptocurrencies live on economic security, regardless of @aeyakovenko's claims that the ETH community is just a meme. The bottom line is always revenue, and it's obvious that hosting your own used chain will always earn more revenue in the long run than outsourcing all used chains, because that's what ETH is doing now, which explains why this is such a crazy bad move from every possible perspective!
Perverse Incentives
Now let’s talk about the elephant in the room: there is orders of magnitude more funding for L2 than there is for L1 in ETH and BTC. Billions of dollars have been created around L2 tokens and VC funding, while only millions have gone into L1 development. This creates a clear conflict of interest and potentially outright corruption. With the incentives so perverse, it could lead to developers arbitrarily limiting L1 capacity in favor of L2. All they have to do is not pursue or support L1 scaling technology…
This is why L2 has become the biggest corrupting force in this industry. Because they benefit from not scaling L1 in the short term. Turning developers into multi-millionaires through L2 tokens and equity. Of course, this also adds a strong bias towards L2 scaling over L1 scaling. This is because L2 makes more money by supporting the narrative of limiting L1 capacity, while supporting scaling through L2; this creates a clear conflict of interest between the long-term success of L1 (ETH and BTC) and the short-term profits of companies focused on L2.
This is also because VCs can rent-seek through “L2 scaling” because these are generally for-profit businesses, while L1 scaling is a public good. It is impossible for VCs to extract a percentage of fees from a well-designed L1. However, this is the current state of affairs in the L2 world. Scaling L1 will not benefit these VCs in the short term, while the “L2 scaling” roadmap will benefit these VCs, even if in the long run, it sows the seeds for ETH’s self-destruction.
L1 Scalability
Both views share a core assumption, which is L1 scalability. ETH’s position depends on whether the L1 scalability tradeoff is tenable. Therefore, it is this technical limitation that justifies the “L2 scaling” roadmap in their minds.
The L1 scaling paradigm is far more optimistic, as it acknowledges that L1 can now scale to meet demand without sacrificing decentralization. Whether through pure parallelization, DAG, or sharding, all roads lead to Rome. The ETH community is ideologically attached to an outdated technological paradigm, much like Bitcoin supporters. ETH is also rapidly becoming a dinosaur, just like Bitcoin, with the same strangely toxic and cult-like ideological trappings.
ETH Maximalism
It’s no coincidence that ETH supporters have gradually become indistinguishable from Bitcoin maximalists, as they have adopted the same philosophy and narrative as a coping mechanism/belief system.
Precisely because this is all a result of the same systemic flaws in the governance structure that led to this situation with BTC and ETH in the first place. Environmental pressures thus create a particular type of belief system, much like convergent evolution in the biological sense. I am also convinced that if formal on-chain governance were implemented, not expanding L1 would never be seen as a realistic option.
Governance
Ultimately, it comes down to “who decides”. The ugly reality is that a relatively small group of people can decide BTC and ETH. This is the essence of “off-chain governance”; a highly centralized decision-making process. This can be captured by small groups with bad incentives (such as for-profit L2) that directly benefit from not scaling L1 in the short to medium term.
On-chain governance allows all stakeholders to vote on proposals in a completely transparent process, which can of course lead to very different outcomes. Most importantly, this benefits L1 rather than the interests of whichever group happens to capture the centralized governance process at the time.
From a political science and philosophical perspective, these off-chain governance processes are often easy to capture and distort, as the “GitHub dictatorship” is far less powerful than a nation-state. On the other hand, on-chain governance processes with a large number of stakeholders, coupled with more complex checks and balances and divisions of power, do have a chance of withstanding the test of time and the worst that our human nature can offer.
This is why on-chain and governance must be viewed as a mechanism to preserve decentralization, rather than a repetition of old-style/legacy governance. In fact, it’s quite the opposite; off-chain governance replicates governance systems that predate blockchains; which, I might add, is done very poorly in most cases. On-chain governance is something entirely new, one that relies on the inherent strengths of blockchain technology and aligns with L1 and collective decision-making. It’s no surprise, then, that the idea was outright rejected by BTC and ETH’s leadership. Whoever has the most influence has the most to lose if on-chain governance is implemented; which is why the incentives are stacked against its establishment if it’s not established early on.
Real Solutions
The solution lies in abandoning ETH and voting with our feet in favor of its scalable competitors because as stakeholders we have no real say in ETH’s governance process.
We can certainly admire the effort to lead a full-scale rebellion against the status quo in ETH, similar to the block size debate in BTC. However, as a veteran of that civil war, and on the “losing side” (big blocks) at the time, the odds are not good. Because at the time, most businesses, miners, stakeholders, and users supported larger blocks. Yet, the core developers still prevailed, and 8 years later, the block size limit is still 1MB!
Even stronger evidence that effective centralized control over the rules of a decentralized network is not even theoretically possible. ETH has nowhere near the same level of support for the revolution as BTC, so I don’t see how it can succeed, especially without formal on-chain governance.
In this free market of cryptocurrencies, we have to take into account another strong demographic effect: those who supported L1 scaling left ETH, and those who did not eventually joined it. Who is left to fight for L1 scaling now? The same effect happened to BTC, turning it into a monoculture with no real potential for change. All these shifts start from the top of the leadership structure, gradually moving the entire ecosystem away from its original goals.
We used to believe in “fork governance”, but this was wrong for two reasons: the threshold for “agree or fork” was too high, so it evolved into an effective tyranny. The second problem has to do with the fact that the market does not actually bypass the problematic chain through forks, but chooses the next generation chain. This explains why the market did not bypass BTC through BCH at the time, but eventually upgraded and all turned to ETH.
History repeats itself
I went from being a die-hard Bitcoin supporter in 2013 to being a Bitcoin alarmist in 2015, but by 2017 I was a detractor.
I gave up BTC and believed in ETH's promise of on-chain expansion through sharding. In 2015, I became a hardcore Bitcoin supporter. In 2022, I became a Bitcoin supporter who sounded the alarm. But by 2024, I had become a comprehensive critic of Bitcoin.
You can comment on my stance all you want, but one thing is clear: despite our protests, BTC and ETH have changed under my watch, and I have been very persistent. Radically changing the economics and purpose of blockchains by arbitrarily limiting their capacity is radical and the exact opposite of conservative; we should not allow them to use "conservatism" or "social contract" as excuses, because these principles have been completely violated.
The real tragedy is that we have squandered opportunities for global adoption twice, quite possibly setting us back decades. Thankfully, we can clearly identify the problems and implement solutions in the latest generation of blockchains to finally break this terrible and painful cycle.
in conclusion
Let’s go back to the first solution and why ETH is doomed to fail. Because for the sake of decentralization and the cypherpunk dream, we must vote with our feet and support ETH’s competitors.
If you really like Ethereum and Bitcoin, you have to be able to let go for them, for their original vision. Precisely because this is far more important than the price of any three-letter ticker. Focusing on the big picture is focusing on the biggest prize:
Change the world with financial sovereignty, censorship resistance, and true monetary independence!
