The State of Filecoin - TOKEN2049 Singapore 2023

hey everyone great to see you in Singapore uh this is a fantastic uh place to be a great conference uh I'm here to just give you a quick update about falcon as you know web3 brings trust to internet interactions the key thing that we're all working on is bringing rights to the internet layer so that we can coordinate much more effectively on all our applications but in order to do this web3 must scale currently web3 is not at all web scale you can't when you think about most of the consumer applications that you use on a daily basis almost none of them use web3 because the underlying Technologies aren't yet scalable enough to be able to support these kinds of applications think about social networks think about your personal storage think about your productivity applications how many of them actually use web3 close to zero the underlying problem is in the scalability of blockchains themselves so I'm here to tell you how filecoin itself is working on this large-scale problem and aims to bring that scalability to web3 so that we can cross the chasm popcorn is a crypto storage um Power a crypto powered storage Network that creates a Marketplace between storage providers that store people's data and clients that hire those storage providers to be able to store them the data the rough plan that we've been following is this three-step process step one build the world's largest decentralized storage Network step two onboard and Safeguard Humanities data in that storage and step three bring compute and applications to that data so in step one we've basically achieved over the last few years we've built the world's largest by far decentralized Source Network and it is now large enough and big enough that that it can compete with large-scale storage networks from the centralized Cloud we have over 10 exabytes of capacity in the network which is an enormous amount and all of these are um high grade scalable data centers just like the ones that you would see in normal uh web 2 applications there are large-scale Hardware vendors that are leaning into the Falco Network things groups like Seagate AMD and many more that are helping the storage providers and helping those data centers get Scaled now what the community has been focusing on the last year is an onboarding and safeguarding Humanities data in this massive scale Network and we've been scaling up the onboarding of people's data into the network uh we've been doing that through a set of on-ramps um and now we have about 1400 clients each of those uploads either small amounts of data or huge vast quantities uh and we have about 23 are uploading large quantities of data so there's over 100 terabytes a few clients are uploading vast amounts petabytes uh scale the onboarding of data into the Falco Network happens through a set of on-ramps that help certain applications bring their data into the network and a set of crypto economic incentives that enables that data storage to be extremely cheap an example of some on-ramps include things like web through the storage that it may that powers over 50 000 users to store over 100 million web3 objects this is starting to hit like proper uh scale of of web web 2 level achievements but this is all for small data so the total amount of data here is quite small uh similar to Webster storage there's things like NFTA storage which is a specific um storage platform for that kind of assets there's lots of different kinds of applications using systems like these so everything from social media applications to uh video platforms games all kinds of things building on top of the pathway Network but what I'm most interested in is large-scale data sets coming onto the network these are the kinds of things that you need for large-scale compute platforms this is the kind of thing that you need for scientific research this is the kind of thing that really puts an impact in um in the world and this is where a large amount of the community has had an enormous amount of impact over the last year we've been scaling that that data onboarding and we now have truly massive scale data sets onboarding onto the network so these are specific examples coming from universities around the world that are trusting thought point with their super valuable precious data and using the network I mentioned the crypto economic incentives one of the interesting things that that Falcon enables is this um under undercutting of the prices of web 2 storage markets and it also enables certain kinds of things like um certain new kinds of structures that give you additional incentives so for example um biddy managed to create this two-sided Marketplace between stores provided some clients and it's able to use that incentive that I mentioned earlier to even get Negative with price storage so that means that large-scale clients might even get paid for bringing the the data to the network whenever that the economic conditions mean that applications desire that data so now what I've been focused on um extensively for the last year and a bit and what a lot of the communities focus on this year and next is to bring compute to the data to enable an application runtime that can do that scalable computation that I talked about earlier so we're doing this in kind of three parts the first part is fvm which brings uh smart contracts to the network uh that shipped earlier this year there's already a large community of Builders um deploying their applications on fem um there's a Defy is one of the main components um of the network because there's large scale businesses all the storage providers that require Capital to be able to grow their operations and so you have an example um fully decentralized Finance style economy providing and supporting for real businesses doing traditional um kinds of things so this is an example of the power that blockchains give you realizing transformative change in some industry in this case the underlying platform that we can build a lot of applications on so um The Source writers use D5 for things like um being able to uh borrow uh platform for collaterals being able to raise capital for to finance their operations um and so on people are also of course building applications on top of fem so things like content delivery networks or other different parts of the the entire storage runtime uh these will be think of these as entire l2s that are building and booting up on on platform we're also getting Perpetual storage so one of the big complaints that the community had about Falcon is that we didn't have Perpetual storage uh contracts and now those are finally arriving into Falcon this year which has been great to see so now you can um for certain amounts of data you can now hire the network to sort the data for you forever now now that fem has shipped and it's uh growing a lot of the community is turning its size to the next part of this plan which is to bring compute to the data so this means enabling a set of computational l2s on top of five coin uh so there's this large set of trade-offs in compute where when you think about um zero knowledge proofs of or fully common encryption or optimistic computation they're in different parts of this large trade-off space between verifiability privacy and performance and so these will mean that you'll have different networks using different um different cryptographic schemes and different hardware and different teams uh orienting for specific different use cases so instead of trying to build one compute platform that tries to do everything but can't actually succeed at that we're building Falcon as a way as a great home for all of these different networks to bring their computational Primitives to the Falcon data uh we we said in when you think about large-scale data storage data has gravity once you have a lot of it it's very difficult to move it um so once the data has been onboarded onto falcoin it is much easier and much cheaper to bring the computation to the data run the computation there and then export the results uh so a lot of the um computational uh l2s I'll give you some examples so one of them is ZK Lambda where they're trying to build a um a a structure similar to Amazon Lambda type computation but entirely in zero knowledge that you and that is an entire computational L2 another example is um the back allow project which is building Primitives for these type of cod networks you can think of this as kubernetes or Docker or those kinds of runtime environments where you're able to Define some program ship it into the network and have it compute uh some of the example applications that people are building is generative AI use cases or llms or things like that and you can now use the storage providers and the hardware that the search providers have for these kinds of kinds of Networks so then the last step in this plan which is um where some amount of the community's focus is going is into scaling the blockchains themselves our view is that blockchains today fundamentally need to scale dramatically we need to get to billions and trillions of transactions per second this is something that I've said for many many years and we're finally getting there um the entire Community most of the blockchain space is targeting something like a hundred thousand or maybe a million transactions per second uh but I think we need to go dramatically higher scale to be able to support the applications that I talked about earlier so that's the kind of stuff that we're we're building and we um we designed a protocol called IPC uh which is stands for interplanetary consensus and that's what enables us to scale chains uh to that level of um uh scalability cool uh the last thing I'll mention is is there's a lot of Builder it's a huge Builder community in falcoin um there's everything from hackathons to accelerators grants and VC funds that have been built this great pipeline for developers so it's an amazing time to get involved especially right now in the winter when capital is scarce um coming to a community that can help you build uh can help partner with you and build your build your business is great and so um I encourage all the Builders out here that are trying to build startups to come talk to us and um and take advantage of the programs here uh cool so with that that's uh all I have for the update and I have I think a few minutes for questions if there's any sure any questions anyone no cool yeah back there yeah great question so um the question here is in the can I get this life back up so I can go back to it I'll start repeating the question so in the Saturn CDN um how are we decentralizing the gateways themselves because that's a big um problem in in um here let's see there we go um so the way that Saturn Works uh a CDN starts for uh sorry the question was um for the CDN how are we decentralizing the gateways the gateways are these uh services that are in between the client's computer when they have no verifiability and the rest of the Network that has verifiability and the gateways are a hard to trust um thing because they could be cheating and delivering something something false um so yeah great great question um so the way that Saturn works is that it has it has a layered architecture so it has different layers of cache and so you can think of the source providers as the bottom layer of the cache and that's kind of where the data is but it's very difficult to move it out to the user and each layer makes it faster getting it closer and closer to to the user and so the last layer the first layer is kind of where the gateways live and so for that we're we're solving it into in two ways so one is we have some JavaScript code that can be so there's different kinds of clients there's clients that can run JavaScript and clients that that cannot for the clients that can run JavaScript we we are enabling a service worker upgrade where there's a specific um of code base a specific endpoint that is well known in the network where that delivers just one payload that you can that everyone can kind of agree on and that loads um a client into the browser that can then verify the rest of the content so that all of the content is precisely what it's supposed to be now there are some clients that cannot run JavaScript and so for those you have to run a set of trusted gateways in a traditional setting but you want to minimize those and you want to enable more parties to run them and audit them so today we run the ipfs Gateway which which handles billions of requests for per week and that's um you know thousands of applications use that Gateway and that's the trust model that they're relying on we don't like it um and so we want to go to deliver JavaScript to the user's browser to be able to verify it but yeah it's a great question uh any other three more minutes any other question yeah back there so um I know this is mainly for fight coin but I have a question for ipfs sure so in your load map what's the plan and the role for ipfs in the future ipfs is a large project with a fairly large roadmap there's lots of different Builders working on different kinds of things and so I would encourage you to look online there's a number of different roadmaps depending on what what uh what area you care about so there's like about there's a few hundred companies working on different parts of ipfs yeah any other questions cool awesome thank you