Dreaming big with Protocol Labs

Transcript

Super excited to have you. Let me let me let me do your formal bio for us and then we'll get to the get to the real stuff. So so Juan is the inventor of the interplanetary file system, more commonly known as IPFS, a protocol to make the web faster, safer, and more open. And Filecoin, a cryptocurrency and incentive storage network. The IPFS project has grown into a large open source movement to re-decentralize the web, safeguard our data, and improve our applications. Juan is the founder of Protocol Labs, the internet tech R&D lab that develops IPFS, Filecoin, and actually many other things as we'll talk about. And he studied computer science at Stanford. It's actually our last founder talk of of of CSX. I can't believe

we're here and it's really it's actually pretty meaningful to me to have the opportunity to talk to Juan as part of it because he more than anyone is like the reason that I'm actually here in that he was the first crypto founder that I had the opportunity to work with. I think I said to many of you when we when you were doing your CSX pitches that that I'd sold a company to Google. I spent about 5 years there and my after-school activity was working with early-stage crypto founders and as part of that really just fell in love with the tech and the people in the space. And and the person I fell in love with was Juan. So There's thank you.

So no, amazing amazing to have him here talking about this. And the thing that we're going to get into is that you know when when you think about the most important, most powerful, most significant companies that have been started and grown really I think in the history of the economy but certainly in the history of tech. They typically would combine really two core properties. One was, uh, just a huge vision about, uh, what was possible, uh, that and and and a lot of that vision was things that nobody else thought was doable with, uh, with just relentless, determined execution. And like you all can think of those companies, Amazon,

uh SpaceX Tesla Nvidia uh Apple going way back, many others. And I think that in crypto, uh, One really embodies that, uh, that core set of capabilities. So we're going to talk, uh, a lot about that today. Um, so may- maybe just kicking it right off here. So in your in your bio on X, uh, you describe yourself as obsessed with knowledge, science, and technology. And, uh, just curious like did that is that something you've always had? How did you cultivate that? Why do you say that? Yeah, so um, and by the way, first off, uh, super honored to be here. Thank you so much and thank you for, uh, the super super kind words. And more importantly, thank you for joining us in the crypto world, uh, helping us make the world a dramatic- a better place, so. My honor.

Uh, so the being obsessed with, uh, science, uh, know- science, technology, and knowledge, and and so on, um, came fairly early. Like I I think I started getting into, uh, computing pretty young, like, I don't know, um, 12, 13-ish. Um, probably the most formative one of the most formative pieces of media for me was, uh, Carl Sagan's Cosmos, and that shaped a lot of how I think about, uh, the world. Um, I watched it very early, and I watched it a few times, um, uh, when I was kind of like growing up. And that gave me like this super deep appreciation for how humanity has come so far in in a fairly short span of time. Um, and for me, that set off a a

kind of long-term quest of discovery and and uh, falling in love with, um, science itself and like knowing things and um, discovering new phenomena about the universe and or now multiverse. Um and the kind of ever increasing frontier of the massive tree of knowledge that we're all building. The kind of technology side came later when I was thinking uh realizing how it's really through technology and embodying those concepts that we might discover in some um concrete product that you can diffuse out into the world that you really get to upgrade the species. Um and so that's where I I and nowadays tend to look at

science and technology as two parts of a whole as opposed to two completely separate uh pieces. And the knowledge piece uh to me that this is a whole like uh you know, in in a parallel uh timeline uh I spent an enor- I would spend an enormous amount of time um thinking about memetics and the process by which we get to figure out the strands of information that we this describe as um pieces of knowledge that then go into our like knowledge computers and then express out um you know, behaviors and work and whatnot. Um so you can you know, really think of humanity's success as having unlocked um properties about how we dis- distribute knowledge and how we refine knowledge

and how we embody it in technology. Uh and so like that's a super super um fascinating piece. So anyway, long long winded answer and it's like very highly compressed um statement but like yeah, I'm just kind of super obsessed with the whole the whole thing. It's it's kind of like I think humanity is really about you know, being able to do that process really well. Discover things about the universe, embody them into the technology and then kind of bootstrap our own capabilities. So that's like that's the biggest thing we're sharing from, you know, the rest of the species on the planet. And I mean actually and and you and you definitely I think have woven that into uh everything that you do and I like one one example I'm thinking of is that when when you you you you guys refer to Protocol Labs as as PL uh and sort of

every every time you talk about uh a new evolution or iteration of the company, you change the version number, right? So, it's always you're you're you're always thinking about organizationally product-wise people-wise, how do you upgrade, ship the next version, and keep it going? Yeah, that's like an surprisingly good design decision and kind of hack to level up an organization. So, the idea here is like most human organizations over time develop like cruft and become too stuck in their own ways, and it becomes difficult to change them and whatnot. But, if you just do the very simple, straightforward rule of like, "Hey, let's version the organization the way that you would version some software." And you like force yourself

to like ship new versions at a particular point in time, then you enable everybody in the organization to fully buy into the idea of reshaping yourself on some cadence, and it becomes easier to schedule significant changes, it becomes easier to like question fundamental things about the the organization itself, and rebuild it separately. So, kind of in PL's history, we've probably had, you know, we're kind of like in PLV10 now, like that 10 years of of PL at this point. Um and in there, we've had like three major shifts during that history with a lot of smaller, minor, significant, but, you know, smaller smaller shifts. I remember vividly remember the first time we met in 2018. It was at my office at Google.

It was late in the day, and I I knew you only by name, barely understood like what you were building, and we ended up spending like probably close to 3 and 1/2 hours, and you you took me through your whole thesis of what you were trying to build with PL, you know, talked about the the connectivity back to what Bell Labs had been back in the '60s. And I remember I remember walking out of the meeting thinking that one of two things had to be true. Either or like you were batshit crazy or you were this visionary you know genius.

And happy to say that the latter turned out to be true. But is that I mean was that I think that's very accurate and like that's surprisingly or I guess unsurprisingly there was a similar thing that I heard when I got into YC early cuz PL went through YC in 2014 cuz there was no other there was no CS X at the time to go into and at the time you know Sam Altman was running YC which is the big reason I went because he had a much stronger technology oriented deeper tech oriented than the kind of like the rest of YC at the time.

And I remember like in my interview he was just basically like man that sounds crazy. Either this will be like this will totally work and be huge or like it'll fail catastrophically. It's awesome. Let's do it. And like I I think a lot of a lot of high risk reward plans feel that way. Like if you're onto something very significant it'll it'll seem and sound crazy at the time but yeah you're probably seeing something about the future. Well so yeah so let's let's get into the crazy talk talk about what is what it is that you have now been working on for 10 years in terms of building with PL and and where you see it going.

Yeah so and and you know PL has had you know where where PL came from was that when I was kind of like going through computer science at Stanford and so on I was A I became pretty obsessed with the internet an amazing tool and technology for all of humanity. I kind of grew up more on the internet than any place. Like when people ask me where did you grow up? I usually default to say on the internet. And the how the internet is made up is just a fascinating like, you know, patchwork of um RFCs and protocols that people like talked about and then implemented and

then bootstrapped an entire like global nervous system. And that's a super super super cool. So, I kind of became fascinated with that um you know, whole technology stack. And then separately, I also became fascinated with like um how you can develop a technologies and again upgrade the species in a significant way. Um today you can take technology and just in pure pure software use the internet as a transport mechanism to just give superpowers to everybody, right? Like today you can like dream up some crazy new capability for all of humanity.

Literally package it as a bunch of software and ship it out to the whole world. And that is crazy. Like if you try to go back, you know, 150 years and describe that to people, like and just be like, "Yeah, we have like this magic machine that we get to like um grant new superpowers to people globally." Um you know, super super amazing. And so at the time like um one other thing I kind of got I sort of like realized um was that progress had sort of stalled out and slowed down relative to um a lot of other um So, I think software and computing had uh kept going, but like a lot of other parts of technology had slowed down massively. I was really into bio at the time, I was really into other areas and was kind of

uh bothered by why. And that kind of set me on a path of like trying to build an institution that was able to do the kinds of projects like IPFS um that uh you know, kind of at the time I looked around when I first were thinking about building IPFS and I couldn't find a single organization that could really give it a good home. Um you know, academia that wouldn't work out. Um doing it as a startup wouldn't work for IPFS on its own because it would had a needed a strict separation between uh a business model um and then places like Google and so on at the time um the project would sort of like become co-opted by other interests that were sort of like more important. And so, I wanted to kind of build I was very

inspired by Bell Labs, and I was sort of trying to build an organization or institution that could um steward the building of those kinds of technologies. And that kind of sent me on a path of like figuring out why those kinds of organizations don't exist anymore, what are the structures that need to exist for that those kinds of projects to be able to be built, um and how do you form teams, projects, capital to be able to um uh run them in the long term. And it took me the better part of like the last 10 years to figure out precisely like why it is that that it is so um difficult to push technology these days in in of certain sorts. And and the kind of like the gist of it is like the there's a thing called the valley of death or the innovation chasm um in between deep

conceptual research and the early periods of product building. Um and that area, which usually today is like crossed by startups, by you know, individual founders and um individual startups take this like super hard leap of like taking some early basic research concept and get it to a point where you can actually produce a product um and then be able to like get um uh investment capital. That part um is extremely long in certain areas of technology. So, in software it's a lot tighter uh because it's not that expensive, but if you're trying to do something uh much harder like in in hardware or, you know, massive scale infrastructure, that valley of death becomes extremely long. Um and there's no good ways of funding that part. So,

the kind of like long-term plan of PL is to build capital structures to be able to fund that area and kind of make the valley of death dramatically shallower. Um so, yeah, I don't know. A long-winded answer to to a And and then and and just to expand on that a little bit, right? I mean, you you you see this uh you you see this playing out into uh many different uh kind of bleeding-edge, almost science fiction aspects of uh what we would call of the science fiction technology. I mean so certainly AGI, brain machine interfaces, the intersection of bio and computation, and and many others.

Yeah, and and the way that we kind of, you know, if I had is that we think about driving breakthroughs in computing to push humanity forward and they're really trying to be at the edge of like what is now suddenly becoming actually possible. There are things that are kind of too far out that are just, you know, 15, 30 years away and there are things that are now, just now have become very tractable. So yeah, brain machine brain machine interfaces are things that the Overton window has been opened. We have some really good um existence proof that you can you can now really do it and there's now on the order of like three to five startups that are seriously pushing that field forward.

That wasn't the case 10 years ago. 10 years ago that AGI was like the contrarian thing. And yeah, excited for what it will be like 10 years from now. And so then then bring that back to Filecoin in terms of how how does that serve the overall strategy and and where does it fit and where do you plan to go? so in 2014 I I kind of really So kind of the the the the long-term goal of PL felt very remote and harder and like had a a much longer term quest. Um I had the idea for IPFS separately. Um I got into Bitcoin in 20 13. Like I had seen it before in college but it kind of dismissed it because it didn't seem to um

It's funny like Ali and I used to like mine it in in the in the myth cluster in in in the CS department. But that, you know, like we probably both like didn't care very much about it because But you said you said you have like lost hard drives with like Yeah. tens if not hundreds of Bitcoin on them. Yeah, so I don't remember how much how much Bitcoin it was but yeah, I think 2009 or or so like we we used to mine a bunch and um e- there were had been a lot of other cryptocurrencies before, so Bitcoin was not the first significant cryptocurrency. There were a bunch of them. And it didn't It wasn't clear why this one would make it in ways that other others didn't. Uh turned out to be the mining process, which is a super super key um feature of the whole thing.

Um But anyway, um in 2013 was when it was starting to hit like larger scale adoption. I mean, larger scale here at the time was maybe tens of thousands to hundreds of thousands of people. At that point it became clear, "Oh wow, you could actually do um a proper cryptocurrency now. That's pretty interesting." When I was designing IPFS, uh there was a clear need for a medium of exchange in a currency. Um and so Bitcoin seemed like a very attractive thing. Um but I wanted to use the Bitcoin process to incentivize the distribution of the of the data the storage the long-term storage and distribution of the data.

And that's where Filecoin came from. Um and then as soon as I kind of had the skeleton of what Filecoin would become and the skeleton of what IPFS needed to be, that became viable to now actually start a company and start a startup because you could bu- build a business around uh Filecoin and you can um develop IPFS as a a normal open-source uh public good for everybody. Um And yeah, so so Filecoin was kind of like the first major um business and product for uh for PL. Most of the last 10 years have been on building IPFS first then uh uh a side quest into building libp2p, which now powers a a lot of the crypto um space as you know, the networking stack. And uh and then Filecoin. Uh and so a Filecoin has been like a building

the Filecoin network itself um has been the better part of a of the last 10 years for most of the people uh through the PL network. But now kind of in the last 3 years um PL has grown into being a a larger network of lots of different startups across the space, probably around 60% in the crypto space and maybe 40% outside uh kind of another another verticals. And in the crypto space we kind of divide it into two categories. One is more web three focused where it's building infrastructure for all of the tech and the other is around building infrastructure for like better economies or better governance systems.

So this is like public goods infrastructure projects or or governance systems. And actually one of the things you've said for a long time and I think continues to be true today is that that you know one of the most one of the most powerful things in blockchain networks is is essentially the incentive mechanism, the block reward mechanism. Even even though it's one of the best understood, it's also probably something that's way more powerful than people even understand in terms of aligning incentives. Talk talk about that a little bit.

Yeah, it's extremely under appreciated and underrated. Um the Bitcoin hash rate graph is the most important In my view, it's the most important graph in the entire crypto space. Um it shows you the magnitude of the power of a single incentive structure. Um the block reward model where you have What what is effectively happening and it's not um immediately obvious because the way it's programmed and the way it was shaped like they were trying to solve for a different problem but they happened to create this like very neat structure um which creates like an ongoing real-time public auction of a certain amount of currency at a a decreasing amount amount of currency

over time for whatever value the network is able to put up at that moment in time. And that free permissionless auction to the whole world um paying a certain unit of currency for a particular resource is an extremely powerful way of organizing massive amounts of of work uh globally. And so today the Bitcoin network um has so many people um building uh infrastructure and mining operations around it that you have like energy use at the scale of nation states, you have like countries that are like now taking serious policy decisions about mining Bitcoin or holding Bitcoin.

And all of it came out of like that one major incentive structure. It also created a a way of like um, distributing the currency to a lot of people and getting a lot of people to care about it. So, we use that same kind of same kind of structure in Filecoin to organize massive amounts of storage. Um, that's why the network was able to grow, uh, into, you know, exabytes like super super fast like in in a in a very short span of time relative to, you know, how long it took, um, the centralized companies, uh, to amass that kind of, um, hardware. Like, you know, it it was it was a huge build out for the, um, kind of Web 2.0 era to build out to get to like hundreds of petabytes and then eventually exabytes.

Um, and sure, we got to start like much later, so the, you know, we got the scaling laws, but still organizing exabytes of capacity through just a protocol is amazing. You get to like deploy this incentive structure into the world, and then like summon all of this, um, uh, work out of it. And you can use that same structure in lots of domains. Um, you can organize tons of work in a bunch of different different ways by just using that that same, um, same exact model. And and so, maybe to bring together, uh, some of your long-term original thinking about PL and the and the power of the block reward, maybe, uh, talk a little bit about your recent trip to Argentina and how some of these ideas are, you

know, coming into play and could be used to essentially like reshape the economy and the uh, organizational structure of the nation. a few months ago, um, a friend and I were on a ski lift and we're like, "Oh man um uh, Milei got got elected, uh, it is like totally reshaping the economy and is like kind of vaguely pro-crypto. Like, it's not clear that like he's super pro-crypto, but like has said it some like strongly pro-crypto things. Like, this could be like a major opportunity for a whole country to to adopt um crypto cuz when you think about like the the underlying infrastructure that we've built as a as a whole community, we've built these amazing kind of like legal jurisdiction structures where you have a

you know, you have a a way of doing contracts, you have a way of like transacting between a bunch of different parties, you have ways of making all of that verifiable, you can create currencies, you can create like all kinds of assets. It's the machinery that you need in order to like run any kind of economy, including national economies. And so, you know, I've been wondering for a while like which country is going to be the first like um adopt it fully and just leapfrog the rest of the world by building a a crypto-native infrastructure. And so, it uh we were on the ski lift and we're like, "Oh man, like uh maybe uh maybe now's the chance in with Argentina." And it turns out that Argentina has something like 10 to 15% of crypto adoption because of the last few in the

last 20 years have seen like this horrible devaluation of the currency because of a ton of printing um and a and a bunch of other policies. And uh that has led to a ton of people in Argentina needing to use crypto in order to be able to hold value or be able to move uh transact globally and so on. And that has led to this massive amount of grassroots bottom-up adoption in the entire country that I haven't really haven't seen almost anywhere else. Maybe maybe China, um but but almost nowhere else have I seen like that level of of of broad adoption.

And um and that coupled to now kind of like a tops-down administration that is pro-crypto creates like this crazy opportunity to to potentially build out uh the infrastructure. Now, the the problem is in order to um uh like really build an economy for a nation that way, you need to like fig- build out a ton of protocols and products to support users, right? So, you need like point of sale uh systems, you need like um uh you you might be able to do like house mortgages this way, you might be able to do the house land registry this way, you might be able to do um a corporate registry this way, you might be able to like do all of the kind of like debt instruments um on chain. All of that kind of stra- stuff.

Each one of those is a whole startup and product in and of itself. Um so, then it's like, "Okay, well, we got to like stimulate the creation of a bunch of these startups." A few friends uh and I decided to like host a pop-up city there in Argentina in May and we just went down for a trip uh to to just talk to a few founders. And we were blown away by like the depth of like interest in this that in like two or three days we like enabled we we like sort of like catalyzed this like movement that it now is growing on its own. Um so, it's called like Crecimiento uh movement and that, you know, there's on the order of like 500 Argentinian builders from, you know, startup founders to investors to um uh

people running companies to government regulators um and so on, all in this Telegram channel talking about all of the problem space uh that Argentina is facing, super energized about um using crypto rails to upgrade the economy itself. And that, you know, that that that shows you like the the level of power that crypto has, the opportunity space to solve like deep hard problems for um for entire nations. And um and just the latent interest out there to to to actually do it. Meaning, um almost everybody that we talked to had a super deep understanding of crypto already, um understood the potential, understood the the the problems with the traditional structure, and just kind of like needed the the the

like the opportunity of like hey actually let's let's realize that everyone knows that everyone knows that this is a better structure. Let's just go do it. And and that yeah that's been like super super cool. And you had a pretty cool heuristic coming out of the trip which was I think I think you said you know in your in your view Argentina was like 10 successful startups away from from really catalyzing huge change in the country. And so if you if you apply like your normal you know nine out of every 10 startups fail but one succeeds like you just you you need roughly 100 new companies and then you can you you know with with with good founders with good ideas with good execution and you can actually like

change the inflection point of the whole nation which is like pretty powerful idea when you think about it. Yeah yeah we're we're super excited about it. It's like uh yeah there's like this massive opportunity. So if anybody here is working on anything related to like DeFi structures or like you know stablecoins or land registries or ways of organizing work like you know think of like the regular everyday sort of use cases. If you're thinking of those kind of structures and currently you're facing like massive amount of red tape in other jurisdictions where it's very difficult to figure it out like come down to Argentina and like build build there because there's just a huge amount of interest in and doing doing this kind of thing. Um and come come out to the So

these pop-up cities let me just kind of explain what that is. So uh this started because of the um Zuzalu pop-up city in Montenegro and before that kind of like the network states um uh whole movement um it's become viable now to create like to bring a lot of people from around the world for like a month to two months of like very focused knowledge dissemination and learning and then project formation and project creation. And it's sort of like a wave like creating like mini Silicon Valley type culture environments very quickly anywhere in the world. Um and that's been dramatically successful in a bunch of places. And um and so that we're hosting one of those things in um in in May. Sorry, in August.

May May's now. Uh Uh and so like I said I would say that's that's uh a lot to have done, thought about, worked on over the last decade. As as you think ahead to the next decade, like what what are some of the things that you're either most excited about or you think um society is broadly underestimating that will become possible or or incredible? It's a hard question. So, um I think so maybe give the I'll give it two ranges of answers. One on crypto and one kind of outside. Um in crypto side, crypto gives us the ability to create um human systems that preserve human rights. Uh that that kind of embed digital human rights in their operation. Like that's a major uh

possibility space. So, I think like doing that well is a is a huge priority and I think could be could be massively transformational. Um I think economies that figure out how to do the entire like techno capital structure of like building startups, having investment funds, being able to do those operations, being able to charge for users. And if if we can create like a juris like a full kind of like common law structured um jurisdiction in a box. And like whatever country adopts that first is going to be able to like leapfrog uh a ton of other people. And I'm sorry, a ton of other countries very quickly. And that I think is going to be transformational to how uh so many other things happen. Like when you think about any other infrastructure or service or whatever, it needs that initial kind of operating system of of a you know, corporate

structure, a way to employ people, a way to pay people, a way to like uh do commerce. Like all of that structure. If you can if we can just kind of dramatically improve that, that that'll be super transformational globally. Um And then the kind of like outside of crypto, I think like, you know, of course the the AI AGI landscape is um I think the probably the most important thing that humanity is doing right now. Like if if if we manage to make AGI, it's like that that that's kind of a totally transcendental pivot point for for what it for our species. Um another one of the same category is whole brain emulation, which I used to think was maybe 40 years out, um but I increasingly think it's probably closer to 15.

Um and so I think it when when that happens, that's a whole other transcendental type of shift. Um and so I think that that's the other one to like really be um orienting towards. Um maybe failing whole brain emulation, it might just be very high bandwidth brain computer interfaces, um which I think will be in in market in 5 to 10 years. Like so so that's like super fast, right? So um just to put put it in context, 10 years ago I used to argue with people about like, "Oh no, like major AI systems are coming and AGI is like going to be here somewhere between 20 to 40 years out." That was that turned out to be like an overestimate.

Um and so I think these kinds of things are going to start happening fairly fast. Um and I think it's a super super exciting. Cool. So let's just close by talking about uh something you do that it that that actually think is um in in some ways a source of some of your uh deep thinking, which is which is Prometheus seminars. Talk about, you know, what those are, how you started them, and and why everybody should emulate the model. So this kind of started start partly in in in hosting like kind of like a group house.

Um in in your day-to-day, whenever you're kind of working on whatever field you are, um usually all of your time and focus has to go into like the main thing that you're working on, and you leave little time to deeply think about or learn things that are kind of far afield. And so it can be very like once you leave the more structured academic learning environments, a lot of people can find it very easy to just fall into just learning around the space of things that you're currently working on.

And so creating an environment that enables people to like talk about the weird, edge, controversial, contrarian type things that they're thinking about or learning about or curious about. And enable especially things that are like um currently globally misunderstood. And and explicitly creating an environment where people can like talk about them, get get into them in detail, and so on. Um can be fascinatingly valuable for a group of people to learn an enormous amount, change their thinking, and whatnot.

And uh for a bunch of us, like that you getting together and having these Prometheus centers around, what kind of like Promethean shifts will we see coming? Meaning, you know, when you think of Prometheus, like the myth of Prometheus being um you know, the beginning of fire or like um stealing fire from the Greek gods to give to humanity and like totally change the the trajectory of humanity. What are those kinds of like Promethean fire type innovations that are going to happen to the world um that and and when are they going to happen? How are they going to happen? Um what will sort of like how will shifts occur? Um and that be be creating a space and an environment where a lot of people can talk and discuss these kinds of things is super valuable. So, I think dinners are a good thing. The other thing is group houses.

Um and I guess now pop-up cities. I think the the cultural environments that you can create for people to be able to like especially in multi-disciplinary disciplinary setting where you can grab people across a bunch of different fields and you get to spend a lot of time together. Um It is a super super super helpful in um catalyzing an enormous amount of learning, um catalyzing the formation of projects, um relationship forming, and whatnot. Um so Well, and just like briefly describe the tactics, right? So, you would you would typically like send out a last minute invite to a broad broad group of friends like, "Hey, everybody come to this house at like midnight or or like way way past my bedtime."

No, it it it used to be like usually I used to plan these with more advanced notice as as my life has become a lot more like travel heavy and like unpredictable. Now, it unfortunately happens like very very quickly. But yeah, it'll be like a kind of like get an invite to go to a dinner. Usually the invite itself needs to set the context. So, like the imagery that you use in the invite or how you describe it needs to like shape the expectations of the people that are going to come to this event. Once you uh come to dinner, usually you need to like do context setting and kind of shaping of the conversation which can be, you know, as like straightforward as um talking through like the the range of ideas or range of things things that you that that you want to get through. A lot

of it is about like the people like so selecting the set of people that can generate that kind of conversation is is one really key part of it. But the other which I I think it's actually potentially more important is really kind of um curating the setting like enabling like shepherding the conversation to to like go into those those kinds of environments and doing it in a way that is not overly formal. Like if you if you make it too formal then like it feels it feels not right. Yeah. a fire, right? Where where people can tell their stories. Yeah, yeah. So, Prometheus dinner with like a actual fireplace is like a really key key setting or like some like large fire object.

Really really key part of it. By the way, like humans have spent on the order of few hundred thousand to a few million years like around fireplace fire camps. So, like it's a very natural thing that we um have gotten very used to. So, there's something that of like psychologically interesting that happens with people around a fireplace. Um it's just very uh very deep. All right. Well, let's uh we'll leave it at that and uh and open it up to you guys. I'm curious. So, like in the in the last talk with um Ben Harry, they they talked about a fully decentralized way of running their organization. I wonder how you you view running uh PL um and

how that's been you know, maybe evolving over the last couple years. I I guess it, you know, what what you have what you've arrived at now is vastly different from the initial state. Yeah, so I'll maybe talk of about a few things. So, maybe the first iteration of PL in you know, PLV1 in um 2014 or so, I was going down the path of creating a traditional-style company in a single place. Um and you know, got a had a house in Palo Alto, was like starting to hire people, and was like, "Okay, great. Like, come to this house and like work from here." And the response from uh the early folks was like "Uh do we have to? How about No, how about like I just work remotely?" And we like accident- accidentally fell into just

working remotely from the get-go um because a lot of the beginning of IPFS was an open-source community on IRC. And uh that meant like we built the beginning of PL and the beginning of the IPFS community and the beginning of the Falcon community um as a fully remote distributed team uh everywhere in the world, not by initial intentional choice, but rather by by it being a a way to lean into the open-source nature of the thing. And that enabled us to really build open-source software for everybody in a very collaborative environment because if you have a small setting in person, it's much faster and easier and and more efficient to like just talk something through in person in a whiteboard or

whatever, and then trap all of that knowledge in your heads or in your physical infrastructure and not, you know, distributed to the community. So, for us being distributed was an advantage. Of course, it had costs. Like when you when you're a distributed team, you don't you aren't able to like form culture as easily. It is harder to work through all kinds of like important challenges for startups. And so that, you know, that was kind of like one of the pieces that was that was valuable. Maybe to your question around decentralization, which is less about like maybe physical decentralization and more about the organizational structure itself. Um so, PL m- mail there's two parts that one is the Falcon network and the other is the PL network. So, the Falcon network um v- from the get-go was like a pretty decentralized network right away because

the like even before the the we built the software and we launched the network, lots of communities started springing up around the world that were really excited about the Falcon project and and the potential of the network. And so, right away we started like, you know, there were dozens of different organizations around the world very interested in developing and growing the Falcon protocol and the Falcon project. And so that naturally forced us to um pick ways of organizing the community in a in a very decentralized setting. Um including a lot of the software build out. A lot of the software build out happened through lots of different uh projects and companies that worked together on on on building it. Uh and then on the PL network side, that was a

transition of going from like a single centralized company into turning it into a in into a decentralized network. Um and that that that's more around like how do you organize like companies and funds and capital flows and whatnot and how do you create like a like a distributed culture. Um so, I don't know if that's like a good enough answer, but Well, and you you guys have have been like way ahead of the curve on all that. I mean, to the point where when when COVID hit in 2020, like you didn't skip a beat because like all the things that every other company was scrambling to get in place like you you'd been running that way for years but I Yeah, so so funny story. So we've been distributed for a very long time and uh we were like on the run up to like

launching Falcon and uh we're like okay, well like if we're going to um we started like trying to make certain things go faster which are very very difficult to do in a in a fully remote distributed team. So for the first time we're like okay, we're going to have an office for a while in Toronto and we like open this office which is like a unusual for us and immediately like COVID hit and uh and uh we it hit China first and so we saw it through all of our China communities um and they were telling it like we were on Slack with them discussing like how bad things were in their communities and so we were super aware of how bad COVID was going to be because of that and so while we're building the Falcon testnet um we had like the the Grafana dashboard of

the Falcon testnet and the uh Grafana the the like world dashboard of like COVID infections in China and we're starting to see the the graph going exponential and we're like oh that's going to be a huge problem. Um so right in January as we heard of it hopping, we like canceled the plans for the office and then we sent everybody home and everyone was like this is way too early, what's going on and they're like no no no like this is going to be like a massive disaster. We were like going around with masks in January like you know, people would have thought we were crazy. But you know, sometimes so by the way, one thing that I've learned over the last 10 years is you will have intuitions that are run against the grain of what everybody else is doing and you have to like

figure out whether or not you're right, like really convince yourself, are you actually right or is the rest of the world right? And if you are indeed right and you can't convince yourself that you're wrong chances are like you are indeed right and just do do the thing that you think is correct um and then later the world will adapt um and that's you know, a uh, have a lot more faith in your ability to figure out the world. Uh, unfortunately, human organization, like large, as human organizations get large, they get a lot slower in adapting to like changing circumstances. So, something might, like the world might be doing something that no longer makes sense. Um, and you'll be able to like front-run the rest of the world by by just thinking more clearly about things.

Um, you talked about kind of like the big chasm to cross and also like the juxtaposition of that with software, like the ability to like really easily package something out and ship it to the world. Um, and then also talked about like these kind of big pivotal pivotal moments for our species coming up, like AGI or BCIs, etc. And so, um, I guess how does the fact that those are closer than you initially thought like change that? Like if you've just kind of like realized the whole chasm thing, then it's like, how does that mindset change now that like that's going to shake everything up?

Like I just mentioned about the world not adapting, I think currently the like we're in this very weird period where the global conversation, like the mainstream perspective, has truly no idea how different the future is. And we are due for like some like major rearrangements, and it's unclear how they're going to happen. Um, hopefully they happen smoothly and we are able to like surf our way through these like crazy waves. Um, but it there could be like some, you know, Look at what happened with COVID. Like a you know, one virus took out like how much of our infrastructure and

operations and reshaped how the world works. Like um world's stability is not as There isn't as much buffer there as there really should be. And so, we we could see some massive shifts happening as these these technologies start occurring. Now, I'll say one piece which is like the if if you look back through history, especially the last two 200 to 400 years, like um the people like that discover the science and build the technology usually understand a lot more about it than the people that like are just finding out about it.

And so you and they usually have positions of leverage in shaping how the thing gets used. So, you can do a lot to ins- ensure safe transition on some things and enable like really good safe outcomes for the world by ensuring that the teams that are building those things take that into account. So, today we have a super free and open internet and that came from the actual culture of the people building the internet. It happened to be built by academic labs who had the principles of an open like free oriented and and it happened in the US, right? So, you had a lot of the US liberty perspective and the academic scientific oriented culture and that led to a super open internet um

in a way that like if it had been built elsewhere, it might not have been um that way, right? And so, if you think about like the web 2.0 uh data monopolies, like those were built by through like businesses and and certain certain business and corporations that are oriented in a particular way with a particular culture around openness or or or a closeness and so on and that ended up creating a totally different type of platform. So, my best advice on that is like really think about the future, think about like the the things that are going to change over time.

Like you probably want to spend a little bit of your time thinking very far away and most of your time thinking very close. Um and so try to like just orient whatever you're doing such that like as things change, you can like shift what you're doing to match reality. Um but then really focus on like building something truly very deeply valuable, like, you know, 10x improvements to whatever else there is. And try to do so in a way that like maximizes good outcomes for as many people as you can. Um, and just, you know, remember that like um, humanity's just figuring it all out. And just because things used to work a certain way does not at all mean that the future works that way. Like, just pick up any history book and like look through many transition periods and you

can see how quickly things can change. And so, just like orient for that, orient for like things that really make sense to work on and build um, through those through those periods. So, for example, I would not be working on things that are about to like get totally replaced by a foundation model, right? Like so many companies that people are starting right now are like dead on arrival or dead pre-arrival, right? You've you've spoken a lot about how things will be different in the future and how we're hitting a point now where everything's going to change radically. Um, and it's super super exciting.

Um, but Jeff Bezos has a very famous quote where he says, you know, it's easier, everybody talks about how things are going to change and how different it's going to be, but for his business it makes a lot of sense to think about what's going to stay exactly the same. So, for Amazon, you know, people will always want cheaper goods and they'll always want them faster, so he can build Amazon Prime, basically. But, I'd be very curious to think to hear what you think about what is not going to change and what can we rely on in the in the world of startups? You have to think about this not as binaries, but as probability distributions, right? So, there's a large probability distribution of possible outcomes and you think you have to think about like what is most likely to happen and then what are the

low-probability, but very significant shift events that you still have to pay attention to, right? Like, you you don't have to care about low-probability, low-impact events, but you have to think about both high-probability events in general, both high-impact and low-impact, and low-probability, super high-impact events. And so, I could probably, you know, list out tons of things that I think are most likely not going to change. Um but, there are like low-probability type of events that will shift. So, um the coming of AGI and things like whole brain emulation and so on, those things are like um low-probability events that are super transformational.

Um kind of like setting those aside because like it gets extremely difficult to like reason about the future. Um with those, setting those aside, um you can reason about like um people building people like it is very unlikely that um the world will invent new ways of organizing people and groups and so on. Like corporations are pretty strong, stable structures. Um like today, most R&D money comes through corporations, not governments, and that is that is accelerating, and so that is likely not going to change. It's going to be, you know, barring a a war effort, like that's likely going to be the case.

So, then figuring out more and better ways of like leveraging um businesses and products and so on to like drive capital formation is like a good thing. I don't know. I'm I'm speaking very abstractly here in like very large-scale type of change. I think, you know, the the bulk of the answer here truly is most of everything will remain the same year over year. It's really in the 10-year or 20-year timescales that that these major shifts are felt. And so, I would um Yeah, even in in Bezos's quote, like you know, he didn't say people are going to want physical books, right? Like that really changed. And so, but he was totally spot on on like, yeah, people are going to want cheaper things faster.

Um and and, you know, think about like what are those kinds of properties for for humans as they get access to all of these like new technologies. And and and try to find things that are like good in both cases, right? Like whether like the world remains the same, you should have a good product. The world dramatically shifts, you should still have a good product. And ideally, those things are like very close to the same, so you can like maneuver. Um working in crypto, like just getting to like normal everyday sort of stuff, um I think you can you can um bet on like the the underlying power of smart contracts to organize like tons of people is both currently very undervalued and just going to keep opening up a lot of opportunities. So, I

would just like think of better and better ways of like leveraging that for for a super powerful type of outcome. And I wouldn't worry too much about like how do you have to dress it up today? Like let that be more of like an implementation detail. Kind of like Amazon starting to sell books, but then later on moving to other things. That's more of an implementation detail as opposed to like the underlying, you know, core foundational improvement that you're making. In the morning, we talked about hiring great talent, performance culture, people, et cetera. Uh over time as you kind of built these great organizations, what kind of views have you kind of come to or strong opinions that you hold when picking great talent or interviewing people? And also on the other side, and how do you think about reward, especially in our

industry, you've got this tokenization that's so unique to us and has the ability to attract people or or great talent because there's liquidity much earlier than in most other places. So, any any points of view or the strong opinions that you hold on just picking great talent, interviewing them, and also on the reward side of it? Human projects and human organizations are primarily made up of humans. And so, like you you have to like pick extremely capable like humans that are going to work really well together. So, um you know, caveat that might change with like AI models, but anyway. Um the like what I have found like first off like the the table stakes are like

the internet is rife with tons of really good material on how to like do really good interviews, how to like how do you build a good team, how do you build good talent, like, how do you find good talent, how do you develop and grow good talent, how do you attract people, like, all of that kind of stuff. So, I won't kind of like go super deep into like the basics. Um something that I think is like maybe different that um wasn't as well spelled out was that uh for me personally and for PL as an organization um getting very concrete about what we mean by capability differences um the getting very concrete about that became very useful. So, con- what I mean by that is you know, like, the leveling systems that were built in the in the um

in a lot of companies like Google and others. I think this comes back to goes back to Intel on that. I don't remember, but like um along the way Silicon Valley developed like the leveling systems and frameworks and whatnot. I found those extremely useful. Um especially if you build if you get like very concrete and very clear about what you mean by all these different levels. And um we developed a thing we we call like level color coding where we have a uh CircleCI build like this competency matrix for software engineers where like you have like every single level as a column and then the the rows are like different modalities or whatever and each cell is like expresses a very concrete behavior or capability that a person is should be able to like master in that particular level in that particular area.

And uh what we do is we take that matrix or not we developed our own matrix, but like we we just color code it based on like grading people how well they're doing that right now. And that gives you this really nice property where like over time as people are growing, they're coloring that sheet and it's you know, turning green along the way. And that is super super useful to very concretely and and clearly talk about what are people doing really really well, what are people what are people kind of like doing kind of well or not well, you know, yellow, and then what are people like um what what are the gaps. And that gives you also like a very good growth trajectory people can now look at like the next level and start going towards those things and you get concrete about the things. To your comp

question, um another one of these things that were a ton has been written about comp um online and and and so on. So like lean on all that for for the table stakes. What I would add on top of that is like maybe two larger large things. One is you you want to make compensation proportional to impact. Like so many companies get get this wrong and I've just have seen over over time so many cap tables that are like messed up when like when people look at it, they're like, "Okay, well, is the ownership of the company proportional to the impact people have in the company?"

And like you really the answer's always like it can't possibly be exactly yes because it's impossible to get to that perfect answer. Um you know, you would have like compute sharply values and so on like which no one's going to do. Um but um you how wrong you are like really matters. Like you should strive to be as like close to perfect as you can. And if you build that good incentive of like um if you have an organization where like the impact you're having to the organization is proportionally rewarded, like you get an amazing level of of outcomes where everyone can super align with each other and can focus on doing their work and and whatnot. And so that's one of the key things that I think Silicon Valley pioneered as well as like highly um uh highly meritocratic compensation

structures uh associated with like the impact like very direct evaluation of the impact that people are having. Um and my guess uh is that we can do that in crypto networks at scale in decentralized protocols. No one's done this yet, but I think you could do it in in a fully decentralized setting where you can like evaluate people's contributions over time, figure out their impact, and then do automatic reward structures and systems. Um so I think that there's like a ton of possibilities there to like build the traditional kind of ways that corporations organize people in a fully crypto native decentralized setting.

It's just there's a lot of work to do to get there and usually it's kind of like the second or third, you know, order away from like the main problem that someone's trying to solve so people don't invest as as deeply into this, but but uh but I probably would and slash um like a bunch of the projects that we're working on are like working on these kinds of things. If you could just communicate to us through books for instance just to like really get dense information passed on, if you could just choose like your favorite non-fiction book and maybe your favorite science fiction book that would like inform us and like help us like like be able to understand that, please. I think The Beginning of Infinity is one of the best books that humans today should read um and it's it's written by David Deutsch. Uh it

talks about it's a very good framing of like how science works, why it works, um uh there's also The Fabric of Reality also by David Deutsch. Uh I think it's like a best philosophy work in the last among the best philosophy work in the last 50 years. Um I think there's a kind of kind of like a bunch of important like science popularization classics like um that that are that are I think quite good. Um things like, you know, Surely You're Joking, Mr. Feynman and um uh things like um The Selfish Gene and whatnot. So there's a whole like category of books that are worth like, you know, a few hundred books that are like that. Amazon recommender is awesome. Like you can just like pick a few of these books and

then just go look at like viewers also bought uh or like, you know, whoever bought this also bought these other things and you just like keep going down the list. Um very good way of like finding Oh, the the Idea Factory about Bell Labs was super formative for PL itself. Um I think for sci-fi um huge fan of sci-fi in general. I think it's extremely extremely useful in like helping think about the future. Um it in that it you know, puts you in a bunch of different perspectives. Uh recently a book that uh I've come to really appreciate is uh the uh We Are Legion, We Are Bob uh book. It It's the Bobiverse series and tells the story of a von Neumann probe uh going to like discover the galaxy. Uh someone gets like cryogenically frozen.

This is not spoiling anything. This is in the first chapter, but somebody gets cryogenically frozen and then he wakes up and um things happen and then he's suddenly a von Neumann probe. Um it's super fascinating. But it talks about a lot of these kinds of things. Of course, all the Neal Stephenson books. You can go back to like, you know, the classics like Dune and Foundation series and so on. Like definitely highly recommend. Yeah, I would I would also like orient towards reading like usually when something is very popular, that means it's a part of like the mainstream opinion now, which is A, you should definitely probably read it because it's kind of the table stakes of what everybody else expects, but B, it's probably not where the contrarian ideas are.

And so I would look for things that are not popular yet. That maybe are popular in a few circles. And so kind of extract a bunch of recommendations like that. And with that, I would say Nick Nick Bostrom's papers are some of the you know, like highest impact papers that I've read on philosophy ever. Like when I when I was kind of like um growing up and going through college or whatever um there were all these like important philosophy classics that they you know, people are told about. You know, you can go back back through like the bunch of different philosophy history or whatever. But at some point, philosophy forks and there's like the traditional like actual, you know, proper philosophy, you know, capital P philosophers that are in

universities and whatnot. And like they do like a very different philosophy than, you know, David Deutsch and Bostrom and a bunch of other people that are like actually thinking about how the future is going to unfold and helping humanity reason about it. And that is like super super high value value stuff. Um, but it's probably, you know, denser than, you know, the sci-fi books. Still amazing. Thank you. There's also a lot of great sci-fi in blog posts. So, there's a lot of like really good blog posts that are like short stories that are extremely good. Hey, back to the disruption conversation, if I may.

So, you know, they say that just in the US, there's 3 million jobs in in customer service. You know, and even before the last version of chat GPT came came out last week or the week before, I was looking into some uh phone services, automated uh AI phone services for customer service, and it's pretty it was already pretty good, right? So, 3 million people in just one industry in a country of 340 million, that's like 8% of the population, right?

Um, are are we at the 10-year mark right now for and how do how do you think that that that we are going to deal with that major shake in So, are you are you concretely asking about AI specifically, or are you talking about in general how technology tends to reshape jobs? I'm thinking more of AI right now. So, yeah, I do think that like, you know, people have been talking about the amount of automation that computers were going to bring for about a hun- 80 80-ish years now, and in that entire period, it was kind of oversold.

Um, and so now most people are like attuned to the idea that people are going to talk about this, and it's not going to happen. But I do think that now it really is going to happen. But But the weird thing here is like, you will simultaneously have AI systems obliterating a certain certain of jobs, but at the same time creating a bunch of new ones. And so, the hard part here, or like the the the opportunity, is how do you use these new systems to rapidly retrain people to do the next generation of jobs that are suddenly available. So, today you have all kinds of industries that are super, you know, like They're like talent talent hungry. Like, there's not enough

people that can do the work. And there's like millions to tens of millions of people around the world that don't have jobs, or like Is it hundreds? Maybe tens, right? Um Maybe hundreds. Um How do you like solve this? Hey, you have like these amazing superpower like superpower machines. Like, you have a way of like training everybody through software. You could do that. Um I've often thought that edu- direct-to-consumer education tools that are oriented with very concrete jobs at at the at the outcome, um could be tremendously transformational and helpful in this transition.

Um Separately like I do think that we are going to be facing a moment where Like, if things continue as they are right now, um as we unlock higher capability intelligence, um That's going to be like a like a That's going to steepen the the power law distribution with like people's access to to good ideas. Meaning, like, the people that know how to leverage these systems are going to be able to leverage them even harder than people that like just don't know to do that.

Um And so, without some kind of economic shift or like a ton of education really quickly, um this can leave behind a ton of people. Uh so, I do think that like, you know, in a hundred years, we have new Like, it is hard to envision our current growth trajectory and and think that we still have the same economic systems and the same economic structures and things look roughly the same. So, clearly something something between now and 100 years from now is going to be radically different. Um when is that going to happen exactly? Is that like the next 10 years or the next 30 years?

Or is that later? Unclear. What are those new systems? We with crypto, we have the capability of investing in that question in a way that almost no other group can. We have the tools and capabilities of creating new economic models and testing them with millions of people globally. And that's like a super powerful thing, right? Like when you think about most economists in history, they never had this ability to just like dream up a new economic structure and then deploy it and see if it works. Like that's amazing. So, we should be figuring out these new economic structures now.

Um and that might be like an enormous contribution that we can make as these shifts happen. Um I do think that like as we go now in the current trajectory we have vast amounts of capital and now we're starting to get like, you know, by by the way like the human I don't know if you noticed, but like now there's three or four large-scale like humanoid um startups, humanoid robot startups. Like forget about the call centers, like the the if those ro- robots got good enough, tons of other jobs would get displaced.

And so, if we're not careful, we can accidentally create an economic structure where like an enormous amount of unbounded economic value gets created by entirely automated. And that goes into like who owns those things, who owns the corporations that run all this. And if it goes the current way, then the kind of like owners of those corporations are very small amount of people in the world. Because today the common thing that all economies are oriented around is jobs, as opposed to ownership. Right? So, if we if we shift that and we create broad large-scale ownership of these infra of the infrastructure, ownership of these systems, and enable people around the world to like own a piece of the future

that way, I think that could be super transformational. Um but you have to like educate people on the on how economics works, why you want to own something, why owning something is more valuable and producing than having a job and so on. Which is like, you know, hard hard problems. Yeah. That's why we need crypto. Yeah. Exactly. No better place than to leave it than that. So, uh thank you, I'm I'm off. Thank you.