 All right, welcome back everybody. Our next and final speaker needs no introduction, but I'm paid for introductions, so he gets one anyway. Juan Benet is the founder of Protocol Labs, the Research Development and Deployment Lab, which developed IPFS and Filecoin. He's committed to building a better future using science and technology. And today he's going to round out Crypto Economics Day, Filecoin Crypto Economics Day with a talk about solving large scale problems with Filecoin. Hey everyone, great to be here with you. Awesome day today. So what I want to talk about is how we can use CryptoEcon on Filecoin to solve massive scale problems. Over the last, certainly over the last 10 years, it was already getting this sense, but I think really the accelerate over the last three to four years, I've come to see the stuff that's happening with Web3 in terms of incentive design and mechanism design as the largest levers that humanity has right now to fix massive scale problems, because they enable this kind of incremental growth of some solution in a way that other microsystems have not quite enabled. Really awesome day to, and great to dive into the depths of the platform CryptoEcon. This last talk is gonna go more broadly into thinking about how we can use this CryptoEcon principles to shift large scale, other large scale systems, and how we can use the facilities of Filecoin itself to do that kind of thing. So I gave a talk yesterday at Shelling Point called Achieving Paratotopia with Crypto Economics, which is kind of a prerequisite. So if you are in there, bummer. However, now I'll summarize it for you. The basic idea, the basic summary of the talk is there on the right, which is if regen, whack me, else, and GMI. The point is, if we can create regenerative Crypto Economics, we're all going to make it and we're going to advert disaster, otherwise we're not. This has been, this is an extremely critical century, though there's been this amazing range of global improvement throughout history. We are confronted with certain existential risks where we now have both the ability to wipe out ourselves and all of life, probably not all of life, but most of life, and we might screw up this massive computing phase transition that's happening. And unfortunately right now, our macro systems are inadequate. Though they've been excellent to get us through the last few centuries of global improvement, maybe debatable whether excellent is the right word, but really good. We've just had this impressive demand of global improvement. However, we're headed into these kinds of problems that those government systems are not quite equipped to solve. Now what we want to get to is kind of this part of this utopian outlook where we can use mechanism design to get everyone to be thinking in massive positive some ways where you're not thinking through small resource increases, but you're thinking of order of magnitude level increases and you're getting people to stop negative competition and engage in collaborative competition or just straight up collaboration. Ideally you want to create an environment where the use incentive structures to warp the actions that various different groups might take so that you align them towards this collaboration and it can get to the prior utopian outlook. Prior utopia is this conception that once you kind of realize that the future is so amazingly positive some, if you can get there, then it sort of makes sense for everyone to sort of band together, align on that cooperation to do the science and technology diffusion that we need to get to that fantastic future. So think of crypto economics as this like tool that we can use to kind of reshape the landscape, whether it's by kind of like boring through the whole, creating a new hole through like this incentive field to try and get to like a better optimum or by like moving mountains in that landscape itself, you can use mechanisms and drop them into the network to achieve this kind of thing. Just to give you a sense of, I think everybody here is already paged in into how powerful mechanism design is, but it's just every year, I need to kind of like step back and remind myself that Web 3 is really programmable. The tooling that Web 3 gives us enables us to really reshape any kind of coordination structure and it's extremely fundamental. It's eating law, it's eating finance, it's eating all kinds of structures. So we're gonna be seeing over the next five to 10 years and beyond this kind of mechanism design rippling through major parts of the economy and major parts of governments and so on. So it's an extremely powerful lever. Let's make sure that we use it for really good things. So for example, the Bitcoin hash rate has been incredibly illustrative to me over the years in showing just how powerful of a lever this mechanism design is. One straightforward incentive structure both enabled the world to get into the world of cryptocurrencies and actually enabled the cryptocurrency to grow, to compete with fiat currencies, but it also, that same incentive structure also created a massive run of equilibrium that just wasted an enormous amount of energy. And like, there's taken what, like 12 years or so? It's in 12 years, we went from a network that didn't do anything to now one of the largest energy consuming things on the planet, wild. And that's kind of like, without the massive, that's also having to bust through the walls of getting people to accept cryptocurrencies and get into that idea and reforming the entire weaponry movement in its wake. Now that the weaponry movement is here, you could do, you could redo something like this in probably two to five years. With Bitcoin, we've gotten a taste of just how powerful these incentive structures are by creating an open and permissionless network. In a year and a half, we've assembled 15, 16 exabytes of capacity. Like that's an insane amount of storage. So that should give you a sense of just how powerful these levers are. Now you have to design really carefully and it's great that we are having this day to talk about the nitty gritty details of these structures. But the main takeaway here is the more you can think in that kind of scale, the more you can make open and permissionless networks and you design the incentive structures carefully to achieve good results, the better will be. And also be careful, to take the Bitcoin story as a warning sign of the incentive structures you make once they grow to massive scales, might have a bunch of unintended consequences. And so think through that. Now I think that we can use all of this incentive structure mechanism design to accelerate the science of technology translation process itself. A lot of us here in this room and in this community are working towards these goals by thinking through all of these processes thinking about how to enable funding to come to all kinds of efforts and groups that are doing work in science, doing work in technology building. We're creating new incentive structures to do that technology translation. We're participating in the public goods ecosystem. We're creating events to enable a lot of people to hear about these ideas, present some other projects and so on. And we're kind of trying to create a dramatically better kind of funding staircase for all of these public goods type processes. So it's a lot of work going on in here, but I think it's been like, a lot of people have been talking about this and events like Shelling Point and funding the commons and others are really great for that conversation. I'm gonna give a different example today, which is looking at the Falcon Green project. I think this is a great example of using mechanism design as this massive lever, breaking down a massive scale problem into, the picture of this talk is planetary scale problems, right? So how do we break a massive planetary scale problem into smaller bits? And the Falcon Green project is a great example of doing this. So I'll tell you a little bit about Falcon Green first and then tell you how to go from a smaller problem to a larger scale problem. Falcon Green is a project to decarbonize Falcon and in so doing, figure out systems by which to decarbonize the rest of crypto and the rest of the computing infrastructure and potentially other infrastructure. We start small, we don't try to tackle the whole thing. We try to tackle something much smaller. Let's think of just the Falcon Network and the energy use within the Falcon Network. Let's measure it, let's figure out verifiable structures for identifying where the energy is going, what kind of energy is it, and then let's use financial instruments and structures to offset all of that energy use. And ideally, let's try and start applying pressure into the hardware manufacturers to produce better greener hardware. So that's a stuff that maybe will come later that I don't think there's a ton of progress on this yet, but it'd be really awesome to start shifting those incentives. One of the key things here is to take a large problem, break it down into something that you think will scale well that's achievable in the short term. If you try to tackle something way too big that you just can't make progress on, you might spend many months to years tipping away the problem and you won't achieve that growth. By grabbing something smaller that works good enough that isn't perfect, but it's good enough to show people that it can scale, and it's already kind of directionally correct, you kick off a movement and you kick off this broader range of improvement that can happen. So how this project works is that it's building a bunch of different sub-projects and sub-tools that are meant to kind of work together. It's not like a big monolithic thing. It's kind of a modular ecosystem-oriented approach to making this kind of change. It's building dashboards that track the energy use and estimate it, so it doesn't have a perfect estimate. It doesn't matter. It creates a lower bound, it creates an upper bound, it has some estimate. Then using that estimate, we can then go around and buy REX. REX are renewable energy credits, which enable you to one of the really cool things about REX and why they might be better than many other carbon offset type things is that they enable you to show precisely where the energy is coming from, in which grid, from which producer, at which time of day, and what kind of plant it is. That lets you, in a verifiable way, get a sense of precisely what type of energy is being put into the grid, precisely at the time that you're pulling it out of the grid. And so there you can have a strong claim that the energy you're pulling out of the grid is precisely the energy you're buying that you're putting in, which gets rid of a bunch of the nasty problems around carbon credits. So carbon credits is a great idea, phenomenal idea to use the financial system to carbonize the planet. However, if you create a bunch of loopholes, you can end up with a bunch of unintended consequences. There's been a bunch of stories of that in the past. So renewable energy credits can solve a set of problems there of trying to introduce way more verifiability into the picture. I would expect this kind of thing to be scaled, like get to the point where you have machines in the plants themselves that are able to run some zero-knowledge computation or that have like tamper-proof hardware or stuff like that, where you know precisely that somebody's not messing with the estimates and so on. But we got to start somewhere and we got to create the incentive structures for the whole system to get better towards that. By creating significant rewards for this kind of work, people will progressively get better, will get better and better at doing these kinds of things. So the Falcon Green project, in addition, is enabling a lot of other groups and people to get involved. It's enabling individuals who have other ideas or new ideas or want to pursue some of the ideas that the project has had to go try them out. And this might be through different kind of grant systems. There's a bunch of different grants that the project is sort of connected to. And it's building a whole set of collaborations with a bunch of other groups to kind of follow the path of least resistance towards greater and greater impact. It's also enabling a lot of people, individuals, that are interested in the project or interested in the ideas to actually get involved. And this is a crucial thing that I think I see tons of projects forget. If you create good avenues by which individuals that are interested in helping you, where they can follow the information, where they can participate, where they can learn about what you're doing, you can turn something that might be a small project for a small group of people into a massive movement that a lot of people are gonna work on. And so things like meetups end up being really important, creating a sustainable blockchain summit. That is not the sustainable falcon summit. Creating a sustainable blockchain summit creates the opening to get a bunch of other crypto groups to show up and also talk about their projects to decarbonize their chains and how to use that stuff to decarbonize the planet. And so by creating kind of like all these like smaller level levers that align incentives with other groups, you start kind of like tackling this larger problem. By the way, fantastic talks from that summit and can't wait for the next one. It's a bunch of like really, really good stuff there. And you know, like you can follow and participate directly asynchronously in the network. That's kind of like the technology behind falcon green involves looking at the actual physical machines that are running the falcon network, estimating the power consumption and then getting into the verifiable claims stack that the renewable energy credits community has built and gluing these two things together, building a market where you can precisely buy wrecks for precisely where the SPs are purchasing energy and so it's connecting a lot of things together and enabling the thing to work fully and kicking off an incentive structure for the whole network. Now then the kind of the whole social incentive structure, part of this is, you know, you're using kind of news and grants and event forming and so on to catalyze a larger scale action in the community. It's both like incentives directly in the network and the people that are running the network and incentives for the broader community to actually have an impact in the broader thing. So let's scale that up. So if we get to decarbonize falcon and we can solve that as a problem, we can then start scaling it up. We can work with other chains that are doing the same thing. We can talk to other chains about how to approach this problem. We can give them a bunch of the technology. We can encourage them or incentivize them to do this by saying, hey, look, our chains way better. It's way, not only can we get to carbon neutral, we can get to carbon negative. One of the things that I wanna get to is like be able to say that, you know, falcon is like, I don't know, 5x or 10x carbon negative in an area so that whenever falcon is there, it is actually like pulling carbon from the atmosphere around itself for that nation or something like that. Like that, I think that would be pretty awesome. And if we can do that, then we can incentivize a bunch of other groups in the blockchain space to do that. And then if suddenly all of crypto itself is doing this and we get rid of this mistake in our past of like having produced the big massive energy waste that Bitcoin is, and we can not only offset all of the other chains, but also Bitcoin and we make all of crypto, not only carbon neutral but massively carbon negative, then at that point, the whole world is going to start paying attention and suddenly other industries, entire other industries will be incentivized very strongly to catch up to the crypto space. Once you get a bunch of other industries incentivized to catch up with the crypto space, you can then finally get to a point where enormous large groups are getting to be carbon negative and we can kind of unwind this like crazy set of emissions back down. So I think like all of this is start with like massive scale problem, break it down into smaller groups and find ways everywhere along the way from the small scales to the medium scales to the large scales of creating incentives by which like you cost coordination, you lean into competition, so collaborative competition, you wanna get the different crypto communities to be competing with each other of like who's the greenest blockchain because that's what's gonna get crypto itself to be very green and then once crypto itself is green, then we can compete with the other industries about which industry is the greenest and so you wanna use the sense of like highly collaborative, competitive, open environments at every stack. So that's kind of like an approach to solving a large scale planetary problem. Now, of course, like there's an enormous amount of work in getting this done. There's all kinds of systems that we need to improve, all kinds of systems that we need to design and build and so on, all kinds of processes that we need to kind of like refine, make verifiable and so on. This is an area where all kinds of tech is going to run the crypto space from it and now in a different angle, all kinds of other really amazing improvements in the crypto space can help this kind of computation. So all the zero-knowledge proof stuff, all the verifiable computation stuff is phenomenal for this kind of thing. This is this kind of problem and many other problems in the world have a kind of crisis of incentives and trust where you want certain parties to take certain actions but ideally you want to not have to trust them. You ideally wanna be able to provide a verifiability layer to what's going on so that you can trust the entire process and the entire system. I think like all of the zero-knowledge proof type of stuff can come into all kinds of facets of broader industry and broader world activity to improve how those systems work and enable you to create financialized structures that actually work. One of the big kind of points of contention that the broader mainstream world has on financializing the solution of problems is that it's very easy to kind of warp the incentives and end up with all kinds of unintended consequences or just like bullshit claims outright where like parties are lying about the stuff that they're doing but that's where verifiability is really important. Be able to create verifiable structures that keep removing sources of potential problems. The thing like all the deep cryptography work that is going on in zero-knowledge proofs it's going to be like a very large lever at solving a bunch of these problems because you can take those kind of verifiable claim type structures, weave them into these larger industries and then kind of then at that point really financialize them and scale them into massive scales. One thing that's gonna help here and I don't have a slide for this is certificates of impact are gonna, I think, and there's other talks about these. I do think that once we have these verifiable claim structures, certificates of impact present a way to financialize almost anything which is an extremely, extremely powerful primitive. So we'll see a lot of that stuff coming out over time. One other thing I wanna reflect on is the Falcon Plus incentive structure is really, really useful. It gives us a way to incentivize the gathering, collection, assembling, cleaning up of and making usable, yeah, it enables us to identify a bunch of data sets that we can collect, gather, improve, and refine and so on and make usable. And so we should be using these incentive structures to go after massive scale problems. So for example, think of massive scale data sets that could be really impactful in certain problems that are really big and really expensive and really hard to deal with or that are really far away from being able to make verifiable or are very disconnected from the economy. You can start doing things like think through some really important data set that you think is gonna be enable some kind of financialized way to solve some major problem in the world and say, okay, great, what's the data that we need so that we can get a bunch of speculators to speculate on the prediction market about what's going to happen? So for example, let me go to this first note, return. Think of being able to make claims about the planet. So there are data sets like Landsat. Right now, Landsat is not really connected to the crypto economy. I don't know if hedge funds trade on this, but I'm sure they trade on all kinds of activity that is totally viewable through Landsat. So if you had like a real-time connectivity to exactly what's happening on the planet and you had blockchains being able to act on that data right away and you had like massive scale bets going on and precisely what's going on and you had a verifiable way to look at it, then you can start making certain kinds of trades possible and at that point, you can then start warping the incentive field to achieve the outcomes that you want. So this might be ways of like holding groups accountable for certain actions or this might be ways of like responding to certain problems or it might be ways of like settling important forecasts and predictions, right? So the crypto world has, not just the crypto world, but a lot of groups have pointed out that forecasting of massive scale problems could be greatly helped by prediction markets. Unfortunately, prediction markets aren't quite allowed in a bunch of places. That's a bummer, in the crypto space, there's gonna be a lot, I'm guessing there's going to be way more use of prediction markets. So hopefully we might end up with like prediction market Dow things that are able to take significant bets about cases like this so that we can understand what's going to, what certain actions are going to do in certain regions. So you can then at that point get a lot more information about certain kinds of problems. Prediction markets are an extremely useful structure because they let you build a lot of trust in specific actors that get extremely good at predicting the future so that you know ahead of time how much you should trust certain kinds of possibilities and proposals and so on. So all of this kind of data, Landsat is pretty big, especially if you wanna get like higher resolution imagery or you wanna get not just like visual imagery but you wanna get all kinds of other kind of data about the world. But once you, there are these massive scale datasets about certain, like the actual brown for example, like massive mining companies have to generate like the insanely complex, huge data sets about all of the elementary data they're gathering in certain regions. All of that data you could bring into Filecoin. You can use the Filecoin Plus incentive to just cause all these groups to gather that data. And then once you have the data connect it to smart contracts and then financialize it which is super, super, super interesting. Now, the stuff that is gonna enable a lot of that stuff is I'll say it for the 50th time today, the FEM, the FEM is gonna enable that kind of layer to emerge because kind of like what we've been missing is the hook where we can start creating computational networks around Filecoin, right? So Filecoin right now doesn't have smart contracts once it does, then create schedulers and then start doing data pipelines around all of this data. So the missing piece is like, so you can bring all that data into Filecoin. Now you wanna connect it to a smart contract and you need the ability to run arbitrary code in ideally a verifiable computing sort of way. And so like that's a, things that are gonna happen over the next six to 12 months is building out all of these kinds of computational networks. So things like, and when you think about like a Filecoin storage providers, they have these massive scale deployments and they already have a bunch of GPUs. That's precisely what you need to build a massive computational network that can operate over this whole data and that can run all the expensive proofs for this kind of information. So this kind of thing of like taking a data set like Landsat and thinking about a certain kind of action that you want to incentivize and building all these infrastructures around it needs these like two missing pieces. One is like the smart contract layer and then after that like the ability to run verifiable computation over that data set. And that's coming really fast. That's gonna be here in the next six to 12 months. And at that point you can do, you can put all these pieces together and get that kind of action. So this is one example data set. There's probably hundreds of these kinds of example data sets. So we're building these massive levers to like move mountains and move the world. There's all kinds of different kinds of use cases that we could lean on. Probably should have some conference about this. Like what kind of large scale problem is connected to some massive data set where gaining the ability to financialize that would produce some major positive impact in the world. So maybe like get to do for the future. Like maybe another CryptoEcon day could focus on what massive scale problems can we solve with CryptoEcon. Great, so I think probably the last thing I'll mention here is that at the end of the day like the tools that we're building will only be successful if we can get them into the market and use by large groups of people. So that means again breaking down a massive scale problem into smaller and smaller and smaller pieces starting with that construction, creating an incentive structure that enables a lot of other parties to come in and contribute. You want to use collaborative competition because that's a very useful structure to cause many other groups to go as fast as possible towards some goal. You want to again be very careful about the incentive structure you design because if you screw it up then you get the Bitcoin proof of work waste. But if you can do that from smaller to larger to larger and larger and larger structures then you can truly move mountains and you can fix the planet. Great, thanks. And I don't know if there's time for questions but I think the time's up. So I'll stop here. Thanks Juan, after a talk like that there's always time for questions. Questions to the speaker. I'm gonna go back. Yeah, I think the question I have is who designs these incentive systems? And yeah, that centralizes the power in one way in their hands. Yeah, I mean I think blockchain networks are vastly more open about this than like you can propose a design change to a blockchain in a way that you can't to a financial, you can submit a FIP. You can submit an EIP. You can't do that to the US government or to the IMF and so on, right? Which is kind of like wild. So I think there's all kinds of structures that can and should be improved across all these systems to get to a better spot. I think CryptoEcon in particular, there's like all this complexity because you're designing incentive structures there's all kinds of groups that will benefit or lose from certain mechanisms so it very quickly gets heavily politicized and problematic. But if you can kind of like remove that as much as possible or get to a spot where like there's common knowledge that certain things are really useful broadly for the network and aligned with a long-term mission then you can build significant agreement in those networks to get that kind of change happening. A good example of this is the Ethereum community is just filled with moments like this that it had to kind of break through and it shifted its own economic structures. Sometimes like the move off of proof of work into proof of stake is like a massive one that they've been doing for years. It's a huge economic shift but the community had enough interest in that to cause that kind of shift. But these blockchain networks are dramatically easier to change in this way. There's probably like in a machine learning way a hyperparameter optimization thing that we could do here which is like how do you create how do you create way more crypto activity happening and you kind of create a much more robust study of all these systems. I think right now we know this in the blockchain community like the level of rigor that we bring to crypto econ is way higher than many other groups and even when you integrate put all of that together across the best groups and all of them, that is probably way lower than what we want. We ideally want something 10 to 100 to 1000x better than whatever we have going on. And so finding ways of doing that might be really good. Potentially might be like causing there to be 10 to 100x more people going into studying crypto econ. We might start with a crypto econ textbook. I've heard that said before. I think David who helped design a lot of the platform crypto econ I think said in the past like, oh wow this is like super powerful. It took me a while to learn this. It'd be great if somebody wrote a crypto econ textbook so that a lot of people could like have a much faster path at understanding what you need for the mechanism design. Because you have to learn a lot. You have to learn a bunch of game theory. You have to learn a lot of distributed systems. You have to learn a lot of cryptography and so on and you end up in like pulling pieces from a bunch of things that to actually get to cryptic on. So there's like a good high leverage thing for somebody to do. One more? I feel like what, right I've had a lot of conversations this week about privacy anonymity. I think there's a lot of people that value maximizing that in the web through ecosystem. And yet we're looking at like planetary surveillance maximization as well. Do you ever wonder if, or do you add, what are your thoughts on anonymity privacy maximization potentially being anti-progressive in the context of what you're talking? Yeah, it's a huge problem. And it's very complicated. So maybe like in the most succinct way I would say that there's we're an integrated system. Like the idea that there are individuals and groups and so on is like useful fictions for us to operate. But at the end of the day you have like this massive integrated system when you go down to atoms and so on. And so you end up in a situation where at various layers of the stack you wanna have a different part of that solution. For example, at a planetary scale it's pretty important for the whole planet planetary community to be able to understand what's going on with the planet. What nuclear tests are happening what emissions are happening and so on. What oil spills are happening what trash is being spilled into the ocean or picked up or whatever. And maybe even at the range of a city you might wanna understand what kind of like problems are happening in certain streets or whatever like you think of Europe is filled with like CCTV all over the place like walking around Amsterdam you can see all the cameras. I'm sure that they reduce crime. I'm sure they also violate privacy. I'm sure that they also create the conditions by which you can do digital digitally to tell terrarium things and like that's terrifying. So it's a really hard trade off. I think the where I sort of land on this is the more you can create you can agree on the policies and the preferences and what is a good mission and the more you can kind of enforce the technology to follow that and only that and limit the scope of potential misuse the better off you are. And so that means leaning into verifiable computation, leaning into homomorphic encryption leaning into all of those kinds of kind of privacy preserving techniques that still allow you to as a group compute on something together. It doesn't, you don't have to have access to the CCTV camera things. You can have ML algorithms figure out like, oh yeah, this is a problem. And then at that point decrypt. So that's like an example kind of use case. But I think this is a deeply hard question that is gonna have very different answers at different layers of the stack at different points in time and in different societies and it's extremely hard to predict what's gonna yield a better outcome because it's highly dynamic. The world changes a lot and changes really fast. And so today we have a lot of peace and prosperity globally relative to many points in the past but that could change and if it changes then we might feel very differently about a bunch of these structures. So. I think we're running out of time now. So if you'd like to ask any more questions you may get to do that at the social hour after this meeting. I'd like to thank Juan one more time for allowing us to end on a very high note at planetary scale and give another shout out to all of our presenters today and for you, the audience for your questions and your thoughtful attention. I hope we've convinced you that Filecoin Crypto Economics is a really fun sandbox to play in. A really cool and very powerful system to explore some very deep questions and that maybe we've attracted you to come play with some of our toys. And now we'd like to invite you to socialize a bit and to chat about some of the topics we discussed today below in the atrium level at the social hour. Thanks again to all the presenters and cheers.