Can Ethereum Win Again? (Part I)
Three Things #166: April 20, 2025
I’ve been an active member of the Ethereum community for almost eight years. I worked at the Ethereum Foundation in various capacities from late 2017 until early-mid 2019. At the time I was, and I remain, very invested in the Ethereum project and community, socially as well as economically. I want to see Ethereum succeed.
For the longest time, Ethereum had a massive head start over its competition, and no one could hold a candle to Ethereum tech or the Ethereum community. But Ethereum’s lead was always shaky because, while the optics were good—high market cap, lots of TVL, full blocks, etc.—the fundamentals were always troubling. Ethereum has never had many useful apps outside of DeFi and stable coins, nor many users doing things other than trading and speculating. Ethereum had a rather large window of opportunity to capitalize on its early success, at least five years, but that window is rapidly closing. It may already have closed.
Despite being so invested in the success of Ethereum and the community, or perhaps because of it, I find this topic especially difficult to write about. For one thing, I have a lot to say about Ethereum, about what’s working and what isn’t, especially relative to other crypto ecosystems. I’ve thought about this probably more than any other topic over these past eight years.
It’s also hard to write about this topic because I can’t do so without being honest and critical of many of the things that are going on today. After parting ways with EF in 2019, I was very tempted to share my critiques of the Foundation and of Ethereum more generally, but I decided not to because it was too fresh, and too soon, and because I knew that I needed the benefit of time and some distance. Now is the time.
It’s not just a matter of time. It’s also because of how Ethereum is doing lately, and it’s not doing well by a number of metrics. Ethereum the asset has severely underperformed in recent months. ETHBTC is near an all-time low, and has been declining nonstop for years. Sentiment is the same.
I’ve spoken to a large number of “OG” friends who have been a part of Ethereum as long as I have, and I’ve never seen sentiment so low before, not in all of my years in this community. Recently I’ve been unable to find a single person who’s still truly bullish on Ethereum’s future. We’ve been through multiple market cycles, we survived governance crises and the collapses of FTX and Terra/Luna and the deep, dark multi-year bear market that followed, but sentiment today is lower than at any of those times. What’s more, many of my OG friends, like me, feel uncomfortable speaking up and speaking to the things that need to change, which is also an indication of how wrong things are in Ethereum today.
I still think Ethereum has a chance to make a comeback, but that chance is shrinking every day. Here are some of the things that are wrong, and some of the things that need to change, as I see them. This is the first part in what will surely be a series, as it will take a few weeks to get through the whole list.
Thing #1: Research Fetish 🔬
If I had to pick one thing that Ethereum has gotten wrong, it’s putting the research cart before the application horse. Since the very beginning, Ethereum culture, the Ethereum Foundation, and the most powerful, influential people in Ethereum have fetishized research at the expense of many other important things, including applications, usability, and simplicity. The people in Ethereum who get the lion’s share of attention and admiration are the researchers. They get the prime speaking slots and podcast interview slots. They’re top dogs within the Foundation, given almost completely free reign: to work on what they want, to travel and speak about what they want, etc.. Research-heavy projects get the lion’s share of grants and funding. I say this as a former Ethereum researcher—it was great at the time, but I see now how and why this fetishization is a problem.
To some extent admiration for research and researchers is appropriate. Research is essential, and Ethereum wouldn’t be Ethereum without research! Ethereum exists and has been successful because of groundbreaking work on consensus, VM design, ZK proofs and succinctness, optimistic rollups, P2P networking, data structures and database design, security, and many, many other areas. Over the years hundreds, maybe thousands of high quality research papers have come out of, and been funded by, the Ethereum community. Entire journals, conferences, and research institutes exist because of Ethereum. The vast majority of this is a public good: the code is open source, the papers are accessible for free on arXiv, and the talks are available for free on Youtube.
The problem is that Ethereum isn’t just a research project. There are hundreds of billions of dollars of actual assets and application value at stake, and this research comes at a cost. The people designing Ethereum are not application builders. They’ve never built a production application on Ethereum, and most of them couldn’t to save their lives. They have very little understanding of or appreciation for what things are like for builders: e.g., how hard it is to build modern, safe, usable applications in Solidity, on Ethereum, or how difficult key management can be. They don’t appreciate the downstream consequences on UX and the application layer of their designs. For a good example of what I’m talking about, see this recent tweet and the replies.
While researchers are valorized, product people play second fiddle in Ethereum. There were zero product people at EF when I was there, and I suspect there are still none. There’s little to no funding available in the ecosystem for product people. Most teams building on Ethereum cargo cult the behaviors they see modeled by EF, and also don’t hire professional product people. And real product people are anyway turned off by crypto because of its horrible reputation, and because of how difficult it is to build usable products and experiences, even after all of these years.
It didn’t have to be this way. All of the UX issues we face, from slow time to finality, to the cold start problem and needing a special gas token, to key management, we tend to take for granted because things have always been this way, but they didn’t have to be this way. We could have consciously designed Ethereum to be better, to be more usable, and to enable more usable applications to be built more easily. We could still choose to do that. As a few concrete examples, native account abstraction, as deployed by blockchains including NEAR and Spacemesh, addresses many of these issues, such as the cold start problem; it also makes key management much easier. There are now blockchains that use stablecoins as their native gas tokens, which makes gas fees less volatile and helps address the cold start problem. And newer chains have much, much faster time to finality, which has a positive feedback effect on UX up the stack.
All of these things will be very difficult to change today, because they’re deeply embedded in the Ethereum design. It would take a radical overhaul of the Ethereum protocol to address them, but more recent projects show what’s possible. As hard as the tech is, social change is even harder. Really solving this problem would require a radical cultural shift, ensuring that product people are lionized as much as or more than researchers. It would also require VCs to change their toxic behavior, and to stop prioritizing funding tokens that they can dump on retail over actual, usable products that won’t have immediate liquidity. This would in turn require retail investors to stop speculating in useless shitcoins.
In other words, the problem is complex and multifaceted. It’s a complex coordination problem and it’s unclear how to fix it, without more or less starting over from a clean slate. Nevertheless, I have faith in the Ethereum community’s ability to solve coordination problems and I’m confident that, if the appetite exists, we can find a way to solve this one too.
Thing #2: Ideological Purity 🕊
Ethereum was very much on my mind when I wrote recently about the risks of ideological purism. We can look at Bitcoin, Ethereum, Solana, and other chains on a spectrum which is ideologically pure at one end, and totally pragmatic at the other. Bitcoin is the most ideologically pure project in the world, and its values are cypherpunk values. It hasn’t compromised on its values and priorities, such as freedom, personal responsibility, decentralization, censorship resistance, sound money, and security. But as a result it’s also the slowest to evolve and the slowest to add new features. There’s simply no way for Bitcoin to add new features without hard forking and risking breaking what’s already working, which is counter to its values. Another consequence is that Bitcoin also has the lowest transaction throughput, and probably always will.
Then there’s Ethereum. Ethereum is definitely more pragmatic than Bitcoin, and also more useful. It has a substantially higher transaction throughput capacity than Bitcoin. There’s a reason Ethereum attracted the lion’s share of developer attention for many years after it launched: you can build much more interesting things much more easily on Ethereum than you can on Bitcoin.
In order to achieve this degree of usefulness, Ethereum compromised on some of Bitcoin’s values. In particular, because its VM is so much more complex than the Bitcoin VM, the attack surface is also greater. Ethereum has experienced many more bugs and exploits during its history than Bitcoin has, so in a sense it’s less secure than Bitcoin. But the network has never gone down. As it’s become harder and harder to run an Ethereum node, the number of Ethereum nodes has gone down relative to the number of Bitcoin nodes, so Ethereum is also less decentralized than Bitcoin, though it continues to value decentralization highly and in an absolute sense the network is still quite decentralized.
Finally, at the opposite end of the spectrum from Bitcoin we find Solana (and many other new, high throughput chains). Solana is far more pragmatic, and far more focused on shipping usable applications, than Bitcoin or Ethereum. It has a much higher transaction throughput than Ethereum. There’s a price to pay for this pragmatism. Solana pushed the technical envelope into experimental territory, to the point where the network has crashed periodically, requiring manual intervention to restart (developers call this “testing in prod”). It’s quite hard and expensive to run a Solana validator, meaning the network is much less decentralized than Ethereum. Solana is also more complex than even Ethereum, which means its attack surface is even greater than Ethereum’s. But this relentless focus on pragmatism and applications means that Solana tends to attract a certain kind of product-focused builder, one that’s less purist and less ideological, and instead more focused on shipping things people actually want to use. This is very much the Solana culture.
To be clear, I don’t mean to suggest that there’s a right or wrong stance here. Each project has optimized for a different set of values and priorities, and it’s good that each project exists at a different point in the tradeoff space. But it’s important to recognize the tradeoffs associated with each point on the spectrum.
As I wrote two weeks ago, there are severe risks associated with the sort of ideological purism espoused by projects like Bitcoin and Ethereum. It severely limits the degree of innovation that can happen at the base layer, or in the ecosystem. Changes take longer to happen, if they ever get done at all.
We can see all of this playing out in Ethereum today. The reality today is that, while it had the lead for a long time, Ethereum has fallen quite far behind the competition technologically. Ethereum’s block time is 12 seconds, compared to the sub-second block times that are common today among modern smart contract chains. Its finality takes on the order of minutes, whereas other chains today offer finality on the order of seconds. There’s no shortage of technical talent in the Ethereum ecosystem, and Ethereum could have these things and more if it wanted to. It’s a question of desire, and of priorities, and of what Ethereum is or isn’t willing to sacrifice to achieve these things.
One stumbling block for Ethereum is its relentless focus on multiple client implementations. There are twelve production clients that need to implement every proposed change, and if one client team takes a while to implement a change, or refuses to, it slows down the entire process. If you factor in the connections between consensus and execution clients, there are 36 combinations to test for every change.
This is one of the main reasons Ethereum only manages to ship one network upgrade per year, while other chains like Solana constantly ship updates. Client diversity is useful to prevent monoculture and makes the network more robust against certain kinds of bugs or attacks, but Solana has proven that even a network that periodically crashes and then requires a manual reboot is still economically viable. This is a choice that the Ethereum community has made, and it has consequences. I can’t overemphasize how different the Solana and Ethereum approaches to upgrades are: they are complete opposites, and this speaks volumes about the different cultures.
Ethereum still has a window of opportunity to catch up, but it won’t be able to as long as it puts ideological purity above all else as it always has. Historically, the Ethereum community has been totally unwilling to compromise on values like decentralization, censorship resistance, or client diversity, even a little bit. The focus on decentralization means there’s still a requirement that it be possible to run a full Ethereum node on cheap, commodity grade hardware. Other networks have achieved much greater scale, throughput, and usability by relaxing these requirements at some cost to decentralization. It’s harder to run a Solana validator than it is to run an Ethereum validator, and there are fewer of them than there are of Ethereum nodes. Whether Solana is sufficiently decentralized is a difficult question to answer, and only time will ultimately tell.
It’s my opinion that Ethereum has succeeded thus far in spite of, and not because of, its values, mainly because cypherpunk values like decentralization and censorship resistance simply aren’t valued by many people. I don’t believe that Ethereum can both continue to be a market leader and also continue to be totally uncompromising in its values.
The Ethereum community thus has a difficult decision to make. It’s a question of ambition and vision. Bitcoin doesn’t need to innovate as rapidly as Ethereum because its ambition and vision are much more limited than Ethereum’s. The question for Ethereans is, are you happy with Ethereum the way it is? Will you be happy if Ethereum innovation and development continue at their present slow pace, or if they slow further? What’s your vision and ambition for Ethereum? The answer is unclear and I’m not sure which way the community will go.
Of course, these things aren’t entirely black or white. There are creative ways to run experiments more rapidly that the community should seriously consider, such as launching multiple chains (consider Litecoin vis-a-vis Bitcoin, and Kusama vis-a-vis Polkadot). The point is that we should be having this serious conversation today and everything should be on the table. We should be thinking far more outside the box than we currently are. Ethereum has reinvented itself before, and we can do it again, but only if we really want to.
Which leads us to the next thing.
Thing #3: Growing Up 🪴
Ethereum feels a bit like a talented teenager with high potential. It can be anything it wants to be, but it can’t be everything. One of the things holding Ethereum back is that it isn’t clear what it wants to be, for whom.
It’s been this way more or less since the beginning. It’s most obvious in the constantly changing narrative. Ethereum went from being a “world computer,” to a platform for ICOs, to a DeFi chain, to an NFT platform, to an ultra-scalable global financial settlement layer, to “ultra sound money,” to a data availability layer for rollups. Maybe this is fine, maybe the industry is evolving so rapidly that the narrative does need to change with the times. None of these narratives quite found product market fit, so maybe it’s okay that Ethereum has pivoted a few times. But even if we write off the first few years, Ethereum is rapidly maturing and it’s time to settle on a clear vision. That isn’t happening yet, and it’s increasingly holding Ethereum back.
I often get the feeling that Ethereum is building tech for the sake of cool technology, without understanding exactly what it’s building towards or why. It’s like there’s no north star (other than the aforementioned values) and no specific product vision.
Ethereum is about pushing the envelope in several directions: originally, this meant adding a Turing complete VM onto Bitcoin, then transitioning to proof of stake, and now, scaling the base layer while maintaining decentralization, which is very challenging. But infrastructure needs to serve a purpose, to support a specific use case, and it’s not clear what specific use case or application Ethereum is targeting.
The “business model” of Ethereum, such as it is, isn’t working. A few years ago Ethereum pivoted to a “rollup-centric roadmap,” where execution and most innovation were pushed out to a variety of L2 chains while the base layer refocused on data availability. This was a pivot from the original “Ethereum 2.0” roadmap, which was focused on homogenous shards for scaling rather than heterogeneous rollups. Whether this decision made sense is controversial and I’ll address it in a future issue, but for now suffice it to say that the base layer couldn’t possibly scale to handle enough transactions on its own so a solution was needed. Instead, those transactions could pay fees to a L2 chain, and the L2 would in turn pay the L1 for data availability and settlement.
In an attempt to become the best possible data availability base layer, Ethereum later introduced blob space, which dramatically reduced the storage cost for those rollups, and fee collection at L1 fell proportionally. More and more economic activity has migrated to L2s since, and the base layer simply isn’t doing a good job of capturing value from that economic activity. Various proposals exist to address this, but at the moment Ethereum is more focused on continuing to scale blob space and making various other technical improvements rather than on fixing the economic problems, which are only going to get worse as blob space becomes more abundant and even cheaper. And data availability has become commoditized, now that there’s competition from Celestia, EigenDA, and many others.
In my opinion, Ethereum should lean into its strength, which is being a maximally secure, decentralized, credibly neutral base layer and settlement layer. It’s the one thing Ethereum can uniquely do, that no other chain can do. Not all block space is created equal. It’s literally marketing 101: Ethereum should find the specific applications and use cases that really value what it has to offer, and it should sell them a premium product at a premium price. I’m not entirely sure what applications that would be, but I’ve heard multiple builders and founders, including some building large, ambitious AI-related projects, saying that they only trust Ethereum for settlement and dispute resolution. There are probably also some applications that involve moving large sums of money, or politically exposed applications, for which Ethereum makes the most sense.
This is just one possible vision for Ethereum. There are many competing visions, which is the Ethereum way. Which one ultimately wins matters less than that Ethereum is able to pick one, stick with it, and maximize its value for that use case. I’m also not sure how that happens today, given the chaotic, slow-moving, consensus-driven, decentralized nature of its governance process. Companies with chief executives are good at rallying behind a particular product vision; Ethereum, historically, has not been. Put simply, Ethereum is trying to be too many things to too many people. It’s building cool tech for the sake of tech, without so much regard for how it’s used, or how to benefit from it. It’s time for Ethereum to decide to change that, or to face the consequences.
+1 on thing#1