Reflections on Devcon VII
Three Things #146: November 24, 2024
I just wrapped up a crazy Devcon week and I can’t help but write about the experience. It was my fifth Devcon—hard to believe—to say nothing of all the other crazy conference weeks the past few years. But this Devcon really takes the cake. As many others have said on social media, it was not only the biggest Devcon by far, but also the best: the most well-organized, the most fun, and the most intense, with the biggest diversity of stakeholders, ideas, and activities.
I don’t regularly write about events anymore, but I had to make an exception this time.
Thing #1: Bangkok 🛺
At first glance, Bangkok doesn’t seem like the best conference destination. The weather is muggy at best and unbearable at worst, even in November, and rainy about half the time. It has surprisingly few direct flight connections to most major world cities. The air quality isn’t great. Worst of all, it’s notoriously difficult to navigate: it can take ages to travel even a short distance. The traffic is a major headache, the trains don’t properly connect, and it can literally be impossible to get a taxi.
But there’s a logic to it all. It takes a few days to get used to it, but if you make some effort it all starts to make sense. I struggled the first few days, then I sort of gave up and went with the flow, and it suddenly got easier. Advice from friends helped. Locals and old hand expats will each share their own secret recipe for navigating the city. Inevitably these all boil down to some clever strategy for using multiple modes of transport, such as the classic “take a moto taxi to the nearest train station.”
I’ve always loved Bangkok. I’ve always loved the happy chaos and claustrophobia of the city, the narrow streets, night markets, and constant rush of mopeds and tuk tuks. By and large it’s a very safe city—helmetless motorcycle rides aside—and a city full of patient, happy locals who just want to be helpful. That’s more than you can say for just about any major US or world city. And Bangkok has the perfect balance of feeling developed, but not too developed: it feels clean and safe and foreigner friendly enough, has excellent medical care, has plenty of people who speak English, etc., but it’s still rough around the edges. You still have plenty of those “only possible in the developing world” experiences that simply cannot happen in Paris or Portland.
Another plus: the food is excellent, a fact that was on display at Devcon this year. The main event actually had excellent food! Which really is saying something, as crypto events aren’t typically known for good food. There was tons of variety, a rotating menu, multiple dining locations, and food that was just really yummy. (I’m biased: I’ve always loved Thai food.) The side events had good food, too. The hotel had good food. And, when all else failed, I ordered an amazing meal using the Grab app: it was cheap and arrived in minutes, and it, too, was good.
And the city is full of stunning venues, from the rooftop bars with views overlooking the whole city, to the old teahouses by the river, to the surprisingly good faux European restaurants and cafes. The main Devcon venue was also excellent: a modern, clean, bright conference center with excellent infrastructure in a not-so-inconvenient location right next to downtown.
And all of that’s just Bangkok; Thailand, of course, has a heck of a lot more to offer. I didn’t personally make it to the pre- or post-Devcon festivities, but it seemed like two out of every three people I met had either just arrived from Chiang Mai (where there were no less than eight (!) popup city events before Devcon) or was on their way to Phuket or Ko Phangan, home of some of the prettiest beaches on planet earth.
I didn’t meet a lot of Thai builders at the event, but I did meet many from the region: Singapore, Vietnam, and the Philippines, as well as from places farther afield in Asia such as India and Taiwan. A majority of attendees were first-timers. I know that the Ethereum Foundation is on a mission to bring Ethereum to as many global communities as possible, and I think Bangkok was a great choice for this Devcon. Developer activity and enthusiasm in the region is clearly growing. Bogota was nice, too, but Bangkok was a much more practical choice.
Thing #2: The Main Event 🏟
Devcon felt overwhelming, but still manageable. The event had 12,500 attendees, more than double the previous event and by far the largest attendance of the five Devcons I’ve attended. It’s up among the biggest crypto events I’ve ever attended, this year’s ETHDenver and Token2049. But the venue was also massive. There was more than enough space for everyone, and the organizers did a great job of carving out space not just for formal talks but also tons and tons of space for conversations, meals, and informal gatherings. Better yet, there was lots of buffer space between these, so there was no issue with noise from the latter carrying over into the former, as in previous years. And there was a huge lobby area and a shopping mall below where even people without event tickets could congregate, which was very helpful.
I won’t say much about the content. I heard it was excellent. I can tell as much from the program agenda. But I attended no talks this year. This is the first Devcon when I was 100% focused on the hallway track: on reconnecting with old friends and meeting new ones. (The talks are all recorded and available online.) Say what you will about Ethereum—more on this in a moment—but one thing the community does not lack is heavy-hitters, brilliant thinkers who can give powerful, inspiring talks.
I’d rather talk about the people, which were awesome, as usual. Somehow, Devcon manages to mostly attract builders. I don’t know how this is the case, even in a bull market, but I met easily four or five builders for every investor or marketing person. (Lots of love to investors and marketing folks, but this is a developer event, and I wanted to go deep technically.) This means the overall technical level of the discourse is quite high. I constantly found myself pulled into conversations such as debating the merits of the Beam Chain proposal or talking about the risks inherent in the Eigenlayer design. This is one of the things that keeps me coming back to Devcon year after year. It’s one of the few venues for this sort of deep, broad, in-person technical discussion.
Yes, the side events were a bit overwhelming. I attended a few, here and there, mostly before Devcon began and after it ended. The traffic and transport situation in Bangkok made the situation truly laughable. I’d have seven or eight overlapping side events, spread out across town. If I were in a city that’s easier to navigate—e.g., Denver, where I could just hop on a bike or a scooter and the next event was ten minutes away—then I’d pop over to a side event to support a friend or connect with an acquaintance, and come back an hour later. In Bangkok, however, it could easily take an hour just to get a taxi, then an hour stuck in traffic to travel two or three kilometers in pouring rain, which means that attending any event meant committing half a day.
The upside is that it made the decision to stay at the main event easy, and it struck me as quite dumb to organize events competing with the main event, which was anyway quite good. I had no trouble meeting the people I wanted to meet there, one after another, for four straight days, and had basically zero FOMO about being anywhere else. What’s interesting is that this is very different from other big crypto conferences I’ve attended this year, where the action has typically been at side events. I wonder how much that’s down to Devcon, and how much that’s down to Bangkok.
Devcon is many things. It’s about exchanging ideas. It’s a cultural celebration and a shared expression of values. It’s a solarpunk university. It’s actually a pretty good place to get work done for teams that are otherwise fully remote. It’s a very special time of year. (Despite the name, it’s not, apparently, a conference.)
But above all else, it’s a gathering of some of the best people in the world with some of the best vibes in the world.
Those vibes were intense. It takes a lot to wear me out, but by the end of Devcon I was fully depleted, both social and physiological batteries on empty. I slept the entire day on Friday after Devcon ended to recover, then attended the ETHGlobal hackathon, which was also excellent. Then I slept for two more days before I finally began to feel recovered, and ready to do it all over again. (ETHDenver 2025 anyone?)
Thing #3: Ethereum 💎
I’d be remiss if I wrote about Devcon and didn’t speak to Ethereum as a project and as a community. It’s been quite a long time since I’ve done so, since I’ve been so focused on Spacemesh, and since there’s no shortage of loud people sharing opinions about Ethereum on social media. But Ethereum has had a turbulent few weeks and even many longtime supporters have begun to fade the project. Naturally, Devcon is the best opportunity to measure the overall health of the Ethereum project and community. And having been a part of it for seven years, I have some perspective on what’s changed and what hasn’t, on what’s gotten better and what’s gotten worse.
Let’s start with the good news: Ethereum is doing just fine. The network is healthy. The community is as vibrant and active as ever. While its lead has shrunk in relative terms, Ethereum still has the lion’s share of developers and useful, interesting applications. (Note that I say this as someone who regularly attends events and makes an effort to remain active in other ecosystems and communities, and who has high regard for other projects including Solana.) All of this was on proud display during Devcon. As mentioned, this was the biggest, most diverse, most fun, and most well-organized Devcon ever, with the best content.
Also on display at Devcon was the prowess of the Ethereum R&D community. The Beam Chain proposal, though much misaligned (more on this in a moment), was bold and, importantly, remains true to Ethereum’s core values, including decentralization and censorship resistance. This is what differentiates Ethereum from the competition, more than fast block times, but this fact is unfortunately usually lost in the noise of online debate. This is just one example among literally hundreds of people, talks, and ideas that were on display at Devcon.
On the flipside: as I’ve mentioned many times before, this has long been Ethereum’s race to lose. Ethereans have grown used to their huge head start and huge lead over the competition. As is so often the case for the early winner, it’s led to a sense of complacency, entitlement, and hubris. (Note that I say this as a proud member of the community and as someone who truly loves the Ethereum project and community from the bottom of my heart.)
There are parallels here with geopolitics. Ethereum is the United States of blockchain. It worked really really hard in the beginning to break away from the “mothership” (Bitcoin) and establish itself, but it’s grown a bit lazy and slow in its maturity. The vast majority of early stakeholders have gotten so wealthy and retired long ago, so that Ethereum’s success or failure no longer matters for them.
It’s become harder and harder to make or even propose big changes in Ethereum, which is both an inevitable outcome of success but also a reversible cultural shift. There are ways to make big changes, even to a mission-critical, production system: witness the Beam chain’s proposal to touch only the consensus layer and to do so in parallel to the existing Beacon chain, both of which make a lot of sense to me.
Solana and the threat that it poses is the best thing ever to happen to Ethereum. This is clear from the bigger, bolder proposals that are beginning to emerge in Ethereum—something Ethereum has desperately needed for a long time.
And the ugly: Does it have to take five years to get this done, and are four second blocks the best Ethereans can expect, while the competition speeds towards 200ms blocks? And is there anything Ethereum can do about value capture increasingly occurring at L2? These are all difficult, contentious questions. I’ll offer my take, but I’m just one small voice in the ecosystem, and Ethereans will have to find their way through this together (or die trying).
Regarding the time frame, five years feels like a long time when you imagine that you’re sprinting against a competitor, but it’s still early days and in the grand scheme of things—and in the long arc of decentralized software development—it’s really nothing. Yes, it really does take that long to get things done in a Meaningfully Decentralized Fashion. No, Ethereum isn’t going to turn into a corporation, pull an Elon, and do everything 10x faster because that’s Not Its Way. It’s not true to the Ethereum ethos or core values, and unlike for other projects and ecosystems, those things matter a lot to Ethereans.
If this doesn’t make sense, go reread The Cathedral and the Bazaar. Decentralized software development processes are slow, and messy, but they steadily produce better and better products and converge on some version of the best truth. Linux is the best example of this, and Ethereum isn’t far behind.
As for the block time, I actually think four seconds is good enough for the vast majority of useful, interesting applications for the foreseeable future. There’s hardly anything other than some corners of DeFi where faster speeds matter, and as you probably already know I’m not particularly enamored of DeFi. Most users of most applications just don’t care, and well designed and built applications can largely hide this latency from the user. What’s more, L2s can provide arbitrarily fast confirmation times, and there’s a lot of neat tricks that Ethereans still have up their sleeves involving things like preconfirmations. I’m just not worried about this at all. Ethereum isn’t going to succeed or fail on the speed of its blocks.
Finally, regarding value capture, frankly, this also doesn’t worry me. The main reason is that the ecosystem is still nascent and it’s still too soon to tell how everything plays out. My gut feeling is that the Ethereum L1 is still going to be the place to settle high-value transactions, and most cross-chain and cross-ecosystem transactions, for a very long time, and that ETH will do just fine as a platform for value capture on the back of this.
What would it take to derail this? It would take a theoretical future world where direct cross-rollup, cross-L2, cross-ecosystem transactions become a thing, and they skip the base layer entirely. What’s more, it would take ecosystems including Base, Optimism, and Arbitrum each deciding to collect fees in something other than ETH. That would be bad for ETH the asset and the base layer, but that possibility feels very remote to me today. The Ethereum ecosystem may be fragmented, but it’s a thriving socially aligned and values-aligned bazaar. Which, again, is more than you can say of the competition.
The ticker is ETH. And Ethereum is going to be fine.