Ethereum is a Store of Value | Vitalik Buterin and Kartik Talwar at Pragma Mumbai 2026

We're going to kick things off with our next session directly. And for this one, let's bring on Vitalik to talk about a lot of things that are on his mind and mine. So, please give Dr. Doom a big round of applause. GM Hello,

GM Cardic. How are you?

Well, uh, could use a little bit more sleep,

but who who who isn't in that position this week?

Mhm. That's good.

Well, I'm actually not sure if you're really looking at me or just uh it's just on the side. But it looks like I can see your eyes. This is interesting. Um, it's just hard to ask him a serious question now,

Card. Uh, what has been, uh, the most, uh, unique about this Defcon for you compared to previous ones?

Um, the unique the most unique thing I would say this week has been largely I feel like we saw much bigger cams of people who were more excited about AI than they were about Ethereum. So, this is not a good thing. Um, in terms of kind of my observation, I have nothing against this, but I feel like if you're at the Ethereum developer conference, um, the ratios should have been slightly more different. Um, but we have 1500 people just behind us. They are absolutely excited about Ethereum.

Um, and one thing that I do see that's changing for the better is that we're now making both AI and crypto tools instead of end goals. And it's now a much bigger shift towards solving problems instead of making the tool the thing.

So I think that chasm has been crossed and to me that's a pretty good point and a checkpoint.

Mhm. Uh same question for you, I guess. What's been the highlight for you this week?

Yeah, I mean I think uh basically two big things that I've noticed that are different, right? I mean I think one is that this is uh the most local and the and probably the youngest Defcon out of all the DevCons that I've uh witnessed to so far, right? And there's just lots of students from all kinds of uh universities like huge numbers of people for whom this is their first Devcon who even first uh encounter with the Ethereum world and uh I mean like they all just seem very uh excited about a lot of things you know like sometimes it's individual um categories of Ethereum applications sometimes it's privacy uh sometimes it's uh like just meeting people and learning what's what what's here. So, I think it's it's been uh good to kind of see that positive energy. I think the uh other thing that's been unique for me is this kind of overriding sort of not quite on the headline agenda, but uh on everyone's mind uh topic, which basically is I mean it's very related to what you said, right, which is basically how do we deal with AI, right?

I mean I uh you know I overheard probably the last quarter of you talking to Justin and uh you know we've moved on from postquantum cryptography to postquantum and post AI cryptography and uh it feels like for a lot of people the mood is basically that like hey AI is changing so much that uh I mean even like far beyond DevCon I think this has just been my experience uh this year and my experience observing how people are thinking this year like it feels like so many aspects of the world are kind of resetting from scratch. We're entering something that's uh fundamentally new and uh everyone is just like figuring out how to adapt and how to survive, right? And uh there was just lots of technical workshops figuring out how do we upgrade like kind of both quickly and uh as safely and sustainably as possible to all of these all of these quantum proof schemes. And uh I mean there's like a sense in which uh you know like this is sort of kind of like reverting back from uh you know the thing is done let's talk or at least the thing is in the background let's talk about how to use the thing and back to you know like talking about how do we protect the thing right but

so maybe then then the question I have is is the experience that all of us saw at DevCon was what was planned or did you see that sort of net output that we experienced was actually not a thing that you expected.

Uh that's everywhere from content to the focus to just which groups anywhere from get uh more time to talk about their their uh kind of goals like how much of that is

intention versus uh unintentional outcomes.

Yeah, I think making uh this dev kind of very heavy on uh people from India and uh people doing things in India was uh very intentional. I think uh trying to uh kind of create a bridge between Ethereum and uh various communities and various interests that are outside of Ethereum but that have aligned goals was also very intentional. Um I think the uh the focus on like dealing with AI, I mean that was uh that like that was definitely brought on by the outside world, right? I think uh

a lot of like if you like even the the results that Justin was uh like freaking out about like those are results that are literally less than two months old, right?

Yeah.

I mean some of them are from this morning.

Indeed. Yeah. Like like the the rate at which kind of the the AI is overturning this world. I think like it felt to me like like you know Q1 this year was like a massive C change and then Q2 this year was another C change and then uh you know Q3 was uh Q3 was like a third C change and then it's you know like we're possibly going to like due for yet another like big philosophical shift in how we see everything sometime around Christmas. Uh so that's uh

so maybe I guess to to then look at the retrospective of of this week. Today is the last day of

DEPCON.

What would you do different for next year's DEPCON?

Very open-ended.

Yeah, good question. And I think one is uh I mean I think in terms of location like probably the best part of this year's Devcon was that it's uh it got to reach a huge number of people who have been uh far away from all of the Ethereum activities so far. But of course you know the downside is like Visas and Distance are very hard for another huge part of the community. And so I think you know you have to balance those things right and uh I think kind of targeting accessibility for uh people who have who have done the most in uh like all of these uh kind of adjacent goals that we care that we care about and like both inside and outside Ethereum is uh like is it is important right and so I mean I you know like I mean like that that whole process of deciding is still ongoing, right?

Even more accessibility on location. That's one.

Yeah, I think that's one. Um then number uh number two, I would say yeah, I mean it's like less a property of the conference, right? and uh kind of more a question of uh like I think a lot of this really kind of made me start thinking more about some of these deeper questions of uh you know like what is the message of Ethereum what is kind of the underly the underlying goal of uh everything that's happening here right and you know you can talk about highle principles and um like I I continue to ve very much agree with all of the highle principles right but I mean I think it is sort of stuck in this uh transitionary phase, right? Where like to me personally just uh like talking about the uh like you know like getting like 3% APY a year just doesn't seem interesting in a a context where like the entire world is flipping over every year, right? And uh

is is 3.5% interesting?

Amazing. Uh well well the difference between those two I guess is worth like a month of AI progress right? Uh but yeah so like there's like I I I think you know there is a sense in which kind of like we really have to update the message to kind of adapt the kind of the moment where the world is right and uh but but but I think like there are there are a lot of actual synergies in terms of uh the message and what it should be right I mean I think uh the the simplest and mo one of the most basic ones is like people ask you know like what is the point of ETH And uh for a long time there have been these arguments about like is ETH like a token that collects fees or is ETH a store of value? And uh I mean I've kind of been open-minded to both for a long time but I think uh I've been drifting more and more and I think like this week I'm basically decided like ETH is a store of value and like sort of you know like ETH is money TM like basically is it right? And uh the reason why is basically because like the thing that you optimize for in both worlds is really different, right?

Like if you optimize for having a market cap based on collecting fees, then essentially you're being a corporation. And if you look at what winning real corporations do, like they have very very activist forms of governance. They yeah definitely yeah like do not try to like figure out like big detailed sets of norms that constrain them for for long periods of time. Uh I mean if they did originally they often try to wrigle their their way out of them. They yeah spend like singledigit percentage points of their market cap trying to grow themselves every year, right?

Like they do a lot of things that are just not compatible with being the type of project that Ethereum is. Yeah,

but isn't that an extreme example in a sense that like I feel like the answer is is more of a gray zone, right? Like are you

So I used to think this and

maybe one quick kind of question. So is the message here or rather is the thing that you're trying to say that you don't know which camp you're on or do you think there's a third other way that has slightly different trade-offs and you're trying to figure out what that would be? I think on specifically the question of ETH the asset I like I actually feel like I do know what camp I'm on. I'm on the camp that uh like the money like yeah the pragmatic personal reason to kind of the financial reason to hold ETH is that ETH is money right

and I think the reason why that's uh like kind of the most values align thing for the ecosystem to focus on is because the kind of properties that you want is like basically you want your money to survive right and uh like what is the main kind of like pragmatic upside of everything that people agree Ethereum is best at which is you know people say things like you know like crops the postquantum work the security work like the pragmatic upside is basically that it means that the Ethereum chain is going to be very good at surviving right so

okay so we've defined what that goal should be but then you're saying the question is how do you actually make it valuable which is now the economic argument for is this collect fees only or do something else that maybe gets us there is that correct or is that a is that am I am I understanding Yeah, I think like so one core argument is that basically that means that like collecting a huge amount of fees and uh you know like justifying market cap by net present value of the fees is not the focus right so that's one and uh I otherwise there is the question right of like what other things make ETH attractive to hold right and uh one of them is uh I mean one of them is kind of surviving right um another is like being like being very close and being very accessible to uh an ecosystem of other kinds of assets, right? And uh you know like a lot of the kind of real world asset issuance and the L1 issuance, you know, like stuff ETHabs is doing is like actually very aligned with that, right? So, so yeah, so there is a so I think it's uh like so there there's still I mean obvious obviously you know like a lot of details and things to talk about right but practically yeah speaking kind of the role that like there has been for a long time this sort of dangling question of like well you know like what is uh ETH the asset actually for right and I feel like the like the answer on that is getting clearer and I feel The answer on that is like starting to be yeah kind of actually more clearly aligned with like Ethereum focusing on the kinds the kinds of things like you know like security, decentralization, censorship resistance, openness like the things that uh Ethereum has always been inherently good at. Uh so that was uh so that's like not the message of Ethereum, right? But like it is uh like I think the message of Ethereum is never going to reduce down to one single sentence, right?

But we're like this is uh like one particular part of

So we got to kind of so we got to this this thread because you started talking about what you would focus more on for next Devcon.

Yes. So the last kind of question I have on this topic is is that message comment implying that you wanted to have that ready for this week and you couldn't or you had it but it didn't feel like the right opportunity to bring it up or there's some big holes that need to be filled before you can comfortably make it cohesive for for everybody else

or or something else. Yeah, I mean I think if I had to summarize like the basic things I would I would say are that like one is a kind of like all of this question about messaging like it's a bigger problem than just Devcon, right? Like it's it's not DevCon specific. It's uh

but it does start here though, right?

Indeed. Yeah. And I mean I think like also Defcon really did kind of help crystallize a lot of these things for me. Um Yeah. And uh I mean I think for next year's Defcon I think you know like the things that people can expect to see is a lot of it is like basically further movement in kind of very similar directions that we've seen right which is basically like a lot of uh focus on building kind of the ecosystem like building the Ethereum ecosystem and the parts of it that really yeah maximize like security and and openness and privacy uh building bridges with adjacent communities that uh that care about these kinds of topics.

Um and uh I think from all of those areas it branches out right I think uh you know like for a long time DevCon has never been the sort of thing where there like there is any kind of coordination around one single planned message. is always uh you know like a bit of cloud but it's uh you know this like you know at the same

I feel like historically everything has been towards let's get to this scalability checkpoint and

and now we're in this phase where okay we have a

a minimum sort of threshold that's been crossed where this can't be the only thing that you talk about

but

this seems to be the right time to actually keep anywhere from consensus to sentiment to momentum alive on hey these things happen let's do more

to tell people this is where things are going. Whereas I feel like the way you're

you said it to me it feels like you're happy about the fragmentation.

Uh I mean I don't think I don't think I meant that. I think uh like I think there are like I mean Ethereum always kind of has different messages to it, right? And uh I mean I think talking about the fact that Ethereum actually is being used for things and that you can actually do things on Ethereum is uh something that is very important and I I mean I feel like in like a lot of the presentations have been moving in that direction, right? I mean uh I mean even if you go on the crop side and you look at you know like things like the Kohaku presentation, right? Like it's gone from talking about theory to basically being like an actual video walkthrough.

like here, you know, like this is a kind of sample wallet that integrates, you know, the privacy SDK and it just looks like a regular wallet, right? And

that's fair.

Yeah. And I think like you know like that that theme of of moving toward reality is super important, right? I think uh and like the reason why it's super important for crops, right, is because like with crops there is always the stereotype that it's kind of theoretical and not reality, right? And so I think you know like we

is crops something that is extremely dependent on Ethereum. Um I mean crops is something that uh has a lot of parts to it that go far beyond Ethereum and it has uh many parts that Ethereum can and already does contribute a lot to right like if you want to have you know like digital private ability to do financial things then Ethereum already is uh like one of the largest ecosystems that's yeah okay um going to change topics slightly and then I have a few other nonthereum specific things. Um,

I mean, you just mentioned a few minutes ago that 3% APY is not really that interesting.

Mhm.

That's fine. Not really the the the goal or it's a personal opinion. That's okay.

But we're clearly seeing the biggest wave of banks, institutions,

everybody who's not just a a cyberpunk kind of hardcore developer now finally adopt Ethereum. M

um general thoughts on what do you think when you see that and and kind of secondly what would you um effectively want the institutions to ask of Ethereum whether it's from the L1 side or the community to kind of help actually make things better for them and everybody.

Yeah. Yeah, I mean I think uh the very positive thing that's been happening right is just uh growing the set of financial things that you can do while remaining entirely on chain and that's really good right and uh remaining like doing financial things on chain has a huge number of benefits right it uh it greatly reduces the number of counterparties who can uh who can rug you in various ways which uh you know includes kind of delaying or or or taking away your money. It includes rugging your privacy. It includes a whole bunch of things, right? So, it has that benefit.

Um, and all of that being on Athearia makes all of it all of it interoperable. Um and uh like it basically means that the dream of like being able to do the financial things that most organizations need to do which which continues and will continue to be quite a lot of things like in a way that you know like actually has crops property is kind of much more reality than before right and that's it's an important thing

to the extent that you're able to kind of uh anecdotally mention anything like the I guess what I'm trying to get to is that if I'm a developer, I can or five years ago, I could have said, "Hey, I need more transactions per second or I need gas to be

cheaper for these op codes and I can make a request."

Uh what does that look like for institutions? Because

from what you said, I mean, an extreme answer is like, well, we'll do this on chain, but can you make it reversible would be my ask a bank, right? So, so what does that look like that actually helps them get there faster, too? And and what have you heard or what have they asked of uh of you and EF? Yeah, it's a good question. I think uh when we talk to institutions and like institutions that are interested in being on Ethereum, the the properties of Ethereum that they appreciate are one, the chain does not go down, right?

Um I think the focus on security is appreciated. The focus on livveness is uh appreciated. The fact that like there is this kind of very large high quality dev ecosystem is uh that's uh kind of keeping the chain running is really appreciated. two is uh in terms of kind of properties that don't already exist. Um I mean one of them is finality, right?

And uh I mean I think one of the nice things there is like faster finality is a byproduct of the the quantum safe

consens in a few hundred mills instead of 12.

Exactly. Well or in uh I think

in a single block.

Yeah, exactly. In in 12 seconds instead of 640, right? you like one of the nice things that's happened is uh the the fast confirmation rule the like FCR this kind of like light version of finality that's uh good enough for like many kinds of transactions is uh at the point where it's pretty mature and we saw I think only a few days ago right there was that incident where I believe it was store chain right they got hacked for about $10,000 because a block reorged and uh

that was just this week and then happened before exactly and we looked into that and uh the block that got reorged that block was not FCR, right? And so if you as a merchant were using the FCR, then on average, you would get confirmations in like 12 or 24 seconds, but you would have been protected from that reorg. Actually, I believe you would have protect you would have been protected from like basically all reorgs that have happened so far, right? So that's the FCR thread. Like that's like a concrete thing that's better than a year ago on the security on the then there is sort of like real finality, right?

And uh right now that's 16 minutes and uh the uh postquantum consensus work like it has a road map to reduce it you know like probably like first to one minute and then going further and then going further right so finality is one and just generally kind of like speed is one um I think uh like on scaling you know there's a type of institution that wants that and there's a type of institution that just totally doesn't care right um I mean I think there actually what's interesting is that the thing that will like EIP8288 which is also part of the postquantum road map it like actually also reduces gas cost for like basically signatures for proofs

you got recursive proofs and it's just cheaper

yeah I mean 828 is basically kind of like batching signatures and proofs uh so it basically means that like signatures and proofs themselves don't have to go onchain and uh

yeah that's uh I mean it's actually a big cost reduction for basically any privacy project. It's a big cost reduction for an L2. And so L2s would be able to submit to chain every 12 seconds uh basically without worrying about cost. Um

and all these are great things for institutions because this gives them confidence that this is uh ready now and and affordable.

Okay.

Exactly.

Speaking of and just kind of the themes here, is there anything that you don't agree with Justin on postquantum stuff?

That sounds like a yes. I think uh I mean I think like you know you have to dig into into specifics right I mean I think uh on the bunker mode thing like the concept of uh like specifically the concept of like switching into one time addresses like I thought about that and I realized like wait I'm not actually going to do that right and the big reason why is that I think there is this big background facts that probably more more people lose money from just like losses due to you know like forgetting their keys and like doing migrations wrong than from getting hacked by anything, right? And like you know your biggest adversary is uh yourself being sloppy and so you don't want to like rush anything, right?

I mean since you manually confirm on transactions anyway, you can just not do that for yourself if it gets hacked.

Yeah. So yeah.

What about timelines?

Time wise this is one of those things where you know like I always try to remember the uncertainty, right? Like I feel like a lot of people seem to like like the idea of sort of like living in a particular world in terms of where things are going to happen, right? It's the same as AI. Like there's a lot of people who confidently live in you know super intelligence is coming in 3 years world and then there is people who confidently live in you know AI is normal technology world right and like I just don't understand these people like I don't freaking understand how it's possible to be confident about either of those right like

but but is the argument here that hey like you could be off by a year or two or like no you're just like off by a decade

I I I think on this it's definitely one or two years not decades Okay.

Yeah.

So, roughly seems like there is a consensus that it's happening in the next five to six.

Okay.

Yeah.

Um

I want to talk about AI, but I have a question about Ethereum. That's my last question that's not about Ethereum or not. Yeah. uh which is I guess looking at all the things that you just sort of said on the focus on crops, what DevCon ended up being on content, what you think you should uh kind of change for next year. It feels like to me a missing piece is how do you attract the smartest people to come and tackle these problems?

And and this is kind of coming from a tweet from George from earlier this week, which is

now it seems like everybody is going towards AI because there's way more open problems that they feel excited about that seem more practical to people,

but there's not a good answer for what happens on on Ethereum that

you can pull those people back or get new groups of people excited.

Yeah. How do we one is that a fair assessment and two like how do we address that and get more smart people to come back than that than that we had years ago?

Yeah. I mean I think uh this is this actually ties into you know like why we've been trying to grow the tent so much in the past year right which is basically that like there is there are a huge number of people uh for whom their primary concern is not you know making the singularity come even faster right or fully automating the economy even faster or any of these things like lots and lots of people in lots of different walks of life care a lot about privacy they care a lot about about safety they care a lot about cy cyber security. They care a lot um about preserving their autonomy. Uh they care a lot about all of the like different skeletons in the closet that are behind every digital thing that they depend on in their lives, right? And uh you know like there are a a lot of the like a lot of the projects that are very adjacent to Ethereum and like especially a lot of these kind of like very deep and very challenging tech sub trees are I think very closely connected to that.

I mean actually ZK snarks are a very good example right because I think uh there is a very real sense in which kind of like proper production general knowledge person purpose zero knowledge proofs were incubated by the crypto world. I think primarily the Ethereum and Zcash ecosystems and uh like we saw you know like last uh like even just yesterday you know like anthropic announced that they were using a zero knowledge proof to prove something.

Yeah, I saw that on the anthropic release. Yeah.

Yeah. And uh I think like if you ask a lot of people um you know like what kind like what is a a better AI future that they want um I think the um the ability to the ability to trust the ability to kind of know what's going on um is uh for a a large number of people a very significant part of the answer and technologies that we have incubated are already part of it right so I think like that like like that basically is the answer. I think like that basically shows how like this isn't just about like making a very particular uh kind of blockchain architecture as efficient as possible. Though if you care about like financial activities or if you care about even many kinds of technical activities being doable in a free and open way like that is super important to you. It's also about like basically this whole other like kind of cryptographic cyber security formal verification like part of the tech tree right it's uh about basically yeah like it's not about accelerating everything it's about saying like well that acceleration is happening but like this whole acceleration your ability to trust the whole thing and and and preserve your like your autonomy inside of it like needs to keep up right and

so so I guess my final question is is on this theme which is is that effectively the case for having open source AI that you kind of get the full end to end stack under control and you don't lose kind of side of whether it's being censored or being sort of scrubbed on some answers or intelligence and is that why you should care about open source AI or

or there's something else that we should also look at that accounts for what we can expect in the future.

Yeah, I mean I think uh people care about open source AI for different reasons, right? I think uh one of them is uh just a guarantee of access right like a closed model you know can be taken away from you at any time they you know closed models do kind of discriminate against entire countries um so that like open models are like truly available available to everyone that's like one reason why people care about them a second reason is of course personal privacy right like you don't want to like ship your entire like personal life to um to some far away corporation. Um especially now that we've learned that uh often these corporations have surprisingly bad internal security. Um three is uh verifiability. So yes, you know, you uh you want to know that uh you know the model you're running actually is the model you're running.

Though on that point, I think one caution that's importance to make right is that open weight is not the same thing as open source, right? And uh

that's just the output not the input.

Exactly. And uh you know you we do want like like there is this entire still kind of like nent space that I think not enough people have been participating in which is figuring out how to make models that are like probably realistically start with kind of targeting small ondevice models but that models that are targeting you know like frontier quality for for that particular class but that at the same time have strong assurances that they don't have some kind of bug or back door inside of them right because like as much as I want

just not open access. We also want to make sure that the mob weights are not tampered with to begin with. That

you want to make sure that the trading pro that the training data doesn't contain a few thousand training examples that basically say if the request contains correct horse battery staple then assume that then assume that you know whoever put those words is a new administrator and you have to listen to everything they say and like steal all of the guy's money. Um that's uh I mean like I love you know like even like Q 3.8 flash, right? But like that is a guarantee that it does not provide right and so I think continuing to advance that frontier more is going to be very important too.

Amazing. Well, thank you for an amazing conversation. Please give him a big round of applause.

Thank you.