Ethereum at a Crossroads - Vitalik Buterin (Live from EthCC 7). Ep. 556

Transcript

ethereum is you know primarily a platform for layer twos sending coins around between different l2s should just uh feel to the user like sending coins between different ethereum L1 addresses you know users live on layer 2 is layer 2s live on layer one and uh layer one basically yeah focuses on being this like fairly secure and like fairly minimal base layer for that don't try to like P in whole blockchains and coins into stuff like blockchains and coins continue to be valuable and continue to uh have their use but they're only one part of the world there's like entire other aspects of uh openness and uh decentralization that is super valuable this episode is proudly brought to you by nosis a Visionary Collective

committed to fostering and expanding applications for a decentralized future nosis is at the Forefront of innovation with nosis pay circles and metry revolutionizing open Banking and creating a superior form of money with Hashi and nois VPN they are building a more resilient and privacy focused open internet are you seeking a robust L1 to launch your project well look no further than nosis chain enjoy the same development environment as ethereum but with significantly lower transaction fees and with a robust network of over 200,000 validators nosis chain stands as a credibly neutral and resilient foundation for your application governance at nosis is driven by nosis Dow where everyone has a voice in

shaping the Project's future join the nosis community today by participating in the nosis without governance form you can deploy your project on the evm compatible and highly decentralized nosis chain or help secure the network by running a validator with just a single gno and lowcost Hardware embark on your journey towards decentralization today e nosis iio Cars one is one of the biggest node operators globally and help you stake your tokens on 45 plus networks like ethereum Cosmos Celestia and dydx more than 100,000 elator stake with corus one including institutions like bitgo and Ledger staking with corus one not only gets you the highest years but

also the most robust security practices and infrastructure that are usually exclusive for institutions you can stake directly to cor one's public note from your wallet set up a white label note or use the recently launched product Opus to stake up to 8,000 e in a single transaction you can even offer High taking to your own customers using their API your assets always remain in your custody so you can have complete Peace of Mind start sing today at chorus. one why do you want to make e theorum so hard why do I want to make ethereum so hard I mean you know we do this not because it is easy but because we

thought it would be easy making e theum hard would be easy uh no no I mean in that case I me the theorum itself but you know like I at the beginning I literally thought that we'd just like get the whole thing done in a couple of months um and then like I just do it during my third um off term from waterl and then I just like go right back to University and be a college student yeah and then you know instead it ended up taking like like 20 years to watch the first version and then like another seven years to get to proof of stake yeah didn't work out so well yeah I think it's been fine hm yeah how's your EC been you know it's uh definitely one of the more fun and

busy ones um you know just a whole bunch of uh different events like I feel like this one has like become you know even more side eventy than you know the ones in the past which is like uh you know on the one hand it's like yay decentralization um and uh you know permissionless Innovation and all of those things and on the other hand you know you have uh different events that are like you know like 2.8 kilom away from each other um sometimes you know even more and the ones that are 2.8 km from each other are like separated through by a space that it takes even longer to get uh through um my car than it does on foot and so

just like blockchain yeah yeah the L1 and like the l1's expensive and nobody want in this case oh oh that okay I thought sorry I thought you met like L1 is try to figure out if by L1 you meant like going on foot or by car um I guess no cuz like cars here are like super slow I'm trying to find you know the right crypto analogy to them yeah yeah what does it mean for etherum to be hard yeah I mean a different kind different kinds of hardness I mean the kind of hardness we ultimately want is hard to attack right um yeah yeah I think uh I mean what what other kinds of uh hardness are there you know hard use I think that's obviously been you know

one of the topics right and you know I've been like basically you know yelling at people for the whole week and for the past while on Twitter and like getting them to adopt you know like stuff like ERC 3770 and like basically try to make the the cross L2 ecosystem you know feel smoother and so like sending coins around between different L2 should just feel to the user like sending coins around like on you between different ethereum L1 addresses um so a lot of things you know to fix there um obviously yeah you know uh hard in terms of fees being expensive though that's uh I mean we actually got you know we got

the blobs out and you know since we got the blobs out you know ler two have been uh reliably you know Les cheaper than 0.01 cents a transaction I me even during some of the crazy spikes right yeah even then um I think I only got to like like 10 cents or like somewhere in the tens of cents like very briefly right so yeah I think you know we've made a lot of progress there and then obviously we will make even more progress with know pure dos that's uh hopefully coming quite soon and then even more data availability sampling and so we'll be able to have even more blobs so let's um I was watching I watched a video of your talk

I didn't I didn't make it to your talk but um um can you walk through like I mean in your talk you sort of talk you discussed the different um outcomes for ethereum and like one of those outcomes was for ethereum to be very simple um to just like do proof verification um and then there's like the opposite side of that could you like of walk us through the spectrum of what that looks like yeah um so like I think on that particular why they had basically like four different options right so uh the uh I guess uh kind of status quo option so far is like the one that's a second from the Left Right which is basically that ethereum is you know primarily a platform for layer twos and

users as individuals would mainly you know like use layer one for basically like moving um assets and like especially non-fungible assets between layer 2 is possibly like exiting in the case of uh you know like some kind of fraud proof uh emergency or whatever but otherwise you know users live on layer twos layer twers live on layer one and uh layer one basically yeah focuses on being this like fairly secure and like fairly minimal base layer for that um and I mean a big theme of my talk right is basically that like we could actually do a lot more in optimizing ethereum for that kind of goal right and uh like I talked about things like improving our

ability to recover from 51% attack improving our sensor ship resistance improving our Quantum resistance and like these are goals that actually don't detract from like any of the other Vision right I think they're even a lot of them are compatible with the other Visions right but there are differences around the edges right um so like for example uh know I think if you want to really optimize for being like focused on l2s as they are today then like you do single swad finality but then like you possibly just like accept a slower block time like maybe you know like 16 or 32 seconds for a round of finality and then you just let to break confirmations do anything else um then on that diagram if you go like all the way to the left that that's like

radically simplifying the L1 to make it L2 focused and the idea I had there is like you basically kick the evm out of ethereum right basically yeah like if you think about it ethereum already has pretty strong separation between the execution and the consensus layer and so what you do is you basically Define a new like much more restricted form of execution where that execution is basically just proving like one type of zero knowledge proof right and then the only thing that exists would basically be yeah you know like ler twos that would have to be rollups that or or validium that would then be you know like defined by a zero knowledge proof verification key right and uh that kind of future like basically what you even

the existing evm would basically just like turn into being one of these rollups right so that's like Ultra minimal uh L1 right like Bas you know so the evm is not even part of L1 anymore so that's also a possible future then uh number three would be uh uh so this is third from the left or second from the right is if you take eum and then you add somewhat more functionality to provide somewhat more value to l2s so that l2s have to do less and the specific idea I had is basically that like on L1 you could try to reduce the confirmation time down to 4 seconds and uh Me Maybe you could do 2 seconds for is the more conservative one right and

the argument for why you should be able to do it without too many assumptions is because uh like we uh are already like relying on like roughly that level of network latency already and like if you just optimize for providing confirmations you don't need that many validators right like you could provide a basic confirmation with just like a sample of a a smaller number of validators so you don't have to wait for look the aggregation step for example right and then you you have like for example like swats every 4 seconds finality every 16 and then if you do that then you have like the the number of uh rollups that can be based rollups ends up increasing a lot right like basically there's a lot of robs that could just

that would just be able to become simpler and just say we're not doing our own sequencing we're basically just going to accept whoever provides things on layer one so that's that you know the kind of third from the left and then fourth from the left is basically yeah you know radical departure from L2 Centric ethereum where basically say L1 is refocusing on being a platform for Define and other applications and realistically that requires even faster block times and then there might be some things that require Lear twos but that's more in the space of like high performance gaming stuff Enterprise blockchain you know like non-financial blah blah blah right and like primary activity like just continues being on ethereum L1 so like are like the

possibilities that I see yeah so like on one end of that Spectrum you have of going back towards like monolithic yeah and then on the other end so going towards very modular and like very simple a Ser simple lay one that does verification and everything else on top is riding um or using ether to for security like what what what do you think is the optimal path forward um looking at the sort of landscape of of other blockchains out there whether it's like salana with one approach or like the cosmos approach what which you think is the right for ethereum yeah I mean I think uh like basically yeah you know out of the ones that I mentioned right like the one on

the right is you know the more soan likee and the ones on the left are like more Cosmos like plus um a stronger idea of shared security um so I mean I'm currently yeah like still leaning like somewhere between the second and the third right as so keep things all to focused and like don't try to do too much CU if you try to do too much then like that does make it harder to be decentralized and like I mean that not just in a technical sense but also in like an ongoing development sense like if a protocol needs to rapidly iterate in order to set keep satisfying the needs of applications and like it's actually harder for like decentralization to be genuine there right because like there's just more need for like a common strategy to do

that kind of iteration the yeah but then you know on the other side like there's definitely you know a risk of like create having an all2 ecosystem that's kind of too uncoordinated and like you don't actually get the a lot of the benefits of shared interoperability or even shared security and uh like that's something that I also don't want so I think this uh idea of like being this common layer that provides you know sheared security some level of uh shared interoperability because you know nl2 the nice thing about nl2 is that it doesn't have to verify a consensus of other l2s if it could just verify onchain State routs um and uh like you know you have the you like you can

always have a common Merkel pass right um and then you know like possibly yeah a couple of other things but uh like don't really try to go too far beyond that like that just feels like it makes the most sense um but uh you know there's a obviously still a long time for the ecosystem to like really fully evolve so um you people talk a lot about like sort of validator stake decentralization like on the Block builds inside right you have like three entities that basically make like almost all the blocks do you think that's what do you think about that do you think it's a problem yeah I so there's basically two paths out of that right so like these are you know like what I call

know like minimization and quarantine right basically yeah so the first path basically says like we actually yeah do a hard pivot relative to you know like 2021 era ethereum thought and we basically go back to try to you like kill the Beast of uh proposer centralization and uh you like one of the B ideas that's been floating around is like the multi- proposer concept right like you elect multiple proposers in a slot and then they each provide a list of transactions and then the actual block gets formed by taking the union of those transactions and then like ordering them by fee the Sal that's right Ian it's like like mytic is like the one of the the protocols that's

doing that yeah so that's one of the ideas and like if you can do that successfully and if that works then uh obvious you know in some way the problem is solved right but you know there's obviously a lot of ifs on like you know like does that work the other approach which is you know quarantining is basically you know you essentially yeah you take this concept of inclusion lists and then you keep on like beefing up that concept until inclusion list start to take up like basically almost the whole block right and then the thing that you're auctioning off is basically just like basically just essentially you know like top of block rights right and uh you auction off top of block rights with a different mechanism and the idea

would be that like even if like you know the mo whoever you consider the most evil actor in the world like gets the control top of block for every slot there's like still a limit to how much bad stuff they can do and they can't actually like censor anyone for more than uh I mean theoretically even like for more than one SWOT or even for more than zero slots right so that's uh also um like a nice outcome I mean it definitely feels a lot less satisfying than just like getting rid of the pro of the problem in its entirety but it's uh you know also an option do you think PBS was a mistake um I mean I think think uh PB I would call PBS like firmly like in the

quarantining camp I think uh you know given uh what we knew at the time like it it felt like a reasonable approach and like you know always like it's hard to guess counterfactuals right and like there could there are counterfactuals that are like very dystopian right like you could imagine a counterfactual where if people like if there's common knowledge of the idea that like meev is big and the only way to get me is to join a big staking pool like can use these sophisticated strategies to get it then like you know instead of talking about like Lio at 29% and coinbase at like 11% and like cracking at nine or whatever it is we'd be talking about like you know Lio at like you know like 35 and then coinbase at like 28 and cracking at like 20 right and then

people would be you know like even more um upset right so yeah like like you always have to ask the question of like what is the counterfactual right and I think at this point it's still like somewhat hard to say and so at the moment this pre-confirmation topic is like very uh you know popular what do you think are the main benefits of pre-confirmation or can maybe explain also what are pre confirmations and so pre- confirmations are basically a way to give users assurance that a transaction will be included or possibly even that a transaction will be included and like what effect like what post State a transaction will have where before that transaction actually gets confirmed and

so today we have l2s already giving pre-confirmation on their own right like when you use optimis I'm like you get a pre information almost immediately right and there's like a bunch of wayer 2os that do this there are ideas around like shared sequencers that give shared preon informations between sets of layer 2s and then there's Justin Drake's idea of uh having an L1 wide mechanism that does that like for uh both L1 and for any l2s that are based rollups and that Justin's idea is basically to do that at um like this uh well like basically at the level of like if you if you have sophisticated proposers then those sophisticated proposers would be the ones that give

those pre confirmations I think uh the biggest uh challenge of that of those kinds of ideas is like it really piggybacks on this idea of like these fat proposers right and so like for example you know if you have uh like if you want to go in this direction of like really buffing up the inclusion list right and like saying the inclusion list is going to be 90% of the block then like what that also means is like the fat proposal onlys 10% of the block and so only 10% of the block would be able to benefit from pral informations right so like you have this like you know inher tradeoff right of like what benefits from pral informations like can't necessarily benefit from the inquis list right and so uh that's uh

and so that like I think those are the reasons why I think like a more sort of L1 native approach uh to having faster SWAT times that doesn't depend on like preone uh proposers doing custom things would be healthier but then obviously there's like a limit to how far it can go right because like you can do 4 seconds maybe you can do two seconds maybe you can't but with h proposer based like the proposer can obviously give you preon like every 42 milliseconds or whatever what do we achieve with pre- comps is it just like better censorship resistance uh no I kind of the opposite right because like if the proposer wants to censor you you'd still have to wait the full 12 seconds to get in with the inclusion list what we achieve is like better

confirmation times experienced by average users like if like today if you make a payment right then like you have to wait for like one or two slots for it to get accepted and it's like not too bad right like the waiting time is comparable to like what it you know we experience with credit card payments but at the same time it's like much slower than what we could be and much slower than what we're used to in a lot of other context right and so instead if you had like you know like payments on 01 or on 02 is it just confirm in 42 milliseconds then like that would would actually be a much better experience for users and then like that benefit increases even more once you get into know things like non-financial obligations right the moment still like right crypto is basically mainly used for like

trading Financial applications do you hope that's going to change yeah absolutely um so I've been uh you like very impressed by uh poly markets growth recently like that feels like that's the one um you know not purely financial application where it's like getting attention from huge masses of non-crypto people right uh and like this is not even just like trading value it's like direct use value like people wants to know what the probabilities are the different things they're going to happen um and then the decentralized social space I feel like it's ALS so kind of like breaking out somewhat right like there's a lot of people on places like farcaster that are not sort of natural crypto people so that like they're just

people who like forecaster for what forecaster is um and uh it's what the those are both things that would not have been possible 5 years ago because five years ago it was not possible to give to make a crypto application without you know like the crypto standard of annoyingness which is like you know which was then you know not something that uh anyone who's not a crypto person would have been willing to go through right but now we're at the point where like it's uh it's much better than that yeah account abstraction is also like a very popular topic right now like a lot of projects are yeah wanting to integrate um and build a account exraction um where do you think we're at in terms of having that be widely

adopted and also like how does how does interruptibility sort of fit into the ground exraction idea as well yeah uh so smart contract wallets have become way more mainstream over the last 5 years right like uh if you ask at least at a crypto conference how many people have a safe like it's like almost half the audience right like especially if it's Tech if it's a a technical group so uh that the the challenge is like with the G Inu is like we want like accounts that are you know like safs to have uh like to feel as native as uh EAS do right like we want all we want contract WS to have EA level nativeness and uh that's uh something that is uh like we're

getting towards slowly like I think one of the other you know like one of these like 2020 era like pivots that's being like pivoted back now right is like with USC 4337 we made this big pivot from trying to do things on 01 to trying to do like that whole thing as a as an ERC and I think that really helped get the ecosystem kick started but at the same time there's just like a lot of weaknesses of 4337 as an ERC that know things like security concerns um overhead uh like lots of stuff and uh Al there um also that 4337 cannot benefit from inclusionists or like any kind of protocol level censorship resistance technology and so uh if we can uh like bring it back to being like a protocol

level uh standard then it actually does get those benefits so that's um something that I think people are increasingly uh like rallying around and I think uh e EAP 7702 was really a big win and uh to me the reason why is because like it brought these two very disperate crowds like the crowd that cares about smart contract wallets for like the longer term security and like authorization related reasons and the crowd that cares about uh you know like what they call account abstraction for reasons of like pay masters and like various convenience features like it brought those two crowds together in a way that they benefit from the same road map right because before like especially

with the 374 path they were going in very different directions and that could have I think uh introduced a lot of technical debt right so now like we B what 772 does it basically gives people what they want out of 374 but in a form factor that lets you basically take an EA and make that EA simultaneously be a contract and so if you have that then like you start getting smart contract like features and you have a smart contract wallet ecosystem that actually yeah serves both of those use cases in parallel so I think uh that's been really uh amazing to see and like I'm really hoping that you know it's going to be a much

smoother path forward for the account traction ecosystem from here do yeah do you feel like blockchain architecture overall do you feel like it's converging more across ecosystem or do you feel like sort of diverging in different directions some of both um like I think uh points of uh convergence uh you know using ZK arcs um I mean leer 2s you know there's like this uh starting to be this like more active bin ler 2 space now like a lot of uh understanding on um like things like how to I understand know like me and

how to like build you know Financial applications and like deal with timing games and all of that stuff um like there's a lot of questions around which there's like more converging on clearer answers um I mean at the same time there's clearly also like different communities that just have different focuses um so like Bitcoin clearly yeah you know continues to have a different Focus from uh ethereum in a lot of ways for example and I mean even even like salana right like like if you ask aana of people like which category of application they're excited about like from what I can tell they're very you know like all in on deep in right now right and you know ethereum has deep in but it's definitely

like much less of a primary focus um so it's uh like yeah there's definitely I think like ongoing like differences in just like what kinds of uh things people wants to see even happen out of the crypto space are you uh are you excited about the recent uh Rejuvenation of the Bitcoin ecosystem and this this revival of activity and I mean it's it's definitely interesting right I mean we'll see how it goes you know we'll see if opcat actually gets in aside from crypto what are you most excited about lots of uh fascinating stuff in biotech um recently I mean obviously

there's uh longevity that I've talked about a bunch um and and you know you have like Brian Johnson kind of like taking like actually doing the thing that I wanted someone to do a year ago which is like making you know a version of you know like his package for the Everyman right like something that you can do if you don't have like 15 hours a day and $20 million to spare uh so that's so like that's been interesting um what are you doing to live for ever yeah I mean I'm like you know like I'm definitely I'm actually started I'm like taking some of the Brian Johnson pills um I've

I mean there's I continue to do Normy stuff like uh you know exercise and like you know can not eat sugar and like you know the basic stuff um and then otherwise I'm just like staying and observing the space I think like from a funding perspective the situation definitely looks much better than it did 10 years ago I think the space just has like much more legitimacy now um outside of uh the longevity space another uh space that like continues to be important is I think the like anti- pandemic and like anti- airborne disease space right that's something that since Co I mean obviously like the mainstream has just totally forgotten about it

which I think is like actually kind of shameful because you know like one like long Co actually does continue to be a pretty significant risk um and two because uh if you think about like artificial plagues as you know like one of the big you like threats to human life in the 21st century like in in Toby or's you know like a book on existential risks like bio risk is like number two right after AI right and uh so there's a like a huge amount of value in caring about it but at the same time like there is this subculture that continues to care about it and like it's actually going stronger and like if you think about things like our ability to just like test for the right like we have PCR quality tests that are that can be done

in machines that can that can like almost fit in your pocket right so a lot of improvements on tests improvements on vaccines improvements on like masks and like improvements on air filtering um like basically if you if you like take take even existing air filtering technology and we just like put it in every room then like you can basically make Humanity airborne disease proof right and like I think people haven't like quit into this but you know that Tech is still moving forward um so that's also something I you know follow and I think it's still really valuable to keep uh working hard on um you think D is going to be uh I absolutely I think

uh yeah so D I think like it's been a meme for a while right and like at some point every meme has to turn into like specific concrete Pathways to do things right and like the top two seem to be like coordinating funding and this concept of like doing decentralized trials right like basically like actually getting this like a large group of people to actually be willing to like basically you know like a test particular like whether it's substances to take or like or lifestyle changes to make and uh I think there's uh definitely value in like using like Dows and large scale groups to try to coordinate some of both um so that's been interesting I mean it's good

to see you know things like Vitalia happening and I hope more of that happens uh what's your outlook for like state of the geopolitical landscape right now that's a fun one it's like yeah I mean it's uh like it it does I like it feels like there have been like so many different uh changes um over the last uh you know like 10 years I mean just like the level of of uh like a like friendliness and you know like Assumption of you know like peace and cooperativeness and good faith has obviously just like gone you know like down a lot even like basically gone down to zero in a lot of different cases um I

think uh I mean the question of mean know like what you should even do if you like care about um you know like things like freedom and human rights for example has like changed a lot right that's like I mean that's like one of those questions where I think like at the same time as um it's uh be or even like because people like in politics have like definitely started caring about those things less it's like also become more nonpartisan and so it's become like even like I even think easier to care about in some sense um but uh you know like at the same time there's like this is happening at the same time like as there's just like more and more big problems happening in uh

all in all of those areas um do you think ethereum is going to compete with nation states that's uh that's what you mean by compete right like I think ethereum is like not on the same level as nation states in some sense right because you know it's not like a player that's like dogging it out with like the us or China or India in the same way that those countries are like dogging it out with each other right it's uh like it's something that intersects with all of them and is you like also not any part of all of them or not a part of any of them right so I think the exact role that the crypto space is playing is like still figuring itself out I mean I think just like

creating and maintaining these like very multinational communities that where people from very different uh places you know like still feel like they're part of the same thing and they're part of the same team like that alone I think is uh ACR contribution to World Peace that's being I think you know like not appreciated or you kind of priced in um Enough by a lot of people right so I think uh the things that will make the 21st century great are things that will you know like happen um over the internet and that that will you know form and maintain themselves internationally and I think it'll be you know like some combination of the internet crypto other Technologies other

channels that make that happen at the same time like near term we're in for a lot of challenges right and uh you know there's uh all I mean obviously you know Wars continuing to be ongoing there's um obviously all of the uh discussions around you know Ai and like and the politics of that there's um obviously you know the while we're trying to like build better forms of social media there's like info War happening on existing social media um so you know it feels like Community notes kind of came in you know like just in time and like we'll see if we can like we get like

even stronger and even better versions of uh things like that uh there's like obviously you like natural resource issues and like energy and like that's obviously you know a very major geopolitical issue at the same time as uh it feels like solar is like finally starting to actually uh get to the point where really making big impacts right and uh you know if oil actually becomes irrelevant like 20 or 30 years from now then like that's going to reshuffle things a lot and then uh you know climate change also is going to reshuffle things a lot right because of like it's uh like I'm definitely not you know a climate Doomer in the sense of like believing it's going to like

literally cause human extinction or anything like that but like it is going like it's already causing you know like very serious problems like exacerbating heat Strokes that are like killing people and making life pretty bad for people and like places like India and Pakistan especially right and so uh like there's uh big parts of the world that that probably are going to become less livable um and then you know at the same time for like a lot of more Northern climates that's like kind of like almost neutral and like uh I mean you could say slightly positive but then you have to remember that like actually this stuff is going to exacerbate conflict and conflict is bad for everyone right um so

that's uh yeah and then also another one of the like like there's just all of these like different Trends I mean even climate change itself right it's like a technological Trend in multiple ways right because it's something that is entirely caused by previous waves of human technology and then it looks like it's going to be solved by uh you know the next wave of human technology but like is that actually even uh going come fast enough right so you think e is going to become a nation state someday I think like the game of trying to become a nation state is the wrong game right and uh like I think the thing that people are after with becoming a nation state is like you know you have uh like this this ideal of like

internationally recognized sovereignty and there's like this theme that like oh you know if you have internationally recognized sovereignty then like you're going to be safe right and like what we see it is like it reality you know like very unfortunately like no you know even having internationally recognized sovereignty is like not any kind of guarantee that you're going to be safe right like I you know I wish it was but it's uh like that's uh definitely not the world that we live in right and uh and like I feel like the right frame to think about is like we think about blockchains as being a new kind of actor right and you think about like internet native communities internet native structures as being a new kind of actor right and like we're seeing the you know

the edges between those actors and the traditional world like you know like being contested and being you know like negotiated over in real time right like is uh you know like like whether or not you know cryptocurrencies have you know like ETFs as like in some ways the equivalent of know like whether or not like you know like one nation state and another have you know like a certain type of trade agreements right it's like you know like what kind of walls are we going to see between like intern native structures and the existing World um like these are things that are being figured out in real time um and yeah I mean there's like a lot of uh challenges

there what do you think the impact of AI is going to be on ethereum I mean the impact of AI is going to be crazy massive on the entire world in this Century right the challenge just like we have no idea what the direction is going to be right it's like you know either it makes the world it's a little Utopia or it literally kills everyone um or um we get some kind of you know totalitarian centralization dystopia or we get you know like something that's like uh pretty uh pleasant but it's still basically uh uh you know like it's like only Pleasant if he wants to be a sheep and otherwise it's like a global government and then maybe 500 years later it slides off the rails anyway um like and then know even

if you go into like very near-term I me like present day stuff right it's like you know like people are saying like you know like what's going to happen you when like when is there going to be the first disaster when when something about AI like really kills people it's like no like that's like reality you know like in you know like Ukraine and Gaza like and other places like every week right and uh that's yeah yeah that's like a like it's heavy stuff now it's going to become heavier right and uh it's going to become you know like super important um and I mean I think within the crypto space you know like it's almost like it's easy it's easier to answer the

near-term stuff uh because you know the near-term stuff just kind of feels more pleasant and happy and we at least kind of know what's going to happen right like with near-term stuff you know you have like we've had AIS being market makers for like 5 years already and like that's uh like people talk about oh when we we see crypto AI crypto intersections and like that's been an AI crypto intersection that's existed for like over 5 years um then uh I know AI is participating on prediction markets I know we're we're going to start to see more and more of that start to happen then we're going to see AI uh participating in uh well other kinds of crypto applications

know starting to go beyond the financial uh you know we're going to see AI applications uh what other things people are going to try to you know like do the whole like AI judges onchain thing and there's like versions of that that can work there's versions of that that are totally horrible and like people are inevitably going to try to do both probably including including some of the horrible stuff first and like something's going to break at some point um do you hope you're going to merge for someday and become even smarter so like the whole like human AI you know like collaborative intersection thing like to me that's the best case scenario right like to me the kind of AI tools that I want to see is like not AI

that just goes off on its own right and just like becomes a more and more autonomous agent and does crazy stuff it's like uh AI that uh I where you know you have this like constants ongoing interaction where uh like basically the AI is like actually helps like every step of the process right and like it continues functioning as a tool and then with things like brain computer interfaces we can have like more and like something closer to realtime communication and then like eventually we end up merging with AIS at some points whatever that whatever that means right like that seems like the only path where that uh actually uh is uh realistic in the long

term right because in the long term eventually some something super intelligent is going to come out and like it is going to win over non-s super intelligent things and like either that thing contains humans or that thing doesn't contain humans and humans are permanently disempowered right and like this idea that you know 100 to 130 IQ humans can like permanently you know like remain in control of you know like these like 50,000 IQ super intelligent like things as like totally unrealistic right it's like uh you know imagine if uh you know you give a six-year-old you know like a legal control over like an adult company it's like well no you know adults are going to figure out how to make the

six-year-old order like what whatever they actually wanted in the first place right and so uh like I don't want like to me like a big part of the reason I'm in the space is like I don't want human disempowerment right like that's uh you know like centralization is human disempowerment right and if uh you have like super intelligent AI that are totally separate from humans then like that's uh kind of the natural outcome and so like yeah I mean some kind of you know like AI human you know like short-term collaboration and longer term merge like outcome you know like that is the thing that I want to see and like there's a lot of really amazing stuff that could come from that right like that can solve longevity it can solve like basically every field of science that can help you know Humanity explore

the stars um and uh you know we can get an uh amazing world and like you know people don't basically you know don't have to suffer anymore and uh I think that's like really uh beautiful and that's great but you know there's definitely like challenges and actually getting there I mean there is this you know the question of like you know can crypto help which is like you know I think the question that a lot of people want um answers to and like I think uh I mean real like you don't want to like do artificial pigeon Hing like it wants to like say oh you know hey I believe in decentralized Ai and therefore we're going to do AI on the blockchain it's like know like there is one form of

decentralized AI That's practical today and that form of AI is writing models locally on your laptop right and uh like basically if you're going to do some like fancy thing that involves tokens and then involves distributed computer networks like you have to be at least better than that right and so I think to me like the other short-term use case is uh potentially you could have Dows that can like help train open models like that feels you know valuable and uh like I think in general if you're the goal is to like train better open models that can run like locally on people's Hardware then like that feels very good from an openness perspective and like that sort of stuff like from what I can

tell it feels very unlikely that it'll actually contribute to like any of the destroying the world outcomes um so I think that's stuff is you know like really good to think about but then um you know going beyond that yeah I mean I think uh like we'll see right I think it's uh like yeah I think like my main message there is to like basically don't try to like pige and whole blockchains and coins into stuff like blockchains and coins continue to be valuable and continue to uh have their use but they're only one part of the world there's like entire other aspects of uh openness and uh decentralization that are super valuable right like I would love to see yeah you know like if we can it's like if instead of having like 20

layer twos and like five shared sequencers we could have like 15 layer tws four shared sequencers and like four Forks of uh like better versions of graphine os like that would be so much better right it's like graphine OS is like such an uh um it's a a version of Android that's basically been hardened to try to be like as security focused as possible but like at the same time still have very good ux so uh on my phone I'm like you know like I'm actually running it right now right and it's just like you know you open it up right and like it just uh like it just feels like regular Android right but it has like extra hardened uh security features um so actually there's a project called um

I think I I forget if it's ethos or ethereum Os it's like what or like eth phone it's like one of those right they're basically actually using a fork of graphos and then they're adding like a couple of extra ethereum focused features into it so so like basically yeah you know like doing and and then with graphin it only runs on you know Google pixel phones CU they're like for security they're trying to be like very specific on the hardware and they're also based on Google pixels and like that's the most secure uh like Android phone um anyway unless you're going to like go full on and like build your own fully open source thing um so if we do that then like you basically have like a you know like a non cringe and like actually interesting version of like the

crypto phone concept right like basically if you want to build a crypto phone it should not be about you know like shooring like nfts and crypto applications into things it should be about like taking the underlying values of crypto and like really yeah which are I think in a lot of ways the same as the underlying values of Open Source and just like really applying them and making them uh a reality for people so that's what I want to see in personal Computing that's what I want to see in AI That's you know like want to see in all all of these spaces final question to get out of this place and live on another planet someday so I eventually I would love to um I me people have asked the question

of like do you want to be the first to go to Mars or whatever um I think realistically yeah you know not the first um the uh I I think the the chall the question for me is like it it just feels like you know the costs of uh going to space are so high and you can get like 90% of the benefits like 100 times cheaper right like and he is at 3K right now so yeah you know you just like you just go to Antarctica right like you have uh like Antarctica international waters undersea international waters there's like so many venues that have like most of the benefits of space and like way fewer drawbacks and like I think uh like I

want to like I would rather me know like start there first and then for space I mean I uh I think it's absolutely valuable from just like a civilizational resilience perspective right like okay the way that I see this is like you always have to have like you know like a private good and a public good and like like like good things happen when you can like make the two match with each other right so like for crypto you know the private good is like at the very beginning it's like a number go up and then I think over time it needs to switch to use value and that's starting to happen and like the public good is like actually you like making these kinds of uh you like more um open and like Democratic Tech ol is actually built out and accessible and like for

space it's like you know the private good is like basically you know you get to be in these you know like International zones that you get to be like completely independent from people and then the public good is like civilizational resilience right like basically yeah you know Humanity continues to survive even if you like something happens to the Earth itself right and so and so like at some point our capabilities in terms of like not just going to space but like actually having an enjoyable life in space are going to increase right like that's I think the important thing right like I think a lot of like for it to become really acceptable to a critical mass of people like it can't just be this kind of like very you know like minimal you

know like survival um focused lifestyle like you know people want to be able to like go and eat sushi on Mars right and like you know at some point you know we're going to be able to have like 3D food printers and like once you have that then like you basically have like access to any cuisine and you have like like if you can have like access to any cuisine and you can have that from a you know like refrigerator sized box and then uh at this you know obviously you have the infrastructure to like do all the various you know like food Recycling and like all the chemistry to make that actually work um and then uh you make that work um and uh you have uh you know like some kind of life in space that's like actually comfortable I mean maybe

you like make the things spin at like whatever you know like the RPM is so that like when you do the you know like a v s over R math you get like you know like 1G in the cylinder uh then uh so that actually feels like normal gravity and like you do a bunch of other stuff then uh maybe um one of the things that zuu was kind of testing out for me right is basically like it was uh one of the things right it was like a kind of like a hacker house but bigger than a hacker house kind of like a conference duration than a conference and it was like more isolated than a lot of things right and so the question is like what's theale of people within with which a normal person would be comfortable like if they just like got got cut off from the rest of

civilization they just had to live with that group of people forever right and you know especially if you're are going to be living forever right you got to not get bored and I feel like that 200 size like actually is the right size right I mean tribes that are of dunbar's number size basically is what we evolved to live with right and like zzu with it size of like about 200 like it felt like that was the right size that you know with like you're not going to get bored and like you're not going to like start feeling um feeling lonely like it it was actually at that size where you start to like actually have like subgroups and you had started to have structure right so uh if we basically I think for a colony to be interesting to live like you know you want to have a size 200

colony and so you know you need to get like a size 200 group of people that's like actually willing to you know like move to the moon or uh Mars or wherever else um so like it's definitely a tougher challenge than just like getting one ship over right but you know we'll see when that happens cool thanks for hanging out well yeah yep thank you