ETHGlobal Paris 2023 - Juan Benet Fireside Chat

foreign [Music] next up what I want to do is take some time to interview Juan Benet Juan Bane is one of the founders and the founder of protocol Labs which you may know is the company and the organization behind falcoin and ipfs we're going to talk about a lot of things so without further Ado let's welcome Juan on stage [Applause] welcome thank you so much excited to be here you've you've been a regular at a lot of our events which is exciting because you have so much context about how everything is going so what I want to do for for uh today is talk about a handful of things I want to talk about all the things that you're seeing are interesting let's talk about how do we do things about startups let's cover all the things that you are but generally tell us what are the most interesting problems that you've seen solved in the last couple of years um I thought I was interviewing you um we'll get there um okay one of the interesting things at least from my perspective is that having been down this for the last six years um I get to see what the progress is on the developer side uh the most obvious thing I would say is um we have seen devtools just get infinitely better I remember four years ago you would spend half the time at a hackathon just getting set up and now you can click on a repo and it just magically works not just for basic daps I think the best part is um like I think we just kind of saw with a lot of our partners we're actually making hard things easy too like CK proofs are getting simpler I don't have to know effectively what is happening behind the scenes if I can just remember the guy in pickle costume I get to write stuff in typescript so I count that as a win um obviously a big thing is scalability uh from just seeing everything getting clogged from the launch of crypto cuties to where we are now with all these l2s everybody deploying on op stack like we've definitely gone to a few thousand transactions per second Visa is somewhere between eight thousand and fifty I believe and uh so that's going well and I would say largely the other thing will be overall it is a lot more normal to say that things are on chain these days uh whether it's the convenience or whether you want to call it ux or just how cheap it is I think on chain is becoming common what do you think great job uh what do I think so I think we solved a bunch of super interesting problems uh from my view the entire space has managed to build in the last few years an entire Financial infrastructure that is self-standing using crypto economic incentives to Now do much more financial processing than the traditional wire Network than the European wire Network and so on and we can now do microtransactions and local payments and micro loans and so on and so we have stood up basically a decentralized financial infrastructure for the world that nobody owns and it's a big big that's a win yeah that's massive progress well now I think that where we haven't done as well is in consumer applications right so when you think about the applications that you use on a day-to-day basis how many of the applications you're using today in your in your daily life are fully web3 powered like probably very few right so when you think about communication tools like chat or messaging or you know things like Discord um when you think about forums or even the tools that we use to produce our software like GitHub and so on it's like uh how this this entire conversation is going to end up in a live in a video platform that video platform is centralized and so on likely so from a consumer standpoint and in our day-to-day lives most of our products haven't yet um moved over to web3 and so that's where I think we need to go next which is like we spend the last few years building this really solid infrastructure and financial system now we have to bring Bridgit over to Consumer applications so we can get most of the applications on the internet actually using the web 3D Primitives and have like the right right structure okay that's I mean that's interesting so that in a way okay do you think that contradicts the common statement that blockages are not databases I hear you're saying everything should be more on chain is that the same interpretation or yeah I think so so I think questions are databases but they're databases that have this strong um set of decentralization guarantees about them nobody should control them you should have public verifiability to how you can get something included in a chain um who gets rights to change what part of the of the database and so they're not you know like traditional database that whoever the admins are get to rewrite all the stuff it's instead a public shared database now unfortunately the blockchains that so far have been built um on their extremely tight scalability constraints we're finally starting to open up the bottleneck but it's nowhere close to being able to deal with the scale of use than modern consumer applications require right so think of something like Twitter or YouTube and think of being able to process all of the operations and something like that think of being able to build social network entirely on top of these web3 systems like that's where we're headed but does that mean you're saying the scalability associates we have are just not sufficient or they are not solving the same or addressing the same problem yeah not sufficient so I think maybe a little bit of both uh there are some problems that the current scalability structures don't quite get you there like when I see many of you Builders are going to try today um and tomorrow uh your hacks will end up invariably needing a bunch of centralized infrastructure providers in order to pull off the kinds of programs that you want to build and at every moment there I think like oh damn like this is still a web too uh now you have to think about how do you move that into web3 and that's kind of the the missing set of pieces so some of it is think of like the back end services and back-end data processing pipelines and so on we need to build infrastructure for that I call call that kind of compute over data networks and the other part is just a raw scalability like the the entire industry right now is has been pushing really hard for scalability and we're now getting to hundreds of thousands or maybe millions of transactions per second I think we need to hit not billions but rillions of transactions per second sorry and like yeah yeah start thinking in levels of you know all tweets going through some kind of blockchain infrastructure like we need to hit that kind of scalability for a second collectively not in a single instance yeah so collectively as the entire industry but I do think you can get to this kind of level in kind of One Tree so if you think of ethereum with the set of l2s as he's starting to form a tree think of that programmatically being able to expand um to be able to deal with the demand of of block space um plus you need to be able to do clear transactions way faster than kind of um you know the current block time you need to be able to do operations that are subsequent so think of traditional data centers traditional operations you want to be able to do scalable High throughput database updates basically um within a data center at subsequent speeds once we can unlock that then you can get you know the rod base set of applications Rewritten in in that web3 landscape so like that's kind of the blocker between where we are today and the free internet that we're all working towards right so this entire movement started with a goal of building these New Primitives new foundational infrastructure to build a better rights oriented human controlled human data preserving um internet in order to get there we need to build that level of scalability so that any developer um can Target the cloud or can Target the web 3 Cloud Target the decentralized cloud so think about it that way I mean that's a it's a very uh I would say Contra not contradictory I would say the Bold strategy cotton let's see how it plays out yes uh contrarian is the worst looking for um we don't think about everything I mean even web 2 isn't thinking about having databases be optimized for this scale so uh that seems like a now that explained I feel like okay this makes sense if you're trying to replace something or offer a built alternative it's got to be least marginally better more than marginally better and uh in context this this makes a lot of sense how does Computing over data come into play here yeah so when you think about modern Services uh some amount of the operation happens in the client side so think of like in your phone or your browser some operations happen there some operations happen when you submit a request and you know there's some a little bit of business logic that happens a lot of that has been now built in kind of blockchain structures today but a huge fraction of the computation happens machine to machine uh afterwards it happens as part of background processing so think of like if you upload a video to YouTube um YouTube has to like rescale that video it has to post process it it has to apply filter ring it has to figure out the metadata it has to update a bunch of feeds it has to push it to all the subscribers there's an enormous amount of processing that happens behind the scenes and all of that kind of computing infrastructure today in our industry is still happening in a fully centralized web 2 context we need to get all of that kind of stuff to be able to be operating over a normal web 3 decentralized infrastructure and so we need to get like The computational Primitives to be able to do that kind of processing what's missing in that primitive is it just having the sure capacitor to set up these pipelines or are we waiting for something to be invented I think one part of it is being able to do kind of like job scheduling or kind of being able to run uh kind of what normally would be in like a web server um but as services that are blockchain powered and that nobody runs or controls so you can kind of run in the network and that includes a lot of computation so you can't put it in the modern kind of consensus throughput you should be able to kind of job schedule through through that so think of like kubernetes style operation but over um in this kind of decentralized setting so that's one component um I think a separate component is being able to do um just like these large-scale data pipelines so being able to define a job that that you want to run over some data you apply some function over that data you compute the output and you can feedback the output into the network so I think those two things are gonna end kind of a broader scalability Primitives are gonna get built out over the next couple of years I think in fact the kubernetes is a pretty good explanation because we know how to visualize that and you can set up all these nodes and clusters and run arbitrary code so that actually will be cool um I mean the more you kind of say it becomes I mean it's clicking to me and it's becoming more obvious that maybe this is one of the reasons why haven't seen a lot of good consumer apps it's because we get to not fully replace every part of that stack with something that we think is is better in terms of decentralization um but what's the prediction on how long it takes us to get to Children's transactions per second yeah so I think we'll probably have the prusa concept of this uh and the first versions of this kind of architecture live in the next couple of years and then from there we'll probably have the first applications that are going to push like all of their operations on top of these things um kind of in the year or two years after that so I think that within three to four years we might see the first fully native web3 powered um social network or something like that or or game like an MMO or something like that um that can just run completely on decentralized infrastructure so not next year not the after that but this infrastructure is going to come into play Then those use cases will appear and that will push the scalability uh up so I think hitting trillions probably five or six years out but I think the tech will be there in two to three years now that's uh it's a build move let's see if it's off yeah um okay so we talked about a couple of technical things that you think um could use improvements uh I guess one obvious thing that most people kind of don't realize that a lot of the things that blockchains also enable are just economic uh benefits and sandboxing so um anything kind of that you think is interesting is a theme that people are not experimenting enough with yeah so one of the most promising and valuable things of this entire space is the ability to create new um new funding structures or new governance structures around how humans and groups organize so this means new kinds of ways of organizing people organizing groups and so on and this is where we have the capability of creating new kinds of structures so like Dows are an example of this but we need many more such mechanisms to do things like way better Public Funding public goods funding or network funding or network governance structures that can coordinate us into building way better infrastructure and service systems for society as a whole so I realized a couple of years ago that if if crypto keeps growing at the trajectory it's been it's had it for the last 10 years and we can get some good Public Funding Primitives into the network crypto could be funding science at the scale of nation states so just for everybody's knowledge like some science running around the world is kind of spotty many some countries do a lot more funding than others the US is one of the countries that funds the most and usually this gets divided into different agencies around the world the kind of funding scale is kind of around you know in the tens of billions of dollars per year um for each of these countries so you can maybe add all of them up and land on something like 100 to 200 billion dollars a year for all of science funding worldwide so if you think of like a nation scale so like single billion or tens of billions per year crypto could do that crypto could actually fund science and r d at a global scale we're just missing the the right Primitives to be able to organize that kind of funding structure and the long-term feedback loop of showing that that impact is is returning value back to the entire crypto space so that can create a feedback loop so it's that kind of thing that I think is extremely exciting about crypto as a whole like these new kinds of foundational Primitives for organizing groups and all this kind of within the the broader community of public as funding so you might have heard from like groups like Bitcoin or network Goods or um optimism with retropcs and so on so all of these are kind of super interesting experiments for how do we use crypto native Primitives to build these new kinds of funding structures and I think that's one of the most exciting areas of the entire space no I mean that is actually measurable progress too because we can unlock so much and you can create these special interests groups that care about specific advancements um maybe for for the hackers too in the audience is is what you said a request for better tools or we need better economic models I think we translate that to actionable things so we have a lot of ideas so maybe two years ago we were missing ideas and so um now there's been a ton of Conversation Over the last over a bunch of conferences and prototypes and so on so we now have reached a point where we need to implement a lot of these ideas and test them so we need both implementation of these into tools that we can use to run these kinds of protocols and we need experiments that are actually going to try this and run this and collect data and see how well it works and what are the problems like what parts does it get right what parts does it not get right so we can refine and improve this kind of structure um the the really cool thing is like we're now reaching um in the tens to hundreds of millions of dollars deployed through this kind of public goods funding oriented structures so things like get going things like the network Goods group and um the retropcs in optimism together collectively deployed I think over 100 million dollars now and so all of that in the morning yeah and probably more um and so all of that is using uh public as funding oriented structures and so think of that as creating a new new class of of organizational structures that can help us build a better future without having to appeal to like the just the traditional you know private funding reps that one makes sense and I think that's like actually a good way to think about how large groups or even small groups can actually coordinate this um a couple things you said about kind of compute and scalability are these kind of things that you believe are specific to the Falcon ecosystem or kind of what's the context here no this is everywhere so I think this is for the crypto now crypto world as a whole uh if we can build these kinds of structures then the whole world would benefit and I don't know like um I tend to not see a lot of the the ecosystems it's kind of as fragmented as maybe most people might perceive there's a lot of branding from every different group and so on and so it might feel like driving lesson yeah tribal different communities and so on but in reality most of these groups are extremely collaborative they're in their shared resources and shared code Powers the whole um the whole ecosystem and so there's an enormous amount of collaboration so as an example like the the ipfs ecosystem which sort of like includes the Falcon ecosystem creates a ton of value for the ethereum ecosystem like lipido P which is something that came out of ipfs now Powers ethereum 2 and Powers polka dot and Powers Celestia and a bunch of other networks um and then even something like filecoin started as it was gonna it's sort of like you OG ethereum L2 like the um early on in 20 2016 what yeah yeah so uh we originally wanted to ship filecoin as an ethereum what we call it time a virtual Network on top of ethereum so like the concept of l2s wasn't around yet where is this this is 2016 or so yeah wow and so but we calculated the throughput and block space that Falcon was going to need and it was about like 90 to 95 of the entire block space of the ethereum based layer so we couldn't do that and so we had to like uh then create a create a separate L1 but now we're seeing all of that reshuffling and so my prediction is that we'll see lots of chains and lots of systems totally restructure and reshuffle themselves as the the developer experience changes as communities change and evolve and collaborate and so on like what we just saw this week Sello went from an L1 into an ethereum L2 and so we'll likely see a lot of that uh that kind of thing happening uh and I totally agree um I guess on that theme okay so let's say these predictions are are good on timelines in five years we see a trillion transactions per second which means anything you can imagine is possible on chain how should Founders be thinking about building companies and this is obviously the best time to start what's yeah so um this I like thinking about this in terms of maybe a set of stages that uh Builders should kind of think about uh when you're in a hackathon uh context and this is one of the reasons why I love hacking those as much and where he spends so much um time and thought uh working with Builders through hackathons uh have us give you this great environment where you can learn a ton you can push your skills you can explore a bunch of different areas but you have a limited time to be able to make your thing complete it and deliver it to um to be judged right and you have to also get good at presenting it you have to some some you have to solve an interesting problem for somebody you have to learn how to sell it to somebody and so all of that forces you to go through the entire kind of like product design validation Loop in one go like in a very fast pace and it teaches you a ton about lots of different things and you'll get a ton of feedback so that's kind of an example stage that Builders should go through so when you're first kind of thinking about potential ideas one part of that is get to being able to build like a minimum version of the product like a prototype or something like that hackathon is a great example for that it'll force you to start validating your ideas now at some point you'll find something that you think is compelling and interesting and you think that it potentially has legs and you're getting interesting feedback from others that maybe there's something here maybe this could turn into into a startup of some kind whether it's startup company or startup Network or some other kind of project at that point what you want to do is refine the idea and look deeper into it um try to validate it more think ahead to what it could turn into think about the who would the users be what are the early communities you could work with and so on so it's kind of a deeper validation step and that will start informing what needs to become your first minimum viable product right so you go from like early ideas early prototypes into something that you think will actually work that's like the key word there is viable meaning some product that you can actually create for users and for whom you're solving a problem and for whom are going to then start relying on your thing once you have that in place you have the trajectory for for a startup and again this could be startup company it could be a startup Network or a startup open source project or something like that the mode really depends on what the kind of thing that you're building is and at that point you kind of want to start thinking about what's the scale of this thing how um how much funding are you going to require to build this how many other people do you need to work with to to organize this and so for the startup company trajectories or there's kind of fundraising as an option uh or I guess for all these trajectories there's fundraising it just changes um structure like in open source you're probably raising grant funding in startups you're probably raising venture capital or something like that in in crypto Networks you're sort of somewhere in between um and that pushes you into kind of the next stage which is uh formulating this as a as a company or a network or something like that and then raising Capital going through an accelerator or something like that so right now where every all the abilities here are I would strongly encourage you to be experimenting through with this hackathon and future hackathons to find an idea and get to validating something that could turn into a minimum viable product you won't get to your full MVP through hackathon but you might get to something that'll let you take that next step and so think about that sort of broader trajectory as someone who's kind of been through the traditional funding route and also the an interface with the design Choice worlds any uh common bad advice that you think is given out for building startups that does not really apply to badass and it doesn't apply to web three companies yeah so the traditional startup ecosystem it has to be extremely competitive and extremely close and private because the nature of those kinds of products are that the all of AIP is very closed and highly proprietary or zero sum very serious I'm oriented and this whole ecosystem is extremely positive sum oriented highly collaborative most of what everybody's building is open source you can take in and Fork it and try something else and so you you end up with a much more collaborative environment where all of the groups building whether it's like l1s or l2s or services or products or whatever tend to collaborate dramatically more so one of the big things that I kind of see is you want to coordinate and cooperate as much as you can and you want to be very open with what you're doing you want to create a community early on and build a broader community of develop of users developers and and participants that are going to help your product or your system grow and that's very different than traditional sort of makes sense an hindsight um well I want to close off by uh well in this case asking the same question you asked me uh you've been here for in this space for almost a decade more maybe more uh where would you say we're in the space and I guess zooming out how do you think uh we rate ourselves on the progress we've made yeah so look I keep it maybe short and sweet here I think that this community has built some of the most compelling and important um infrastructure for Humanity broadly especially right now the finance stack is really solid but the broader communication stack and the broader internet stack is following and now that all these Primitives are here you as builders get to shape the next big thing the next important set of components so it's really up to you it's really up to what you want to build um and what you make and what you scale that will Define the future so go make it a good one and you know hack on really cool stuff amazing advice thank you so much thanks