ETHAmsterdam 2022 Kick-off Keynote - Vitalik Buterin

but before that i want to have a quick interview with a person who needs no introduction so uh without further ado let's welcome vitalik buterin the founder of bitcoin magazine on stage hello hello all right so we're thinking about the stuff we wanted to cover today and uh it's all over the place when i talk about the future of the past personal stuff but uh let's get right into it at this point what could stop the merch from happening oh did you hear the announcement we deleted until 2027. it's 2027. no no um i think uh realistically yeah what what could uh i mean what could slow down the merge definitely unexpected security books i mean that like if i you know if you wake up in uh you know some um crazy high months of this year and you discover that the verge hasn't happened yet since some you know security bugs and the clients is probably going to be the reason um things that cannot stop the merge um any kind of like weird minor malfeasance like if miners try to do some last-minute thing to break the merge that will just like mean there's going to be you know some kind of manual fork and it'll happen faster what are the possible scenarios of that and also could you talk about the uh yeah difficulty i mean what what what could miners do in principle miners could uh a 51 attack around the same time as the merge but like with the way the merge is designed it's like basically designed so that like the beacon chain fork choice just picks one of the forks and so we're fine miners could like if they were really coordinated like kind of stage of protest and just all quit um and if they do that then like we set a particular terminal total difficulty and so like they could just make it take a really long time until we get there but like if they do something like that and it's really serious then like once again like people will just make an update that uh just like either changes the tpd or just like sets a particular hash right so there's always a recovery there um i mean what else can they do they can like make fight about centralization on twitter i don't know um like i i think like on honest at this point i mean like practically speaking i do feel like miners have generally you know realized that ethereum is going to proof of stake and like you know a lot of them are not just miners they're also stakers um and you know they're plenty of them are you're very happy to be you know participating in um you know both both parts of ethereum's transition uh so i'm actually not too worried um and you know like lots of them i'm sure like that they're going to want to start mining on ethereum classic or z cash or like any of these other systems um and so you know like they as an industry do have a kind of common interest even going beyond ethereum there so yeah i guess not too i guess not too worried about that i mean i like i don't know what else like i feel like aside from you know like the possibility of software uh bugs which are which is exactly what we're in the process of uh testing to find out it's pretty much a full speed ahead so i think this this obviously kind of touches upon just the complexity that lies in in the base layer and you kind of talked and blogged a lot about just how the layer one should be as simple as possible um what's kind of the philosophy behind that how does that kind of now reflect on the design for the merge and let's go deeper into that yeah i mean i think in ethereum we've been kind of moving toward this like hybrid philosophy where layer one tries to be kind of very simple um and or at least as simple as it realistically can be um while layer two kind of takes on the complexity burden of uh you know providing like very high levels of scaling providing privacy providing a lot of the things that users want right the the changes that um are all kind of going in the direction of i mean earlier one does improve somewhat like it's which is the proof of stake i mean we added tank sharding which accounts a lot of data uh that the chain gets consensus on the roll-ups can use and all of these things but the goal is still to try to like add a limited number of tools that's a very general purpose i mean then the you know layer two and the application layer and it does the rest and uh what's the design process that like do you actively refuse to think about adding changes to the base layer is that more of like you only get to realize this in retrospect how do you kind of think about that as you're planning this yeah i mean it's not just like it's not my decision right like so like sorry not you specifically just as uh you know and i know on one of the you know recent awkward awkward just decided to kind of not really yeah push hard on kind of non-merge related efps for a while um and like there's and like in general i think there's definitely been kind of more and more pushback on adding eips like even things like pre-compiles that would have easily gotten into or three years ago like they just have a harder time justifying themselves now which uh i think is healthy like i think it's uh you know good to kind of not keep on i mean increasing complexity forever and kind of throwing the kitchen sink in and like really step back and think like what do we really need um so you know there's definitely an attitude of like growing conservatism toward changing things um like if i had like if i had to kind of change that formula a little bit like i would say yeah like what i would want to see is uh not like not just conservatism in terms of like the size of the changes but also like conservatism in terms of the final results like not not just thinking about like you know changing little but also like can we make changes toward an ethereum of tomorrow that's simpler than the ethereum of today um so like banning self-destruct is probably one of those examples right i think like i do not want a world where um you know your grandchildren uh you know wake up at some uh you know a dark rainy morning in 2072 um you know take a look at the ai robots outside briefly um and then you know ask you but grandpa like why yeah why does ethereum like just have all of this uh extra two-layer weird state tree structure and why does it have half ropi and half ssd and like why is that why isn't it creating my own node so hard and like you have to answer well you know because we uh some people in 2022 kind of just basically give up and prioritize backwards compatibility right like i think um you know there is something to like the virtue of uh kind of accepting short some short-term costs um in order to achieve long-term benefits and like obviously it's important to prioritize and obviously you can't do everything at the same time but like having this kind of discussion around you know what are things that you know do incur short-term costs but are going to like actually make things better like ethereum a better system and a yeah you know um a simpler system with uh in the long long term is like a discussion that would definitely love to see more of so maybe this is a good time to understand how that got applied to sharding and the design in the architecture could you tell us about the original design what it is now how that happened well in that case it's a it's a bit of a different situation right because like sharding is something new and so like the question of like minimizing complexity of the change and minimizing complexity of the final outcome like they just end up being perfectly aligned with each other um and the design really did evolve a lot in the direction of simplifying uh the bass layer right so you know at the beginning um it's uh like you know there was this really complicated like quadratic and then like eventually cubic sharding design and you had committees on top of committees and you had chains on top of chains um and uh you know it got really crazy um and then eventually that simplified down to a beacon chain uh with a bunch of shards that had crossed that were chains that had crosslinks eventually that simplified further we got rid of the crosslinks we got rid of the shards with an independent fork choice and now we don't even have shard blocks right like the dank sharding design um is actually actually very ultra minimal because it basically even has one proposer that just chooses all the contents for the for the entire block like you know the entire um 16 or 32 mega megabytes or whatever and it ends up being right um and uh that is um much like i mean it's definitely is still complex right um but at the same time it's like vastly better than what we would have had to deal with had we actually tried to like properly implement any of the designs from 2017 while at the same time you know also having these five other road maps of like account abstraction and proof of stake and vertical trees and all um and uh uh all of those things right so i think uh i do think that the ethereum's kind of a research community learned um and like has updated a lot in the direction of like proposing things that are actually implementable um so i think the interesting thing i've noticed is that this is a very recurring theme and it's like a great thing which is we kind of have initial ideas to get kind of iterated on and they just become simpler or even better or just more efficient is that something that you think is because we're able to minimize the feedback loops or is that because we just have a lot more research that's kind of being applied faster than uh in other industries like how does this kind of why is this repetitive in a good way as a theme from your perspective why is it not repetitive why do we see this happen often that the designs get simplified especially in in how we think about scalability yeah i mean i think a big part of it is like the first of all just the quality of the research team improving um sec like i think uh you know the more the kind of research thinks hard the more like it actually can like build things that don't require developers to work as hard um and uh like ultimately development is like a bigger slog than research is right so i think it's a you know it's good when you can make those trade-offs and just think a little harder and make something better um the other thing is also just the proposals keep improving because like you have the benefit of understanding you know versions one to version end of the thing when you build version and plus one of the thing um i think it's a combination of both of those two it's it's just impressive that we see that consistently happen on the research side like and then three obviously it's just that like this ongoing process of like actually just like learning to value like pretty like real simplest anymore well speaking of simplification i want to talk about two eips and uh i think it'll be great to kind of have you talk about what they are first and then what's going to go into uh should or should they not be included um let's start with 4844. what is it and should that exist sure so eap 4844 um aka protozoading um i guess uh okay to explain that maybe we need to start like going into dink sharding in a bit more detail right so dink sharding is the sharding design that is uh kind of the result of like this long chain of like simplifying things and throwing things out after we started from crazy chains inside of chains back inside of chains back in 2015. um and the way ding sharding works is you still have like one block per slot it looks like basically just like a regular non-charted chain except validator nodes don't actually have to validate everything right instead validator nodes like randomly select a few pieces of every block to verify and because of like erasure coding data availability sampling a bunch of like math magic they're like that's actually enough right like it's actually enough for every validator to only verify a few randomly selected uh pieces of this like big 16 megabytes like blobs or i guess blob lists um so that yeah now the problem with the egg sharding though is that it does require like eurasia coating and all this stuff to actually be implemented um and so what happens before then right and so before then uh we have uh proto-dink sharding which is basically a like it's a kind of halfway house that can happen uh before full dank sharding uh uh can really be rolled out and what happens there is is it's kind of like you do dink sharing without actually shorting the data right basically you introduce all of the data formats exactly as they will exist in full dink charting uh like you introduced this concept of blobs as trans transactions with this extra really big piece of data that's much cheaper but that can't be accessed directly from the evm you introduce kcg commitments you introduce all the math you introduce that like all the the ways that interaction between the execution and consensus aerial work you basically do all of the kind of infrastructure work except you don't do this kind of like a random sampler check you still require everyone to just like fully download all the blobs and check them right so in every swat everyone would actually have to download and check maybe about a megabyte of data so this amount of data is still a lot um it still provides a huge amount of space for roll-ups it still stands you know big good chance of making roll-ups way cheaper than they are today um but so does this have any long-term effects on the state size um right so on the state size no because these blobs are not part of the state uh but uh on the history aside um eip4444 is another important one this is the history expiry eip basically making it so that clients don't have to like store a really old history anymore and this is important both to keep the disk size of an ethereum node small and also actually to allow ethereum implementations to become simpler right because like you don't need one implementation to be able to process all past versions of the ethereum protocol that ever existed um you needed to be able to just like process like maybe the last one or two hard forks and then if you really wants to have a node that checks the entire history like you can basically just uh you know download multiple versions of the client and then like you know we'll we'll have to like early verify yeah serial evr file like we'll have to implement like some interchange format so you know you can run one of them then export the post state import that as a pre-state to the second one run a few more blocks export the post data and put it into the third and all that uh but uh like that's all doable and that like that does uh like basically the idea is to try to like allow the nodes responsible for a consensus or allow the nodes for like responsible for the really critical thing of keeping the blockchain running today to kind of be as simple as possible and like there are going to be ways of doing all the other stuff like there are already plenty of ideas on how to like create automated networks that kind of automatically start backing up all of the history and make that really robust so it's like basically guaranteed to know it was like a one of interest model that the i'm history will be uh will be uh actually saved um and uh also you know like actually be able to like check and see what happens in um in the history um but like that would be a kind of somewhat more specialized functionality and like users who care about that will be able to do it but users who only wants to run a node to verify what's going on today would not have to concern themselves with as much with that and they'll be able to like just process a narrower piece of the chin uh so yeah that's the ap four four four four uh what about four three three seven four three three seven is um the account abstraction i believe um it's going to have a presentation tomorrow so i highly recommend people watch it uh but uh account abstraction this is another one of those things that i've cared about for a very long time like basically it's like making smart contracts while it's first-class citizens right so today's there's two kinds of accounts in ethereum there's what we call eoas externally owned accounts which are the accounts that like most people have their you know coins and assets and everything in um and then there's smart contract wallets smart contract wallets use like smart contracts code to be able to impose arbitrary conditions on when the funds can be spent right so you can have a multi-sig wallet where you have like say five different keys that that might even be controlled by different people and you need three of them to approve a transaction you could have social recovery wallet where like a three of five can change the key that's needed to sign you could replace ecdsa with a different signature scheme you could replace it with something quantum resistance right like you could do a lot of things but the problem is that smart contract wallets today they base like smart contract wallets cannot initiate transactions and so actually using one requires you to have some money in an eoa um and like you basically just needlessly complicated and that's a big part of why not many people use it but then eoas they only support like one form of authentication which is one single private key right so i think um so a 4337 does it basically it's like an extra protocol um way kind of like second layer way to make uh smart contract wallets basically be as convenient as eoa wallets are today and so you know regular users for like it will be much easier to like actually have one of these more secure like you know multisig or social recovery or whatever kinds of wallets and so to try like you would actually be able to um you know feel safe and comfortable using ethereum without having to worry about like what happens if someone hacks my computer and steals my private key or or what if my private key gets lost is this just an implementation is it just in the implementation phase or what kind of how do we think about that from just being i believe like it's already on test that's about we if your c4337 is going to be released on the main net fairly soon um that should change how most of the hackathon projects work that's going to be awesome um you already touched upon the quality of researches that are going to come on um in the last few years to kind of improve the protocol but let's kind of zoom out a little bit and go into how do you think the ethereum foundation's role has changed in that same window and just kind of from your perspective what was the role now and how do you think the role foundation serves i think the ethereum foundation first of all kid has really stepped up on kind of improving its quality improving its organizational capacity um like there's just a lot of things that the ethereum foundation was completely not capable of doing just because like it did not have people and there'd be lots of people who would just say like hey like you know can we do this and like i just unfortunately would not be able to point them to people who can actually do it um and like and now all of that stuff is just so much better right like you know there's like excellent people in the in the foundation like steven gilbert referral like who's been you know doing a lot of like um latin america oh work some other things um you know josh stark has been you know great at just organizing a whole bunch of things in the uh in the you know foundation core um and you know there's this uh you know i am a gosh new executive director has definitely been you know great um and uh you know both you know her work kind of organizing the foundation and a lot of the philosophy that she's introducing i think have been excellent so just a a lot of like really good people lately that i'm happy very happy about um so it's definitely a much more capable organization it's definitely an organization you know more capable of starting to like have have breathing room to think about the long-term big questions it's like in 2016 and 17 like we barely had breathing room right like in 2015 like we barely had money i mean that because of barely having money we barely had people and then in 2016 and 17 and beer we had enough people to do things and like finally you know over the last four years we've really managed to exit that kind of you know vicious cycle which has been uh i think excellent we can actually observe it too which is the great part uh just the last two years have been phenomenal and just seeing more transparency the level of effort going into different geographies this event itself def connect i absolutely agree yay def connect do you think the ethereum foundation should just focus on making the protocol better or do you think this expands into supporting or collaborating with other protocol um i mean i think there's definitely room for the ethereum foundation to get more involved at the application layer um and i also think that there's more room for everyone except the foundation to get more involved with the protocol layer right i'm like uh i think to some extent this has happened already right but uh like i think uh for the first couple of years like we have had this kind of like v like interesting sort of like half decentralization where i know there's these like five different lawyer two client teams um or well the responsibility way more um like five different uh twos um if consensus client teams and there's a lot of lawyer two teams there's a a lot of uh execution lawyer teams but like a big part of a lot of their budgets would still still come from the foundation right and so like it sets up the structure but the foundation is still kind of at the uh has this kind of very strong very central role um but like we were already starting to see signs of that diversifying we're already seeing signs of like actors outside of the foundation supporting all of these different teams supporting all these different individual developers and you're referring to just kind of companies in this space that are like companies is one like you know obviously you know there's a consensus um there's like the various different dows there's like well i think like lido announced that they're supporting protocol guild um you know the whole political guild effort in general um i think you know the whole the cell like the nft for eap 1559 um and like there's i think a lot of potential for this to go much further like i think we still need to remember like that we are in a world where i believe this is still true that the the unit swap dow has um more uh capital than the foundation does uh i believe so too yes um i guess what about collaboration with other layer ones do you see a future where we can work with the bitcoin community oh interesting um yeah yeah anything i mean anything can happen i think um you know out of um you know coming out of the the coven pandemic i mean a new century uh you know new challenges um you know it's uh possibly yeah time to get a time to make new friends um i'm it's uh yeah well i i think uh you know especially yeah like both the bitcoin and the ethereum ecosystems are just kind of so different from what they were five years ago um and um you know they're just like clearly here you know do things that are somewhat similar and also somewhat different um and i think like there's a a queer sign that like you know actually coexistence between these two ecosystems is possible um and uh like i think uh the the big thing that you need for like people to feel comfortable cooperating is like the feeling that there doesn't exist they kind of neither can live while the other survives a relationship between the two right because once like like people are at their most evil not out of greed but out of fear right like when people perceive that you know if they don't like fight for their interest and for their own interest that that they're gonna lose to some existential threat that's when people are gonna be the most willing to sacrifice everything for their interests um and you know if the like the competitive part of the relationship is not an existential one that already just like reduces the temperature so much i mean i think i mean technical collaboration is definitely possible i mean i yeah i mean there's definitely some happening already like that okay a lot you know the you know the nerds generally like they're they're happy to kind of like you know ignore the politics and just like work on really fun and cool stuff and interpolate the polynomials together uh which you know i think is great um and um i think i'm you know more and more very possible over time that's great all right so for the next section um i'm going to call that underrated overrated i'm going to give you a topic you uh comment on is if this is underrated or overrated and then a short answer on why sure so first one 85 hot green tea 15 red wine underrated like only a few people have heard of it so far we need to like spread this we need to like make it the new thing the current thing [Laughter] we'll try to serve that at midnight okay one coin one vote uh i mean one coin is a scam um oh oh i see like you know like token voting governance um i think realistically realistically i think still overrated like i've yelled a lot at it but i feel like a lot like the yelling hasn't really yet translated into people actually seriously experimenting with alternatives what's a good framework for thinking about how voting or governance should work overall it's a very broad deep question but specifically for the voters it is i mean i don't think there's one framework i think there's a few different frameworks like one of them is collusion resistance right like how to prevent like basically how to prevent concentrated interest from always beating out diffuse interests right like because like that's uh the big the big problem that always happens in the world um and um you know plutocracy is actually not very good at that like what other things can be um the second is um how to like even like and then or the first one you can think about it in two ways like you can think about it kind of like mild ways where you know you have like the concentrated interest as like a fairly team actor like i don't know paradigm paradigm is tame right um and uh the diffuse interest is like you know just like you know random community members but then on the other hand there's like the concentrated interest is like a hostile attacker that's uh like um you know there was that 80 million dollar defy hack right like someone tried to roll their own stable coin technically not a hack right technically it was a yellow was it it was working as intense it was a special cryptographic operation um yeah um investing in pride yeah so uh basically yeah well what basically what happened there was like there was this uh state of someone tried to roll their own stable coin um and it had a governance token um and the governance token i mean it had like and something like an 80 million dollar market cap and this is like lower than the value of the assets it was protected that was the treasury i believe right well it was yeah the governor's tokens were lower than the value of the treasury right um and so someone did a flash loan and they just like borrowed 51 of the governance tokens right and they voted to transfer the money to them as a governance vote and then they flash loaned it themselves and they got approved yeah right so um you know you want to be resisted to attacks right uh so that's also is delay the best way of thinking about that or well i mean figure out civil resistance figure out civil resistance like i think the delay thinking kind of like it's it could help in the short term it's a band-aid but if you're like in my article i think it was like moving beyond coin voting governance or something like that like i talk about some like longer term financial attacks that can still work no matter what the delays are and so you know i think it's uh still concerning um so there's uh that um so like what that so that's one framework right of like thinking about collusion and attack resistance another framework is how to ensure that the decisions are high quality right like basically how do you ensure that the people with the governance power actually yeah are people with expertise who actually deserve to have the governance power um and like delegation is one of the answers but is the delegation might not be the only answer because um you know there's it's still subjective yeah i mean there's like people who don't know um and who know that they don't know but then there's people who don't know that they don't know right um and um you know if they're gonna um uh just like delegate to like peep the frog number one two four three seven then like that's not also not the best uh so i i think i think uh we should go on to the rest of the topics because there's a lot more there um pfp nfts pfp nft is i think a bit overrated just because like uh you know the problem is like you can all like you can clone them and like we don't really have good infrastructure for like telling clones apart from not clones and even like answering the philosophical question of like when is a clone even an illegitimate clone in the first place and like when is it just like you know yo dog that's part of the game public goods funding still one of them you know one of the most important problems in the world so um you know probably still underrated actually i have way more questions so i won't dig into it um rlp um i mean it's still part of the ethereum protocol so it's overrated [Laughter] hmm play-doh iron play-doh art like the philosophy oh play to earn oh i see it um we can do that i mean art of like the of the greek philosopher i definitely would want to see some more art yes um would you would you have played art is your pfp maybe okay yeah play to earn i think a bit uh probably a bit overrated like i think the problem that i see with it is that like i haven't yet seen a credible story about like who would actually be the one that's like providing the capital that people end up earning by playing is an argument for just not having enough time passed or just mean it's an economically it's just an economic thing like if you have like if you have token outflow you have to have token inflow who's going to pay for the inflow um l2 centralization like centralization of layer 2 protocols winner take all effects in l2s oh um well i guess that's like you have to be clear about what you mean right because like one of them is um like oh you know what if like fuel gets 95 of the market share um but fuel like becomes super decentralized and then basically have like this you know weird kind of awkward two-layer distinction but then the other one is like well what if there's like five different roll-ups but but the sequencers are all run by one guy um i mean the good news about lawyer two is like those are like even the worst outcomes are not as like actually necessarily that bad uh right because you still have trustworthiness um as long as the alters are designed correctly you still have censorship resistance probably the biggest risk is like governance capture like what if the l2 has governance and it does a governance vote and there's too much network effect for people to like actually be willing to exit if the governance turns it into something trusted for example are you worried about that influencing layer one design in a way that you would not intend or just like l2 is influencing layer one design um i mean possible like i mean i definitely do think that like application layers tr like things it's not necessarily always about governance like it's a it's not just an o2 problem like it's ever it's a it's something that could happen in like so many different ways right like there's going to be a lot of conflicts of uh interest between different kinds of applications um on ethereum um right i mean you could go like all the way dystopian and like oh you know is there going to be like some major uh you know a centralized stable coin that starts a lobbying for the base lawyer to have kyc uh which uh no like i feel like that one is like a bit too blatant and i think too many people would just like resist that one very quickly but i mean others all kinds of things like that so i think the answer isn't so much worrying about like one specific interest as it is about making sure that the governance as a whole is like secure against uh like those kinds of like social attacks from all of possible uh concentrated interests last one private keys underrated or overrated oh underrated i mean you know the eu is like freaking trying to ban people from having their own personal wallets like come on it's uh yeah i think uh like personal wallets where like you actually control your private key or it's a social recovery or a multi-stick wallet and your family control some of the keys like that kind of like cyber self sovereignty is i think like a really important bedrock of having a decentralized crypto ecosystem uh even with account abstraction uh even with account abstraction yes like i think like smart contract wallets are not about giving up control they're about like being able to actually choose who to who to trust who he can recover and like like these things don't like even like multisig wallets and social recovery does not have to mean centralization like um you know as i said like you can choose five different friends who don't know each other i mean have them be your recovery guardians like i actually recommend this right um so there's a lot that you can do with that but like like anything like that is far better than an ecosystem where like almost all users like just interact with the blockchain through a centralized provider where they don't actually control their own access yeah i think permissionless needs to be truly permissionless how how were the last two years for you with covet and everything oh it's been i mean it's definitely like the the kind of mandatory break of my usual schedule was definitely kind of like interesting and fascinating um in that um like before covet i was basically on on a plane on average once every six days like i would like i literally have like tracking and like that was my life between about 2014 and 2019 and that obviously had to cut down completely i was basically you know holed up in singapore for about uh some you know most of a year in the middle of that um and i ended up walking to um every district in singapore you know it's a nice small city state 700 and i think 18 or something square kilometers um that's uh i think like was it somewhere around somewhere around like 300 freedom units a bit less um but yeah you know it did yeah in singapore's nice like you could actually do like literal cross-country running there but the um yeah so that was also just a you know a good opportunity to kind of like think about things um you know just like get quiet work done uh but you know at the same time also definitely a lot of like a lot of inconvenient like inconveniences and like a lot of serious stuff that i would have rather not had like there were definitely family members that i was just not able to see for a whole year which i think was you know really unfortunate and uh definitely yeah i think to pick up any new hobbies did i pick up any new hobbies um i mean i listened to a lot of podcasts um you know that was uh any favorites that you want to recommend i mean hardcore history it was uh a fun one alex friedman's are a lot of them are very good um there's also like alex friedman's about history and there's alex friedman's alex friedman with the guy who does hardcore history quite a bit um but um you know since uh then it's like i think definitely is such so nice to actually be able to like also see people in person and like you know see yeah all of the yeah old friends again that i am if i missed for uh such a long time i think that's one of the the really interesting pieces for for us today too there's hundreds of people that have participated at these global events for the last two years which we've never met and we're seeing so many of them come to us and say hi and us knowing that including some of the global employees and team members we've never seen them you know until now and it's awesome to see that so i'm really glad to hear that today is the 22nd of april how was your 420 oh let's see what what yeah i mean 420 is an important day because like 420 is like let me explain it it's the smallest number which is uh divisible by all the integers from one to seven um and so it's uh you know it's a it's oh it's also um the current the co2 concentration in the atmosphere um so you know very meaningful number um but uh anyway um what did i do i uh i i went to an in a mini internal ethereum foundation research workshop and listens to mary meller do an excellent job explaining secrets a single secret leader election sounds awesome we've talked enough about the past eats to the merge let's talk about e3 what's coming what are some interesting problems and uh what's something exciting that we can cherry remember there is no e2 and there's definitely no e3 there is uh the merge the surge the verge the purge and the splurge right so you know the merge we know is completing the switch to proof of stake the surge massive increase in scaling because of layer two's and sharding uh so you know we talked about dank shorting i know a lot of progress on lawyer two is here um and uh that's something that we've really seen that have become alive over the last year or so right like we've basically gone from you know stark ned and optimism and arbitrarium and most of these kind of being theoretical to like really being live i mean some of the some of them have been like around for much longer than a year right but it does really feel like the last year has been this uh amazing and huge uh tipping point um so the surge like really getting kind of the scalability vision kind of fully complete um it is progressing right like i think very recently optimism and uh added in support for a compression and that noticeably decreased transaction fees by i think like another like something like was it like somewhere between 25 and 40 percent yeah yeah yeah so that's amazing and like there are going to be more and more of these right like um eventually you know i could replace ecdsa with something else and that can like add a lot more compression um eventually yeah you know prototyping shorting like a lot more and then full dang shouting a lot more so i think you know we're going to see transaction fees on rollups continue taking steps further and further down over the next couple of years and that'll be amazing the verge this is the switch to of verkaltries uh but virko trees are important because they enable stateless clients which basically means you can verify the chain basically without having anything in your hard drive like you could just verify any random individual block you want without having like pre-existing information about what happened in the previous block and that's really important right um and uh what basically because it makes it much easier to run a validator it's like very good for like proof of stake decentralization um it also means that we don't have to worry about the size of the state for a long time like you know there's always these concerns about oh are we going to implement state grants and like do we have to do that what happens when the state blows up to a terabyte but like if you have proposed builder separation so we can like allow block building to become more specialized market while validators don't even have to run any state at all they just like verify stainless blocks then even a tier of the state being a terabyte in size is like actually manageable right um so virg is going to do a lot a lot of good things um then there is the purge and the purge is like basically like eliminating craft um so we talked about eip4444 no it's not having to store history anymore there's like the very long term vision of eventually doing state expiry so eventually we also have we can also like allow even builders to not store old state um then there is uh just allowing ether and making it a norm for ethereum clients to not have to process and like be able to understand every single previous version of the protocol um is also part of the purge simplifying the ethereum protocol so banning self-destruct maybe like getting rid of and replacing the 2300 gas stipend like maybe getting rid of and replacing logs uh because uh like do we really need in protocol bloom filters when like basically no one uses that infrastructure and everyone is using the graph anyway right like archive notes will exactly like at some point you know i think you do have to sort of ask that question um what other things are there there's um i mean just like a lot of simplifications in the ethereum protocol that i think can be made and that i think should be made and like the purchase all of that and then the splurge is like basically all of the other cool stuff right so account abstraction is really big importance proposal builder separation is really important um if there's this guy like um adding vdfs to increase the security of uh validator set is also really important so a lot of really nice things um and also there is um i guess you could put this in the merge bucket because it's also proof of stake uh but this uh longer term idea of moving to single slot finality um so instead of blocks finalizing after two and a half epochs blocks finalize like immediately in the next slot and this has like a bunch of important benefits there's a bunch of technical reasons why actually it might be easier than we thought um and like you know over the last like year or so you know we've discovered that there are a lot of like just inherent complexities that are difficult to work around um how do you think about reorgs in that world i'm just curious yeah so the point is that reorgs would do not exist unless you're in the special case where like more like a large portion of nodes are offline and so you're not finalizing if most of the nodes are um online then you're finalizing so you can't have reorgs that's like the big benefit um right so like this the other thing that we realized over the last year is that there actually are a lot of like unavoidable complexities and like room for bugs and the way that we combine fortress rules and finality today i mean so moving to single slot just like reduces the number of protocols from two to one and it just makes things so much simpler potentially that'd be huge that's awesome um yeah sorry are we no no i'm done there okay final question um been part of ethereum for the last seven years uh the bitcoin and just the cryptocurrency ecosystem is around for a little over a decade how kind of how does that evolve from your perspective to the space whether we're talking about the research or the community or just overall the blockchain space give us like what the last five to ten years have been and uh kind of how do you see this moving forward and then for the next five to ten years um i think uh the the crypto space today is just like so different from what it was uh like five ten years ago or even five years ago like there's even there's aspects of the old space that i really miss like i definitely miss like how it was this kind of like very small and niche thing that like most people didn't really know about and even people who did know about it kind of thought that it was a sideshow this is an anecdote uh our last event was east london two years ago two years and like a month ago and uh we were struggling to make that a full house and that capacity was 300 people so exactly yeah and then you know yeah there were definitely these kind of like very mysterious and undiscovered aspects of it like you know even like visiting china and like discovering this kind of like dark forest of like various 16 year olds who have actually been mining bitcoin for four years i've ended up already being multi-millionaires and quitting high school um like uh there's just like a lot of libertarians and socialists uh peacefully arguing with each other on bitcoin talk and like having like actually a fascinating high brow conversations um oh um sometime around this time i think is like the 10th anniversary of like either the pirate at 40 hack or bitcoinika or maybe both like i forget um quicker remembers pirate at 40.

wow uh who here remembers bitcoin okay um you know uh uh zotong um who was uh give us a little bit of background uh i think it was like the first like bitcoin leverage trading platform it ended up just getting hacked over and over again and eventually it just didn't have any money and lots of people got angry um and uh pirated 40 was an anonymous guy running uh basically like a funds which was like a way of uh investing in like some bitcoin related stuff and getting like stuff like i think it was like seven percent weekly profits um and uh you know yeah the guy basically said like hey you know trust me and uh a bunch of people did trust him um and uh you know what mark of what mark zuckerberg said 20 years ago about people who trust him uh so uh this guy was anonymous so it was that he basically just like ended up ended up just like oh no he actually did get caught and he did go to jail eventually uh but uh did they just like such crazy stuff um happened back then i mean i guess you know these days used like we have like different versions of that like we don't have that like that sort of stuff severity has been slightly minimized well i mean it's like the sizes like the dollar amounts of crazy stuff that happens are i think about the same but like as they have been all along but like as a percentage of the ecosystem it's definitely much smaller and like i definitely think that's good right um so it's definitely good that the ecosystem has matured um and i know it's definitely grown a lot people understand a lot like what actually can be done what can't be done um you know how uh like what kinds of things make sense in terms of scalability in terms of privacy in terms of all of these issues uh so i think um you know over the next five to ten years like uh what we're gonna see you know the continuation of that process and we're going to see like you know like a lot of crypto applications continue to turn into something that's like really mainstream um and you know i think there are going to be some costs to that like i think uh like the aspect of you know the the crypto space being this frontier i don't think that's going to go away like i think uh the crypto is just like so hard to control that there's always going to be a frontier within crypto but like it's not going to be the case that like all of crypto is a frontier anymore right and but at the same time you know it's hopefully going to be something that like just the lots of people use as a regular part of their lives and it's a good feature before we end any predictions for what the next few years look like oh um the ethereum price will be 12 000 but i will tell you in what currency could be bolivars you know venezuelan could be uh or it could be dosh um you won't tell us what currency or what base exactly yes um and i will tell you if that's a thousand in base 10 or base three um um what are the predictions i mean basically i think um you know scaling and zero knowledge and smart contract wallets and all that stuff are going to be mature and uh you know the ecosystem will then have to like basically either succeed on its ability to actually deliver on the applications that people have been expecting for so long or it'll not do that or something in the middle um i mean i do think that like blockchains have already already proven that they are here to stand but they do have some value but there is still this question of like well you know in what places exactly how much i'm like you know are we talking like the new uh windows of finance or are we talking the linux of finance and uh and not just for finance like well how how does that um end up like playing out in every other other industry um so no i mean it's still um you know a lot of things that remain open and undecided but you know i think uh you know all of the lovely yeah you know hackers of the hackathon here what can you know do a lot to try to kind of make help make the more positive outcomes happen that's a perfect note to end this on thank you so much and uh just give a big round of applause