Illia Polosukhin left Google, where he made significant contributions to Google Translate and Tensorflow, to co-found NEAR Protocol.
Illia and talks about how he and his founding team’s background in competitive coding (all being various ICPC champions, an elite coding olympics) and their work teaching computers to code informed how they went on to build an intuitive NEARs Web 3 Stack, so it could easy enough to use that it could onboard the 99% of developers and now backed by the likes of Andreessen HorowitzPantera, Multicoin, Coinbase Ventures, and Pantera Capital.
Jamie: Welcome to the Founders of Web 3 series by Outlier ventures and me your host Jamie Burke. Together we’re going to meet the entrepreneurs, their backers, and the leading policymakers that are shaping web three. Together, we’re going to try to define what is web three, explore its nuances and understand the mission and purpose the drive its founders. If you enjoy what you hear, please do subscribe, rate and share your feedback to help us reach as many people as possible with the important mission that is web 3.
So welcome to the web 3 podcast today I’ve got Co-Founder of Near protocol, I’m not gonna try to pronounce his surname at this he told me not to. So, I will let you introduce yourself by name. Your other co founder is Alexander skidding off that was easier, right?
Yes. So I’ve been following you guys for some time. Now, I think when we met in Shanghai back when it was fashionable to be flying in and out of China. Every few weeks, I think we actually met at a Binance party on a rooftop. And and since then, I’ve been following your progress, because the way that you described near protocol and your approach, at least from what I can remember, because I would have definitely had a few beers by that point. It definitely struck me that you you had a very pragmatic approach to to realising this stack. It was very intuitive and very, very open in terms of your development cycles with the community. So hopefully, we can we can talk about a little bit later. And and I guess, that should be no surprise because you you Really pride yourselves on the developer friendliness of your product. But before we before we go into that, it would be good to just understand a little bit more about your background by way of summary. So I guess the most standout things that I could see was your work at Google, at Google research, where you worked on machine translate with Google Translate, and you’re one of the main contributors to TensorFlow. I can also see that with your founder, Alexander and some of your kind of key development team, McAllen, Evgeny, you were all icpc champions and medalists which I’ve never really seen in a in a web three or a blockchain team. Could you explain a little bit more about your background, the work that you were doing a Google and perhaps some of the stuff that you think carries forward and has relevance it’s what you’re doing now at near?
Illia: For sure. Yeah. So my work at Google pretty much surrendered a lot around natural language. And I always was fascinated by evolution of machine learning specifically for language. Because in a way, that’s how we communicate. It’s how we transfer knowledge. So I really worked a lot on question answering sounds, the kind of questions like if you ask them on google.com are answered by our models, Google Translate, worked on the kind of models, like was a pretty car, so for hours right now, so new AI on Google Translate. And obviously, when TensorFlow was kind of starting off and becoming popular inside Google, like my team and jumped on it, and we’d like started actually contributing back as like external team to original TensorFlow, folks, and then over time, I’m actually I lead a part of the development for actually developer friendly part of TensorFlow. So, then me and my co founder, like, as I love Google, we started an AI company. And we were actually trying to teach machines to programme. We’re trying to figure out how can a non engineer or maybe an engineer who wants to be faster, be able to describe to computer something in English, and that’s a produce a pretty much working code. Now, it’s a very, like challenging problems. People have been working for it since like 60s. But we thought we had an approach which actually was rooted into our competitive programming background. So like, for God’s sakes, I see champions.
Jamie: Yeah, yeah. So this icpc thing is a kind of international programme where starting from universities and then going to like district, region, national nation and the world is a competition like sport where in five hours you need to solve very challenging kind of computer science problems. And pretty much you actually practice for it, like for any sport, right to actually sit down and you know, do a five hour sprint of solving these hard problems. And the point the only point is is like a computer on the other side that judges how good you are. And so yeah, so I was actually like a national level, my co founder, Alex won gold medal. And so we have a pretty good kind of connection to the space and we actually use that connections to start getting data for, for the kind of teaching computers how to care because in a way, this is how we learn how to programme in the first place by like solving a lot of for simpler and then kind of progressing to hard problems. And this actually were kind of blockchain crypto into because we built a something of like data collection platform where A, we would contract people in Russia and Poland and China to solve problems for us, like go from natural English to care to go from coal to natural language to really generate data for us. And it’s actually was a pretty hard problem to pay them. Because like China doesn’t accept, like anything. Russia has its own restrictions. So and then on the outside, like we actually were building something of an open data set, we really wanted to collaborate on projects, we’ve worked with Berkeley, that actually was done song, a little bit there. And we wanted to kind of create a open platform on the other side where people can, you know, create tasks on this platform. So blockchain seemed like a very kind of reasonable infrastructure to build such an application, such a marketplace, but as we looked into it, and kind of really dug deep, our realisation was that there’s like kind of a lot of missing pieces can across the stack that really like makes, like usable, you know, developer platform. So my my co founder Alex, he actually his background is actually from building a sharded database called Mashable. And, you know, I mostly been doing machine learning but honestly, like at Google, you can get a pretty good practice in building distributed systems. As you know, I give it even sounds and machine learning systems built for, you know, ran like across like hundreds and thousands of machines. So we started just like really digging deeper into kind of blockchain trying to figure out where, where’s the limitations carry from it, but what can I do and what is the awareness within Google or say interest within Google, formally about blockchain? Because we never really hear much externally, but clearly, experiments must be carried out or there’s a kind of general awareness of the technology domain. People like to Make the straight clear distinction between what sits in web to web presumably, you know, Google is the antithesis of that, versus this new paradigm, which is, which is web three. But it’s interesting, because when you talk about near you talk about a developer platform for open web. And I guess you could argue that there’s a lot about Google that is very open.
Illia: Yeah. So open that actually a term that comes from Google and Mozilla, and few other kind of industry leaders back in, and I’m trying not to mess it up. But it’s definitely more than 10 years. It’s way before Bitcoin it’s way before, and it’s actually was more rooted around like close technologies like flash, Adobe, Adobe Flash, being a close technology that was, you know, driving that down. And like html5 was supposed to be this open lab kind of concept that really was trying to rally around developers to build Something on open technologies. So this is actually why we like open lab as a term because it actually kind of showcases that this problem, like it has been exist, had been existed for a while now. And they just kind of changes its shape. It went from being a front end problem, right? Something that is on the, you know, on the user side to be more of like a back end problem, like, Where’s the cloud? Where does the data, who’s running servers, like what services are powered and who’s controlling them?
Jamie: And so, I mean, I really like the idea is that you are directly there’s a continuum almost right? So a lot of people in the web 3 space or who would identify being part of the web three space, really feel something other. And also, they they often feel or think that they’re working on new problems, or have new perspectives and often that might just be down to youth, but I think with any basic history in information technologies, a lot of these themes are kind of recurring, this bundling and unbundling of technology. So, I mean, do you see? Do you believe in the distinction of web three? Or do you think it’s a bit of a misnomer?
Illia : I actually don’t like like, I don’t fully like let’s see, because it’s actually I think, like, if two had had the ring to it, because like, people use this like 2.0 thing as like, Oh, it’s like better. But like, let’s see, I don’t think it actually rings was like outside of, you know, kind of descriptive space people. And that’s why we’ve been we’ve been actually like trying to see what what is a better boat, what is a better term, but also what is a better concepts that people are more excited about? Right? Because, like, I don’t think like decentralisation, for example, on its own is, you know, a, like a thing to target right? Like it’s it’s a means To serve some problems to solve some problems, but it’s not exactly where, yeah, it’s not a destination. So like, the question is, like, what is the destination we can get using this as a vehicle? Right? And I mean, there’s like a lot of interesting kind of things coming out, right? Which is like this concept of, well, obviously, like, something that’s not permission, right, something that’s not controlled by, and we, you know, we know the stories for like, Google, we know the stories for Microsoft, you know, the stories for Apple, whereas it would, you know, bomb somebody from their apples from their store, or, like, kick out somebody from their cloud, yellow logs. They’re like, mission critical application, because, you know, some, some other thing happened. But at the same time, obviously, we have, like, clear, you know, regulatory kind of where, you know, different geographies right now have different regulatory regimes. And actually right now, like, at Google, for example, for you to They have an extremely complex system to maintain other laboratory compliance photos and videos in YouTube. It’s like it’s huge team working on tooling and a huge set of PMS. And then a huge set of like lawyers. And so like to operate within regulatory, of each country of each, like state and whatever. And, like, we actually, you know, part of it is like actually making cheaper sounds and things where, you know, we are kind of, I mean, in the way I’m bundling some of the stuff and making it more available, but at the same time, maybe moving the, like, who wants to watch which content from, you know, the government deciding to individuals or like self censorship or whatever, right? So, so there’s like, a lot of kind of different ideas coming in from this, from this ecosystem. And I don’t think like that, to me is like, hey, like, what we trying to get is, you know, no, like government or no, you know, corporate interest. I don’t think that that’s directly like resonates with the end user. Like, what what does resonate with them is actually like specific solutions and specific, like things that we can deliver business technologies.
Jamie: Yeah. And I mean, I guess that makes sense. If you look at your mission, your mission is to effectively onboard the 99% plus developers that don’t care about web three. And I guess how you talk about what you do is just as important, right? Because there is this habit within crypto or web three or the decentralised space to kind of create new language. Not programming language necessarily. Yeah, sure. Exactly. And, and new, the new semantics and obviously, that that can create barriers, often it’s quite low loaded language. And do you think that that experience at Google seeing operation is software and networks and systems and products at that scale and the complexities of them being rolled out globally. You know, do you think that that is an advantage and, or a disadvantage or a bit of both, right? Because I know some people who are perhaps a bit more dogmatic about what they think their vision of web three is, would, would argue very strongly that there shouldn’t be there shouldn’t be any in between. But it sounds like if I kind of stick to the analogy of web two, and web three, what you’re, what you’re really proposing is almost like a web 2.5. It’s an evolution of the web. But that is an evolution that is at the pace that developers can can use, deploy and build products at scale.
Illia: Well, I mean, what we’ve built is actually like it provides you a infrastructure for anything like you can plug in, you know, a our identity system into your business. To application and be happy, or, you know, you can build a fully decentralised like, you know, like smart contract which can, you know, custody assets and will be resistant to, you know, government attacks, right? Like we actually like on the kind of developer platform itself. We don’t target like this in between we actually have the full spectrum. It’s more of a question like, how do what is the language we can indeed use to bring more developers like to actually excite them was, you know, just acknowledges, right? Because like, I mean, and this is like, you actually asked this question like, how to Google? How do people like engineers at Google, think of blockchain? Most of them actually think it’s a scam? Right? Like, it’s still like the 2017 kind of where, you know, anybody could like write a white paper, you know, take a template of white paper submitted and raise money kind of still looms over it. I think I get introduced but same as just generally, like blockchain in use or do So people just don’t think about it as much. But like real realistically, people are still not there was like understanding what is the value that they are getting by using these technologies. And we kind of need to like clearly, you know, provided there and gives them like gives him nuggets of value to you know, lose them in in a way to like more interesting applications which are, you know, maybe getting closer to what sounds like that three kind of more dogmatic that city people think look like but I also like as I think realistically, you know, we cannot like destroy the old you know, financial system even it’s even if it’s trying to do it itself right now. You know? Yeah, I mean, the thing is, like, it will stay there and like, I actually I just did a talk on this at the non con. And like, realistically, even though the you know, there’s a lot of like huge global economic crisis is No, I’m going to happen and like happening right now. Like it’s still like tremendously bigger than all of the crypto right now. And like all those, the crypto right now is really, really just BTC to us dt. Trading, like, in and out right. So, so like realistically like before, before we get to the state where where like everything is operated by decentralised or like some some even a hybrid of decentralised finance, like we all need to coexist together and like build bridges between, you know, kind of the centralised and decentralised finance centralised and decentralised application, centralised, centralised identity and like how exactly you do it and how exactly you connect developers to it, like really matters how you know how much adoption will get in the coming years. It’s really interesting that you know, you talk about the some people within Google, who you could regard as the elite of the software engineering world, that is almost a dirty word to be working on anything associated to blockchain because of its because of the Ico phenomenon, and crypto more generally the kind of speculative element. And I think that, you know, if if we look at the thing that’s putting off a lot of developers when I speak to them as I’ve travelled around, it is it is sometimes embarrassing when they asked you to what are you investing in? Especially, I would imagine, in circles, the icpc circles, which even more kind of this elite community, that it’s kind of by association, you don’t really want to say that you’re working in this domain. So it makes total sense that in your missions to kind of onboard these developers that you’re taking much more of a pragmatic approach, or, I guess more of an agnostic approach, right. So as you say, you’re providing a solution. bactrim of technology, you don’t really care where somebody wants to sit on that spectrum, your role is to just kind of provide the toolkit that enables them.
Jamie: Yeah, for sure. I mean, I think like there’s kind of few dimensions in this. There’s, like when we’re talking with developers, right? You can definitely do in developers with technical problems and technical kind of interesting nuggets that blockchain has, right? And the reason why we have such a strong team of you know, of ICBC kind of winners and finalists is because like, there’s a lot of interesting technical problems of building blockchain and building on top of blockchain to be solved. And that’s kind of originally what Lou does in as well, where, you know, when we looked at it, and I’m like, wait, but like, we can do better, right, in a way, like trying to figure out how to solve these problems. But then, kind of flipping this on kind of dissolve the side of this question is really, yeah, I think like, there is a Different, like needs for different engineers, right? Like some people, you know, just coming into, like they have a project they need to do, they just need to pick the kind of technology to use. And, like, they don’t have a goal of, you know, building a, you know, self sufficient application that will live forever. I’d like that’s not everybody, like, it’s not everybody’s job in the world to to build such applications. A lot of people, you know, building e commerce website, like for e commerce website, I mean, there’s clearly somebody like actually serving products, right? It’s not like it’s not gonna be like sell, you know, self running application. So, like autonomous application so so like, you do need kind of a variety of things and a variety of tools for different people. And like, we actually like the way we think of it is kind of a stack and then where at the right you know, obviously at the bottom, at the bottom of the stack you have the blockchain is like kind of like lowest in Like infrastructure, incentive structure, you know, finance fabric that provides like underlying stuff, then you have this like we call them reusable components, right? It is this kind of independent libraries and kind of kind of like micro services, but they’re you know, self self operating and they do like governed by community. But then on top of it, you actually have user facing applications. And realistically, most of user facing applications will be run by some entity or people right. So not like you cannot have like really high quality applications, which are not, you know, overseen by some like product vision. And so they can still be written, they can rooted be rooted in this reusable components. Like example I like is imagine me right now wanting to build Airbnb. Well, what is Airbnb? In reality? It’s kind of a composition of few components. One is a marketplace of the like listing marketplace, right? It’s an insurance and it’s a service and doubles this right? So if I want to build In their baby company, what they want to do is like have the permissionless lifting, which means I actually have the sounds that kind of regulatory arbitrage coming in here. And they also can bootstrap it by you know, integrating some marketplaces and other places I can actually use an existing like reusable component of insurance which you know covers some of the risks that you know, people will destroy the house and then what they just need to do is kind of link it all together provide the best user experience people you know, can have bootstrap just lifting set and you know, provide some kind of connection to the you know, somebody to call to and the service right so like my costs on like actually building this will be way and operating this will be way lower. And I also kind of reduced a lot of like regulatory uncertainty that this may had a rich like Airbnb needed to deal with, for like last, you know, how many years so that you were talking about this reusability and I know you have the ability to provide Pay distribute royalties associated to smart contracts that are used. Do you saw on one hand you have to stack it? Is it fair to say that you’ll then try to build a marketplace around that of these micro services? And that is the first step, then towards there being daps? Because when people think about adoption at the moment, they’re primarily thinking about, you know, billions, millions, billions of users tomorrow using a dap? And is it that you believe that there needs to be something that first onboards all these developers, and then this marketplace of microservices will emerge on top there’s phases to mass adoption of these underlying technologies?
Illia: Yes, at least that’s kind of our assumption that they should be testing. And, I mean, I don’t think anybody has like a clear answer to any of this, as solves this problem of mass adoption, but like our assumption is that you kind of first need to attract developers who are First, you know, we’ll be able to tinker, we’ll be able to build some interesting applications, but also we’ll be able to build something that, you know, makes it a living, or kind of allows you to create companies around. And then from there, if you have this kind of, you know, hundreds and thousands of applications being built, then some of them, you know, will actually capture use cases, which are, you know, as novel or solving something in like, you know, better, faster, cheaper way, and actually bring more users, right. Like, I don’t think like, you know, throwing a dart, like in the dark and saying, like, Oh, this is this is application that will, like attract, you know, millions of users, like, maybe this, you know, like, I mean, people have been doing this, but like we’ve seen that haven’t yet, like, realised into anything. So, so far, like what have worked before for all of the developer platforms of the past is actually this kind of a platform. And it’s like working through developers I like Windows as a developer platform is an example. Obviously, like office being like one of the products that like had huge success. But having like, you know, an open platform and being able for the for like, millions of developers actually make a living out of building applications had a huge kind of impact on people using Windows and offers actually, as well. And same for you know, Apple safer, like platforms like parse Firebase, etc.
Jamie: So what are the qualities or characteristics that a developer that’s not interested in decentralisation per se? Why is it that they come and use near protocol what what is it that they can’t currently get from? Let’s just call it the web to stack.
Illia: So I mean, there’s kind of a few things that like wait are way easier right on on this like more like on the platform, which has let’s say blockchain kind of under Right. First of all, anything that like does deal with money or or like valuable assets is way easier to do. Like if you if you imagine like building right now a obligation that needs to do insurance or you know, borrowing or if you want to do like even just like payroll, like the amount of kind of working needs amount of licencing the amount of kind of cybersecurity you need to enrol is like, tremendous, and then you need to GDPR or like HIPAA for, you know, California now. So like, it’s just like work after work after work, just to kind of launch something that you know, in reality can be like 1000 lines, my contract that captures value and transfers it and then the application itself like the front end and kind of user interface can just operate with this with this data and business assets in non custodial way. Right. So and then like this, actually Actually evolves into more things, which is like, right now, we like if you think of like the requirements for data kind of control, and they are huge. There’s actually a lot of kind of sub use cases where people are becoming more and more cautious about giving data to the big cloud providers. The example I really like is Facebook actually not using a cloud for their email, and their file storage, right, they actually use Microsoft SharePoint, that is a cluster on prem. Because they don’t trust Google, for example, with their email, right. And so like, in reality, if we had a solution, which was like, as a company, you can actually own all your data, but still use software like as a service. Like this is actually a pretty big thing for a lot of a lot of companies because like right now they have a huge maintenance costs and running on prem on prem services for free. All this kind of, like, service and I mean, like resolve is back to Microsoft, but they’re like, you know, on prem, there’s like, outlook is not the best piece of software, like, adds like 10,000 unread emails, that thing is just like hanging. So even web two doesn’t trust web two.
Jamie: Exactly. I mean, it’s like, especially when you start actually like becoming like at, you know, the more higher risk. And that’s where it’s like, you know, when people say, Oh, we should like, move away from Facebook. I’m like, I’m all for it. Actually, I deleted my Facebook recently. But like, this is not where the pain points are the pain points actually, for, you know, data protection and privacy is around either like big companies who don’t, who don’t trust as big companies as their data or around like GDPR protection, where if I’m a small startup, I actually don’t have resources to build something that’s like, fully compliant and like while maintaining all this data on my side, right? So it’s actually easier for me to not manage data and give it away. Back to Users and have them their custody of their data. Versus like, building everything, like super compliant setup, like my friend who who is building company, they spend more than $10 million building a like compliant database, which like actually manages, like, their clients records. So there’s like full compliance. Yeah, I mean, it’s interesting. You know, GDPR, I think is a great catalyst for for web three in that it introduces liability. So rather than just be a security risk, and perhaps a bit embarrassing, now, it is a huge liability that can dramatically impact share price and as you say, for for an SME actually becomes prohibitive to do business.
Illia: Yeah, for sure. Actually, I also know like a friend of mine, mines, sold his company to Twitter, because like when GDPR came out, they just like evaluated and had no like, there’s no option they could actually be compliant. So there’s like, just decided to sell it to somebody who actually can’t like either pay fees or pay engineers to rebuild the stack.
Interesting. And so we were talking earlier, you said, No, the technology is agnostic. It can operate on a spectrum on multiple dimensions. It’s just configurable to, I guess, the trade offs. I mean, you could say it could be the philosophy of the developer, but it might also just be the particular trade offs that they want to make for certain use case. But But you also you also mentioned this idea about kind of leading them into perhaps principles that they might not currently be thinking about. So whilst it is agnostic, it sounds like there is
there is still at least fit for you as an individual. I’m not sure for the wider firm. There is a mission there to try to get a better web somehow. What for you are is non negotiable. That’s it. Certain things that are fundamental that you, you you wouldn’t allow to be configurable.
So that’s an interesting question. Like, my point is like, I don’t think we can have like this radical change right, we do need to be slow and like steady and actually like, give value in every step. I think like for me, like I am looking forward to choose a world where you know, we actually shifting away from this like profit as like a fine like profit and like stock price as a final thing that like, is very removed from like, what people are doing in a company. And like this removal actually creates like, like a Google is like a, you know, huge company. And most people don’t actually understand directly how their work relates to kind of what the company does. And this creates like a lot of misalignment around like what people willing to do and what people are doing to actually Like factor thing, right figuring out how like, like this, you know, community owned in a way, like moving into community owned by our users actually deciding, right? Where the products are evolving, where’s, you know, the software is going? I think that’s where I’m looking forward to that world, like where race, you know, very far from that right now. And this is where like, when you ask him, like, what are we leading developers into is actually like, showcasing how they can start integrating some of these components into their own applications and actually benefiting from them in the short term, right? where, like, the benefit of you know, even like talking economics is the fact that you can kind of bring some of the future value, right down to the original users, right? Because it’s like first early adopters and kind of smeared off and obviously you need you need to like value to come in later. As always, like, getting counts does work, but it’s you know, I just Max’s original community, and then the surgical community can help shape the product and help drive it. But the point is that like long term, in any, like bigger company, the like, this kind of loses this connection gets lost between like the, you know, community and shareholders and users like it’s all kind of falls apart and like has very different connection. And like here, we actually have this connection that is, you know, linked kind of through the crypto. And, like, actually figuring out how we can, you know, manifests that into like better products and products and serve users better. I think that that’s kind of my North Star.
Jamie: So that’s super interesting, in that you referenced the idea of crypto assets or tokens, how they play an important role in the participation of the network in the value that’s created. I know, this is a very polarising topic, whether whether tokens are attitudes to open source systems or whether they’re a distraction. Could you talk through why you think they’re presumably a powerful addition?
Illia: Yeah, for sure. Yeah, yeah, I think I mean, the core point is like, the token needs to capture value in some in some way going forward. Right. I think like most of the tokens are created kind of for the sake of it. And that’s where like a lot of the polarisation comes from the the kind of the, like, my, my view on this is, but like tokens, provider is a way to really do a few things. One is like designing an instrument that can provide the early adopters a way to capture sounds like value that’s gonna come in the future, right? And because of that, they will they’re interested in you know, participating and growing this ecosystem kind of early on, to really to, you know, bring sounds it brings that value in the future. The kind of question really around is like open source and tokens is the fact that like, the fact is, like right now open source is actually supported by large companies. That’s, that’s like a real real situation is like most of the open source is just large companies paying people to work on stuff, because he uses himself. And, and this, you know, started obviously, like red had been like a pioneer in this and then like, I mean, I get I worked on TensorFlow, you know, like, a lot of my work and that being open source, but you know, I was paid by Google. So, the point here is that like, that means all of this open source is really just targets, whatever the kind of goal so just pick companies follow. And like, we cannot obviously like discuss, like, Where’s this goal, you know, good and bad. But it’s definitely not always aligned with actual users of open source projects, right. Like, if Let’s imagine that like TensorFlow specifically, right, like inside Google, when I was leaving, we had like, I don’t know, maybe 1000 teams using it, right? Probably now more. But outside of Google, there’s, you know, 10s of thousands, if not, maybe even hundreds of thousands of developers using tensor flow. When you know, decisions are made, how TensorFlow should be evolving? Do you like to whose voice do you think is heard more? Right? That’s kind of like a simple comparison. Right? But like it even more complex is like really, when you think of, you know, if you have something that’s not just a software, but actually a service, right, it’s something that needs to align kind of multiple parties, like it’s either a marketplace of some resource of some service of some value. This is where, like, the softer part is just, you know, a way to facilitate this. I can really you need to align the incentives of the of the kind of entities participating developer itself, you know, the person who builds this and people who build this also there, right? So you kind of have this, like multiple roles. And maybe the people who, you know, upgrading on top of marketplace will have some financial system between them. But like, it will definitely not slow back to developer. And, like, it’s unclear, like, without some kind of economics, right design, like it doesn’t need to be talking, but it needs to be designed some way to kind of make sure that, you know, everybody, like aligned on, you know, evolved into this ecosystem forward. And actually, like, defy right now has like examples of like misalignment where, you know, some of the default applications are built by people who actually not getting any value out of it, because a lot of these applications are kind of as community services, right. They’re just, they’re not really capturing value themselves. And then people complaining to developers, hey, why not building the feature or something? and developers like Hey, I’m not getting really paid for this right? I built you know, the best thing I could and like, but you’re actually getting value out of it. Right so. So this is where like this misalignments actually coming in even in like that free world is there is no like actual clear design how this should work.
Jamie: So you came into blockchain or you you began looking at the problems, challenges or design challenges within within the blockchain space from a machine learning background. What do you if you look at smart contracts at the moment, you know, they’re fairly limited. I don’t mean specific to Nir just generally, they’re quite limited in terms of what you can actually do in a smart contract. How do you feel about the potential for machine learning and blockchain in in combination and to allow for greater complexity in smart contracts?
Illia: Yeah, so I mean, it’s an interesting question that I’ve been obviously thinking a lot about. I think like, I think blockchain has said this is the first design is about like, kind of incentives and alignment and providing this, like backbone, I think the machine learning will come in is kind of more, quote unquote, like layer two, but like, I actually have some ideas how this should come in. But before before, I think like, right now, it’s still too early, like, the reality of machine learning machine learning works when you do have users. And when I have data, and right now, we don’t have users. So right now, it’s like, too, it’s too It’s too early for like, actually, like, trying to build some of these things where like, where we just don’t have kind of this underlying, you know, infrastructure yet. I think like, there’s some interesting kind of concepts around like, you know, like self evolution and like, maintaining things and like how an applications can like live and evolve on their own on blockchains. But again, I think like, we’re still way too early for this, like, even like good prototype kind of cases for this stuff.
Illia: So I mean, I definitely follow the logic, you know, users data then then machine learning. And how do you how do you stay disciplined, as a founder in the space because I mean, your team are, by all intents and purposes, you know, the, some of the elites of the elites, you’re incredibly bright. I’m sure there could be a number of ways you could distract yourself solving really interesting problems, but that are slightly removed from a roadmap that a user with one right now how do you how do you stay disciplined? Because this is one of the things I when I when I speak to a lot of founders or I see a difference between those that are executing and those that aren’t, is it seems to be those that are most disciplined in sticking to the user and what they need. So some question I think, like generally a few things. I mean, one is kind of trust. to focus on, you know, this, like, what, what is the deliverable so that we can cat to, you know, developers and shoes and to users. And through that is like, it kind of focuses you on things that you know, like it kind of removes right away things that maybe don’t matter right now. Like, I’m actually super excited about stable coins. I think, like, there’s a lot of value in them. But like right now, like, at this specific moment, right off or near, can we deliver something? And the answer is no, because well, we need to launch our thing, right? We need developer tooling, we need this and this. So like, it’s kind of like builds up into this, like more like, everything that’s not currently like fully applicable, like similar as machine learning, actually have some interesting ideas. How did the federal like Byzantine fault tolerant federated learning, but again, like, can we do it? Do we have data? Do we have users? Not yet okay, let’s think about it later. So, so really like trying to kind of focus on like, what can we deliver in like next few months? But also like how do we make sure that people are aware of it? What is the messaging? What is the story that we need to tell? That actually gets people excited and start trying things?
Jamie: So you’ve got your main net coming up. And I know you referenced this the ready layer one event that you’re going to be very actively involved in. What are the what are the some of the big things that are coming up on your on your road?
Illia: Yeah, so regular one is very exciting event because it’s pretty much polka dot Cosmos near protocol, labs tasos coming together to really have like a obviously online event given current state of the world, but kind of bring together developers bring together like workshops bring together, people who can ideate around their scenes, how we can practically get things to the users to the developers in some way. That Actually, it will specifically help elevate some of the current problems that we’re facing. Right? I think like, that’s actually one of the important things is like, we are living through probably like generational thing. This is, you know, this will be you know, remember the history like this time and we, you know, we should be mindful of it and like we shouldn’t you know, like we should review our assumptions at this point and like actually think more how we can help kind of people who will be hurting in you know, next coming months. So, so I’m actually very excited about this man because like, the idea is to bring probably like, smartest people in the industry, and like, actually all together work on like, practical things we can we can help.
Jamie: Anyway, I didn’t get an invite then. Yeah. And then you’ve got your you’ve got main net coming up as well, right.
Jamie: So thanks so much for your time Illia. It’s been great talking to you hearing about your progress. Hopefully I’ll get my invite to ready later on. So I can become part of the elite. And we can, we can at least see each other online if not in person soon.
Illia: for sure. Yeah. I mean, everybody’s invited. Ready, layer one to apply for tickets and Fox.
Jamie: Great. Thank you.
If you enjoyed today’s podcast, please make sure you subscribe, rate and share your feedback to help us reach as many people as possible with important mission of web 3.