Immutable Foundations vs. Adaptive Evolution: The Future of Blockchain Protocols

John is going on stage in 5, 4, 3, 2, 1, go. Okay, alright, here we go. You've had your sticky rice, you've had your espresso. We gave you not one, but two spicy debates, and we have a third. This afternoon, be it resolved, blockchain protocols should favor stability through ossification over evolution.

Now is your chance to vote. As before, you can scan the QR code. Give it a few minutes. The votes are coming in. I'll give it 30 more seconds.

Okay, as it stands, we have a majority of voters for the disagree category. Please give it up for Uma, Odysseus, and Philip. Thank you, John. All right. So here, left of me, Uma, so founder and CEO of Succinct.

For those who don't know, they make zero-knowledge easy, basically. And they famously built SP1, which is a fast and production-ready ZKVM. On my right, Odysseus, founder and CEO of Phylax Systems. They built an open-source security protocol for securing dApps on IoT networks by actively censoring hacks on the underlying network. So today we'll be debating about ossification or, well, not doing it, so agility, basically.

So the idea of ossification is basically to make sure that the base layer or a protocol basically does not change or becomes immutable. The idea is, well, for it to remain stable, everyone who builds on it does not expect any changes. It remains secure, so updates could potentially introduce security issues, which with ossification is not the case. So again, as for both of the other debates today, it's kind of a spectrum. So you have hardcore ossification where nothing ever moves and full agility where everything, like kind of as a startup, move fast, break things.

But then, of course, you have anything in between. And maybe a good example for ossification is Bitcoin. Maybe they still change a bit, but they are pretty much ossified since the beginning. So the 21 million tokens, all of that has always remained the same and has not changed and probably never will. Let's see.

And then on the other side, maybe Ethereum is a bit more active and does hard forks and whatever. Of course, this is the case for blockchains, but it can also be the case for protocols, bridges, smart contracts even. Like a smart contract is fully ossified, fully immutable, but then you can also do upgradable smart contracts which are fully agile or anything in between with governance, whatever. Yeah, so to start off, actually I think the basic state is to be agile, I would say. So I'm curious, Odysseus, why do you think ossification is actually necessary?

I think it's both valuable and a necessity culturally for Ethereum. I think we have to lean into our strengths, which is not shipping very fast, and lean into the value proposition of a certification, which is something very lindy. If something worked for 100 years, it will probably work for another 100. And there is deep value in that, right? There is a reason why most banks run on COBOL.

If it doesn't work, if it works, don't change it. And Ethereum can provide this value proposition for a bedrock of the global financial system that may be suboptimal, but it works. And we know what to expect. All right. Do you want to...

Yeah, do you have anything else to say for your opening remarks? Sure. So there's three pillars. First of all, ossification is valuable, right? Do you know why rockets for SpaceX have a certain size?

It's because of the train. They have to carry them. And why the train has the size? Well, it's because of the trucks. And why the trucks are eight feet, not four feet, like eight inches?

It's because for 200 years we had the same size. And when they built them back in England, they used the tools they used for carriages. So basically, the size of the rocket for SpaceX, it's the size of two horses side by side. Essentially, the size of this measure has been ossified. The second is that ossification means stability.

It's secure. It works. That's the value proposition of basically the banking system in Switzerland. I can deposit some money there. Probably my grandchildren in three generations will be able to get them.

Bitcoin is ossified. That's why it's used by institutions. Probably my grandchildren in like three generations will be able to get them right Bitcoin is ossified. That's why it's like used by institutions. What that's why the ETF of Bitcoin is Accepted right because they know what to expect they trust it It's Lindy The second thing is that Ethereum can't ship.

Ethereum originally, if to at least had a thousand shards. And then 64. And then no shards. We gave up on scaling because we just couldn't figure it out. And it wasn't the time where there were scaling solutions.

Polkadot, Near, they had to find ways to enshrine scaling. And we just, you know, didn't do it. We invested so much money in Ethereum R&D, Greenfield Research, never came to light. The closest thing we had to product manager for the protocol was Danny Ryan, and we ended up burning him out. Agility requires a strong direction.

It's like a vector, right? You need to go from point A to point B. If you don't have the direction, no matter how much investment and power you have, it's like a Monte Carlo. It's a random walk. Maybe you end up where you want to go, it will be after too much investment, too much time.

I think this is why Solana is winning of this vertical. It's laser focused on one direction. Is it right? Is it wrong? Doesn't matter.

They're going for it. And if we try to play in this agility game, I think we'll lose. So that brings me to the final point. The L2s, the current landscape, is this a very happy coincidence. We threw in the tunnel, the towel for scaling, and basically we privatized innovation for our Ethereum L2s, right?

We said we can't find a solution, go figure it out. It's like privatizing the tracks and rails as a dysfunctional government. Some people may die. Some people may get scammed because of scamule twos. But the system as a whole will survive and thrive.

So what we can do here is we can ossify the base layer. We can offer this bedrock that has never ceased to work, that has been the same since the DAO reorg, right? And then allow innovation to thrive and do all this crazy shit in L2s. And if they go out, it's contained. Who cares?

It's one roll-up that died, right? The protocol remains. So this is how I want to set the stage for this debate. I want us to imagine what can be, but burdened by what has been. Awesome.

All right, Uma, what is your opinion on agility then, I guess? I think that was a very good take for your side. I think that my side is definitely easier to argue for. I think much of the audience agrees that Ethereum should not be fully ossified and be agile. I think maybe to start off with, a question to ask is, what does ossification even mean?

I think ossification is a spectrum, it's not a binary. So for example, Bitcoin, you have them bike shedding over whether or not to add opcat. When I first heard of opcat, I thought it was a complicated snark verifier on Bitcoin. No, it turns out it's just adding an opcode for concatenating two strings or two vectors. And it was actually in the original version of Bitcoin.

And then Satoshi took it out in one hard fork. And now they spend endless years debating whether or not to add it back. Do I think Ethereum should go to that level of ossification? Absolutely not. I think that has actually been very bad for the Bitcoin community, where for many years they can't innovate at all on the base layer, and the only thing they have is this digital gold, which is doing pretty well, I guess, as a narrative, but fundamentally there's zero innovation there because no one can build anything new.

On the other side of ossification, there's things like L2s, which, you know, I think in their final form should upgrade more than every six months. Maybe there should be a rolling upgrade on every L2 every month or maybe one day even every week. Do I think Ethereum, the base layer, is well suited to upgrade that quickly? Absolutely not. And so I think ossification is a spectrum.

So maybe this debate even doesn't make that much sense. It should be more like, where on that spectrum should Ethereum lie? And are we moving too quickly on the L1 today or too slowly? In general, I think being as ossified as Bitcoin doesn't make sense because we don't know the end game architecture of blockchain scaling. We've all been debating these topics for a very long time.

I think there's, you know, even zero knowledge proof technology has made tremendous leaps in the past years. That wasn't in a state where we could have ossified even five years ago. And so who's to say what technology there will be in the next five years that will change our mind on what sort of primitives we should have on the L1. To make a very concrete example of that, today on the L1 we have these BN254 opcodes on Ethereum that allow us to verify Groth16 and Plonkish KZG proofs. But there's a lot of technology that's coming out there, such as Starks, that maybe could be a better fit for the L1.

And maybe in the future it would make sense to have opcodes for those things. And so ossifying and committing ourselves to old technology while technology and innovation is always going to progress doesn't seem very smart. Even other examples like cryptography changes over time for getting starks or snarks, which is kind of the world I live. For example, quantum security is a topic that has come up recently. And obviously if we get more advances in quantum computing, we will obviously have to upgrade the base layer to become quantum secure.

And as another example of economic attacks that aren't cryptography related, MEV did not come into existence for many years during Ethereum's history. So Ethereum started and then there was MEVGIF, but that was actually, I think, probably more than five years in. And so what will the next five years of these style of innovations on the crypto economic side look like? We're not so sure. And so it seems silly to ossify even solutions to things like MEV and the protocol at the current point in time when there could be more innovations down the line.

So overall, I think because technology is moving so quickly and then there's a lot of people working on the Ethereum protocol and also broad Ethereum ecosystem that are very innovative, it doesn't make sense to have a very strong commitment or point of view to the current base layer as it is today. We have to be willing to innovate even if it's just to accommodate innovations in L2s or MEV or any of the very rich ecosystem around it on top of Ethereum. Thank you. I think you actually made a very good point regarding quantum computing and the need potentially to change cryptography. And actually, what would your point then be on that side?

Like if we need to change the base layer because of cryptographic cryptographic issues, how would you do that? Or for example if for layer 2s we need new primitives, new opcodes, should we then be able to be a bit more agile and update the base layer? Or we can't ship too fast, we should ossify directionally. We should define what is good enough and then try to get there, get there as fast as quickly as possible so we can leverage the ossification value proposition. So for example, if a new, if quantum computers came to realization and suddenly sure someone can break Ethereum, yeah, probably that's a good, to put it into that ossification spec.

But an opcode that would make some Zk-snark a bit more efficient, I don't think that would be, I would count it as, it's good enough. Got it. But I guess you would then prefer a new opcode to make things faster, for example, right? I think I actually agree with Odysseus' point of view. It depends, like, what the trade-offs are.

I think Ethereum's main problem right now is that there isn't the clear end state we're striving for. There are some people who say we should scale the L1 completely and not have this L2 scaling roadmap, or others are saying we should not only do DA and settlement. I think right now there's kind of this not clear vision of, as he was saying, what is like Ethereum's, not to use this word because it's been on Twitter a lot, North Star, or what is the desired end state fully. And I think, for example, Solana does have that to an extreme, which is we want to just do as much TPS as possible, as cheaply as possible. I think Ethereum, there's some people that say we should really scale the DA and just become this settlement layer that verifies proofs with ETH as an asset and just do DA for all the roll-ups.

There's other people who say, oh, DA isn't even a critical part of thatced to like EigenDA or Celestia. There's other people who say, no, we should do everything. We should do DA, we should do execution, we should do settlement, we should be money. So I think it's really hard to ossify anything right now because it's not even clear what of those four things we want to do. And it's not even clear, maybe there's gonna be other things that come up in the future.

So I think upstream of the ossification conversation, I think first it has to be like, what's the clear end state? And then once we have that, then I think talking about ossification makes a lot of sense. And so maybe that's actually a reason why Bitcoin ossification is fine. They want to be this pet rock. They want to be the protocol that definitionally is ossified.

And so they've done a really great job at that. For Ethereum, it's not even clear what the goals are. So it's not even clear how we can accomplish goals that are uncertain, in my opinion. Good point. So yes, I guess Ethereum's vision is maybe a bit fuzzy currently.

So with Pectra coming up and then those ideas of being more, you know, building on account abstraction and potentially everything linked to layer two, do you think we should maybe not do that Pectra hard fork? Should we slim it down? Or what would your vision be for Ethereum? I think Ethereum doesn't have an Earth Star because we have like an Earth sky. Everyone has its own idea of what is Ethereum.

And, you know, this is why I think, I mean, this is I think why we need to ossify because we can't agree of what Ethereum is. And I don't think that will change. Because the community and the culture perpetuates and recycles. It attracts this kind of people. People that like the Solana, the Anatoly dictatorship, they will be attracted to join Solana and exist in that community.

People that like this thing here will join in this thing here. So I don't think we will ever, probably ever get a North Star. Maybe apart from decentralization, right? I think this is probably the common denominator, right? So I think ossification is a necessity based on that fact.

So let's agree on what we agree as a community, which is decentralization, you know, like rational specs for running a validator node, et cetera, and we also fire around these parts. Obviously, we continue to improve on the database and the, you know, client implementations, like, you know, ossification doesn't mean that we should stop developing, you know, ref, meth, death, whatever. It means that we don't develop the specs anymore. So I think Ethereum as it is now, if it would ossify, sure, it wouldn't be a very good DA, probably not a very good execution layer, maybe too expensive to do verification But it would be this thing that has never stopped working since 2015 the same right and I think this is a very Good value proposition for This the new financial system sure, you know Solana will be Nasdaq and meme coins and fart coin and whatever, and people will gamble there. But Ethereum will be where institutions deposit half a billion dollars and make your DAO.

Where crypto banks will live, and then maybe the users, like you have Venmo, you have a roll-up for your bank. So I would actually push that and say we can ossify around here because by the time we go to the end state, it will be too late. Don't trust us. So basically keep Ethereum as is, like not even do Pectra, potentially do a hard fork in case quantum computing becomes readily available, and that's it. And then improve on layer 2s maybe.

Yeah, because my whole argument is that if we go into this mindset of let's find this end state, let's decide what's the end state, which is very different from what we have here. And let's go to that. It will take like 25 years and I'll be dead by then. Makes sense. Yeah, and then to his point about Ethereum being safe and being able to add or put billions of liquidity on it, isn't agility kind of how we got into this mess about, you know, bridge hacks because of upgrades or new software updates or hard forks or whatever?

Security-wise, how do you see this going? Is this not a risk of being able to upgrade contracts, protocols, systems, anything? I think the current Ethereum core dev upgrade cadence is actually very, very geared towards security. Most of the bridge hacks are actually because people are, well, just they're using multisigs, they're not using the blockchains that are meant to be used. I don't think it has much to do with Ethereum's cadence.

I think the core devs test things very extensively. They upgrade every six months or maybe even longer. So it is actually a pretty slow cadence. I think things are intentionally not rushed. It's a very large committee that decides what's going in.

It's usually pretty non-controversial by the end because so many people have to agree to it. So I actually don't think security, anything insecure is being pushed in for the sake of it. I think the Solana core devs are much closer to that where they are pushing in stuff constantly. Sometimes they even do stuff that breaks a chain. I think Ethereum is not at risk of that at all.

It leans far away from that. I think one thing that is really needed for ossification, if we do want that as an end state, which I'm not sure is a desirable end state, but if so, let's suppose it is, is that the L2s really need to break free from thinking of themselves as EBM compatible. So today L2s, most L2s view type 1 EM compatibility as this holy grail, where every opcode has to be the exact same, the gas pricing has to be the exact same, they can't innovate on the bytecode format, they can't innovate on the state tree, and type 1 is very, very desired. I actually think that's really bad because if, say, the Ethereum doesn't do EOF, they don't do any changes, it's actually a very inefficient system. The Merkle-Patricia try is really horrible.

The bytecode format leaves a lot to be desired. And so if L2s are kind of pigeonholed into having to follow that, then that's very negative for innovation. And so I think that's like a mental blocker that needs to be overcome in order for us to feel better about ossification at all. I think this is very Ethereum centric, let's say. But I think ossification can also apply to protocols or bridges or smart contracts.

What's your opinion on that side? Like should the smart contracts be upgradable, especially for bridges that hold hundreds of millions, if not billions? Well, if you look at rollups, for example, which are smart contracts that are bridges that hold billions of dollars, I think at this point they have to be upgradable. You cannot have a rollup on a base layer that upgrades that isn't upgradable. Now I think upgradability in and of itself is not bad at all.

You can have time locked upgradability, 30 day time lock, where if you decide to put in an upgrade and your users don't like it, they can exit no problem within 30 days. So I think upgradeability in and of itself is actually not a bad thing at all. Time lock upgradeability I think solves most people's problems with upgradeability and it's doable today. I think most people should have it. I think the problem with upgradeability right now is you have upgradeability that's governed by a multisig that's instant that can just take all the funds, and that is bad.

At the same time, sometimes in these current systems that we have, they've only been in production for a year or two. So it is quite reasonable that you'd want to safeguard that can step in instantly without a time lock. And in those cases, I think it's fine to have kind of like pause or freeze functionality that can't take funds but can at least pause things. And that's like a very pragmatic trade-off for the reality of the world we live in today. That sounds reasonable.

Well, maybe a similar question to you then. For bridges or contracts in general, do you think they should be upgradable or not? And if not, what happens if there's a security issue, for example? I think whatever is above the protocol should be as innovative as possible. Upgradable, malleable, go crazy.

Build an EVM in PHP, on-chain PHP, I don't care. I think this whole EVM equivalence is actually very interesting of a social layer specification. It's like the social layer almost ossified the EVM because people like the value proposition of specification, which is, I know the EVM, how it works, its quickness, I know all the stupid mistakes it does, I know it. I feel safe to store billions of dollars of dollars. And all these other EVMs, sure, they're nice, move is very sexy these days, but I don't know it.

But we shouldn't ossify the EVM at the L2 layer. I think that's a bad move. I understand the value proposition and why people like it, but I think the L2 should be pushed to become as innovative as possible, you know, like Eclipse, like SVM, really go crazy. Because that's how the Ethereum as a whole is competitive and also stable. So to come back to your question, I think I understand what, you know, bridge should be upgradable, and I think they should.

If you have an organization that is operationally mature, then that's not a problem. Got it. So if ossification is the desired end goal, would you say that layer 2 chains at some point also need to ossify? And if so, should they start by being ossified? Should they slowly move towards such an end state?

Or should they, as the innovation layer, let's say, stay agile? It's like saying what's the best business strategy for a company of the private sector? It depends on the product they want to sell. If Citadel wants a roll-up, probably they will be EVM equivalent. They want something boring.

But if you want to gamble meme coins, probably you want something super optimized where it's very cheap It's very instant and whatever because you don't care about the value proposition of ossification You want other things you want innovation you want the latest tech? Which means you know scalability cheapness blah blah blah So I think It really depends on the value proposition of every roll-up and how they compete with one another and how they compete with other blockchains. And I think, you know, BASE, for example, now, it's competing with Solana. Like Solana is not competing with Ethereum. I think Solana is competing with Base, right?

Because Base offers that the same or similar value proposition, which is very quick, very cheap to use, very cheap to deploy. And who the fuck cares about security? Right. So for Base specifically, do you think they should ossify at some point? Well, to be honest, actually, we just activated the fault proofs.

So they care a lot more than Solana, maybe. But Bay is a very good example of competing with Solana on the same vertical. Ethereum is not competing with Solana. Right. So Bay should remain agile then?

BASE is probably the most agile rollup we have, right? They are constantly improving the size of the blocks, yeah. They're adding new innovations, they're experimenting with new OP stack things. Saxon is doing a lot of things with OP stack as well. This is what we want, right?

Experimenting and playing. And if it breaks, maybe OP goes down. But it's okay. Ethereum will survive, right? Right.

Yeah, so not as spicy of a take as I expected, and certainly not as spicy as the previous debates, but that's fine. So maybe, yeah, so Ethereum is moving from the Merkle-Batusha try like you just mentioned to vocal trees, right? Maybe. Maybe, yeah, maybe if that happens at some point. But this would potentially break a lot of things, right?

So Ethereum mostly remains backwards compatible, which with this change will no longer be the case. What's your opinion on that? Well, I think vocal trees are just a very bad data structure for many reasons. So I think Ethereum, even if it is willing to throw away the Merkle Patricia try, it should migrate to something else that's more SnarkStark friendly, that's a more performant data structure that has a lot of the benefits that it wants. So there's whether it should do Virkle.

Then there's the question of whether it should migrate away from the Merkle Patricia try at all. I am of the opinion that if it poses significant benefits for the protocol, especially given that the state of L2s today is that they just mimic Ethereum without thinking about it, then the Ethereum protocol should do it. There is a good question as to whether people in the Ethereum community will agree to any given proposal. But I don't think philosophically I'm against Ethereum moving away or doing something that's not backwards compatible. In fact, all the tech debt that we have to keep around to keep things backwards compatible kind of slows down the core development, it slows down all the clients, it slows down everything.

And so I think being willing to make these changes that are breaking but have a longer term benefit should be on the table. Makes perfect sense. Can I say something? Yeah, of course, go ahead. So I was reading a few, like a week ago, but how Solana basically shipped a database in like six months, a new database.

We've been talking about for like three years. That. Right. So, maybe to your point of, well, maybe both of your points, I guess, that Ethereum for one maybe doesn't move fast enough at this point. Yeah, I'll just leave it at that for now.

We want to move. We want to move, but... But we can't. Yeah. I do think maybe one thing I do disagree with you on is I do think Ethereum should have this North Star.

I think it is really important to have a strong, maybe not one leader, but a group of leaders at the top setting the direction saying this is our goal. Here's what we want to accomplish. Let's work towards it and uniting everyone. Because otherwise you get a lot of infighting. You get people pulling in different directions, and even though there's a lot of energy and momentum in the Ethereum ecosystem, and there's a ton of money, there's a ton of people, there's a ton of, more important than the money really, there's just a ton of very, very smart people working, because they're all moving in different directions, the effort doesn't result in the outcomes that we want.

And so I do think it is actually very important to have this goal and North Star that we are broadly all aligned towards. And I think, yeah, maybe that's... I think the North Sky has its problems. I agree 100% with you. I would love us to have a North Star.

I'm just saying we don't. And I don't think... But don't you think they could change? I don't think... But don't you think they could change?

I don't think culture changes very... Like, you know, in countries, like, historically, if you change the educational system in order to realize that change to the society, you need, like, 30, 40 years to see the effects. I think people change... Probably don't change. And if Ethereum community was to change, then it means that the majority of the old guard would have to be pushed away and then introduce other people that would introduce their culture to the community to change it, right?

So because that will not happen and people don't change, I think that obviously the Ethereum community will evolve, but the critical mass of a very specific culture will remain. So, what I'm saying is that we have to accept it and lean into what we're good instead of trying to be something that we aren't. I don't think we'll ever be Solana. And that's OK, because there is a lot of value in being ossified. That's my point.

I think you can still have a really clear, strong North Star and be ossified. Like, you can still have everyone be very excited. I think if Vitalik came in tomorrow and he said, you know, instead of writing six blog posts about the merge, the verge, the splurge, the whatever, like very technical blog posts, he wrote a blog post being like, oh, this is the end state vision in a language that a lot of people could understand and get excited about. I do think it would be possible to change. If he did, that would be, who are you?

And what did you do to Vitalik? Yes, ideally I would love to. I think for me personally, what I find most interesting about Solana is actually Anatoly. He's a multi-billionaire and he's in the trenches on Twitter all day fighting with, you know, or, you know, MERT, going to the DGEN conduit conundrum and saying, you know, if you come to Helios, I will personally paint your servers purple. All right.

Let's leave it at that for now, because if not, we will not have any time for questions. So let's start with that, actually. I don't know if anyone has a mic. I can't see much. Hi.

Hi. I don't know if you discussed this exactly, but I think Martin Köppelmann at EVE Denver gave a talk on why Gnosis shouldn't become an L2. And one of the arguments was that L2s will have to be upgradable for a very long time because the L1 is not ossified. And so things on the L1 that change, they necessarily mandate L mandate to be upgradable I think some arguments were made that it's the quality of upgradability is important like I think you might you said multi six are not a good model or something like that but yeah I almost feel like I mean I can't make the argument for L2s to be truly decentralized and not controlled by anyone without the L1 being ossified almost. So I would love to hear your opinions or whatever.

Thank you. Well, even Solana is upgradable, right? They do upgrades and then sometimes those upgrades break, and then they do more upgrades. Or even Gnosis Chain, for example, is upgradable. So I think Gnosis is not upgradable because it's an L1.

That's not even true. I think the quality of how you do the upgrade, I think, is the question. So if your upgrade is a five-person multisig that can instantly upgrade and take all funds, I agree that's probably not that acceptable. If you have a time-locked upgrade by well-known people, maybe that's actually probably quite acceptable, but maybe that runs into regulatory issues. Now maybe you have an upgrade that's governed by token holders or something like that, and then it's decentralized, that becomes even better.

So I guess I challenge that L1s are better than L2s on this dimension because L1s also do upgrade. It's just the validators of the L1s, and ultimately who recognizes the canonical fork of the L1 that ultimately chooses. And then there I would argue that maybe Coinbase or Circle is actually the ultimate upgrade authority of a particular L1 or even an L2, right? So I actually don't think there's much of a different, I don't think that's a strong argument for not becoming an L2. I agree.

And also the underlying implication that Gnosis is like decentralized? I mean, okay. What's that supposed to mean? Anyways, let's move on to the next question. Since you both agree on so many things, let me disagree with both of you on something.

I don't think that we have this incredibly wide open sky of all these conflicting possibilities. I think we have a fairly tight zone. You know, we're not all pointed at exactly the same place. But one of the things that's driven Ethereum forward is the Ethereum project has always been about very passionately exploring how do we simultaneously maintain a reasonable, accessible degree of decentralization, strong security guarantees, and within those bounds, maximize performance to reduce costs and make opportunities accessible. And I think that's a fairly clear target.

I think that the reason that we spend so much time on the subtleties of things is that sometimes the trade-offs are difficult. I think sometimes figuring out the technical problems is difficult. And I think sometimes we're stuck in the situation where we're working on technology that's rapidly changing. And so if we commit to something now and something else comes later, we spend a lot of time trying to figure out if like, we've got the best thing yet, or if we've got the best thing that'll at least last long enough for the amount of time and effort that we have to put into it. And so like, I think these, these challenges arise naturally out of our strong sense of direction that we have.

And as we proceed further down that, that, that story, you know, like things do sort of get checked off, you know, we're not trying to move off of proof of work anymore, you know, we do have some degree of scalability, it is literally possible to do something on the Ethereum network that doesn't cost $25 every time you send money, but just because we've checked off part of it doesn't mean we're finished, and we're never going to be finished, and that's the point of the direction and destination of Ethereum. So yes, maybe some components get finished over time and then those components ossify. But simultaneously, until we stop solving new technological problems that fundamentally impact this thing that we're passionate about in the Ethereum community, we're going to keep moving forward. And we can't just ossify the entire protocol and say we're not going to do that anymore because we have this strong sense of direction. Even if it's not spelled out in like this is exactly what we're going to do and even if we don't centralize that power in like a CEO or like a central entity who can just say, this is what we're doing this quarter and then like if it was a mistake, it's a mistake.

Oh, well, we'll have to fire Steve Jobs. But even though we don't do things that way, it's because of our North Star. North Star is about decentralization and not putting that type of risk into the protocol. So that's not a question, but I'm curious what your take is on that. Please.

Do you want to? So, you said a lot. And I didn't take a lot of notes. But I will start from the last thing. So, you said that we're solving technological problems.

And I think that's where I disagree with you fundamentally. We're solving human problems. The point of working in Ethereum is not to mentally masturbate on new and new technologies, right? We spent $2 million on doing research on asynchs for the random generation function in Ethereum. That went nowhere.

We're not here to solve interesting and technological problems. We're here to solve human problems. And if we keep changing this, then people will not be able to use the whole value proposition of a certification. Like institutions will never trust, it will never become the backbone because it will always change. And because we don't have this CEO thing you said, then we'll not go somewhere.

We'll be just like a crazy person that's running around the street instead of going to a certain direction. So I would say that you said, and then finally, you said at the start, you said that we have very clear goals, right? Sure. It's like everyone wants to save the world, but everyone disagrees on the how. Right?

So, yeah, we have these goals, but it's like people can't agree on how to solve them. We want to shard and scalability. But we couldn't. Like, we were working for years on sharding. And we ended up saying, ah, fuck, we can't.

And there was Polkadot at the time, and Near, with good research and good implementation. They failed for other reasons. We didn't take these design patterns, we didn't adopt them for whatever personal reasons. Maybe we don't like Gavin or whatever, right? So sure we have these goals, but we have proven that we haven't been able to coordinate in order to achieve them within reasonable either, you know, with reasonable time or with reasonable investment.

Like, the Ethereum R&D was much bigger than NIR and Polkadot. That's my... Just to acknowledge what you said, like, I certainly acknowledge that Ethereum doesn't just solve technical problems. That's one of the things that the Ethereum community does. I would certainly not say it's the only thing, so I just want to acknowledge that.

Uma, comments maybe? I guess I still maybe disagree that there is a clear canonical direction everyone's working towards. I think, for example, even in the landscape today, I think everyone wants to see more applications and more users on-chain. I would agree with that. But today it feels like, okay, well, I think everyone wants to see more applications and more users on chain.

I would agree with that. But today it feels like, oh, well, I want more applications and more users on chain, but not on your L2, on my L2. And so even there, I think, and I think that's for a really good reason. I think that makes total sense. I think, you know, I would also be the same way.

But I think it is not a clear direction that everyone agrees on because I think we all agree on the end results, but we don't agree on the best way to get there. We don't even agree on, for example, the best rollup architecture, optimistic rollups, ZK rollups, whatever. And so I think that social division is maybe what is maybe showing up more in the Ethereum fabric right now. And I think it's a very good... I don't think people are being very, very selfish or very, very bad people or anything like that.

I think it all makes sense given the incentives at play. Yeah, to that I would just say that of course, rollups are not great. Yeah, let's keep it at that. If there's any other questions, maybe we can talk about it outside because we are running out of time. So before we end, if you have any closing remarks, now is the time.

I'll start with Uma this time as you started before I think it seems like Odysseus and I actually agree on many things I think probably our biggest point of disagreement is that it's not possible at this point to have like one clear North Star or one almost centralized leader or centralized authority or voice. But yeah, I think the current pace of protocol upgrades, in my opinion, I would like it to be a little faster, but I wouldn't want it to be a lot faster. I actually think the cadence is fine. I think the other big thing is I wish the L2s really broke free of having to be on parity with the L1, because the entire point of Ethereum's architecture is if you have an ossified base layer, thing is I wish the L2s really broke free of having to be on parity with the L1 because the entire point of Ethereum's architecture is if you have an ossified base layer, you can have more fast-moving sub-layers underneath it. And having the ossification and slow movement of the base layer doesn't work unless everyone else underneath is moving faster.

So all the roll-ups need to move faster for this entire architecture to work at all. And yeah, I think in general, like we do all agree on the outcomes of wanting more people on chain, wanting more money on chain, wanting more applications on chain. So that's always important to remember on top of like all the Twitter drama that like perhaps other motivated parties stir up to like, to stir division in the community. It's like actually we do have very much similar goals and are working broadly towards much of the same thing. Thank you.

Your turn. I fully agree that we should allow the roll-ups to run free. Even though I think it's interesting how they have this natural EVM-sification direction, not because they were said to be ossified, but because they see value in being ossified. But I fully agree that they have to break free. We have to see innovation.

Roll-ups need to try crazy things. A lot of them will fail. Some of them will not. This capitalism. And Ethereum as a whole, as a community, will actually evolve.

And we'll be able to compete with Solana, with Celestia, with many other different chains. Because rollups will be able to compete with them by choosing what trade-offs they want, right? And then allowing the markets to decide or the VCs to decide what rollups are good or aren't. Yeah, I guess the only disagreement we have is you trust more to people than I do. In the sense that this culture can change and finally find a direction.

While I just think that if a group of people haven't showed a behavior for an extended amount of time, probably will not change. And I don't think that's bad, actually. I just think we just accept who we are. We lean into our strengths, which is this plurality, this linkedness where we've been through so much, right? If you think of the Maker, the Maker hack, right?

It was crazy. We've been through so much and we're still here. Ethereum is as big as ever. Billions of dollars of investment flowing into it. We have the ETF, which means that institutions or the financial, we have the acknowledgement of the financial big daddy have the acknowledgement of the financial, you know, big daddy.

And I think, you know, very soon we will start to be attacked as we see institutions deploying on Ethereum. And, you know, people, you know, Wall Street doesn't care if you buy Ethereum peer-to-peer or if you put $1,000 into ETH. Goldman Sachs cares that the pension fund from the Workers' Union in Canada doesn't invest $200 million and they put them on Maker instead of giving them to Goldman Sachs to manage. That's what they fear, right? And this is why ossification and this plurality can work.

Because when that happens, this is when we're starting winning. When the financial system that brought 2008 starts hemorrhaging billions of dollars, not through, you know, VCs that try to shoot and spray and, you know, and somehow explain their fees, but through funds that find value in using Ethereum, in deploying on DeFi, in deploying on Lido. And to do that, we need to offer them something deploying on DeFi, deploying on Lido. And to do that, we need to offer them something as boring as fuck. It needs to be stable, relaxed, like a green field of peacefulness.

They don't care about speed. And when that happens, that's when we win. All right. Well, thanks, both of you, for this very balanced debate. Maybe relaxed, kind of like you said, in comparison to the one before.

What happened before? Well, it's a long story. Watch the review on, I don't know where, actually, but whatever. Maybe we need a third option, like disagree, agree, and maybe a long story. Watch the review on, I don't know where actually, but whatever.

Maybe we need a third option, like disagree, agree, and maybe, maybe not. But I don't know if we can get that set up now. Yeah, abstain is not political enough. Come on. All right, let's vote again.

We'll conclude this debate with a final vote. So be it resolved, blockchain protocols should favor stability through ossification over evolution. And just to recap, we started off with 10 in the agree category, 17 in the disagree, and 9 in the abstain. stain. We might have the exact same result.

Well, we've achieved nothing then. Now, remember that the width of a rocket is equal to the width of two horses. Okay. We have the same results. 18 in the disagree category.

Well, almost the same. 12 in the agree.