 Tubit Verified is a verified computing platform as the name might indicate. So what Verified Computing does is it takes these compute functions that you've built much the same way you would build a serverless function that you might deploy on the edge or you might wanna deploy anywhere and it makes that function trustworthy by verifying the code that's executing, verifying the data that goes into the function and verifying the outputs of that code. Hi, this is your Supreme Pathya and welcome to another episode of TFR Let's Talk. And today we have with us Blaine Sims, head of product at Tubit Blaine. It's great to have you on the show. Hi Swap, it's great to be here. Yeah, it's my pleasure to host you here today and this is the first time I'm talking to someone from the company. So I would love to know a bit about the history, the history of the company, how old is the company, what problem that you folks are trying to solve for the market. Absolutely, Tubit traces its origins back to 2015. We started as a research project. Our founder, Dr. Jason Toich is a researcher in computer science and game theory. So some of the work that he's done has been really important in the world of Web 3 to contribute to the question of how do you make blockchain scale? How do you perform compute in Web 3 that may not happen on a blockchain directly and make that compute secure? When you look at blockchain, how have you have seen the evolution of blockchain and how your industry is leveraging these technologies? You know, blockchain, I think when it, you know, when blockchain started hitting the scene, again around that 2015 timeline, for example, was when the Ethereum blockchain was still very, very new on the scene, really a lot of excitement because it seemed truly as a disruptive technology. The reason that blockchains were created was to facilitate transactions between parties that had absolutely no basis to know each other or to trust each other. And while that in many contexts is equated with cryptocurrency and indeed a lot of blockchain technology originates in the world of cryptocurrency, I think what's really caught fire since then is this notion that there are all sorts of different types of interactions we have, transactions we have, whether as individuals or as businesses, where our assumption of trust is really one that gets harder and harder to understand, you know, that basis for trust as we, you know, as our world becomes increasingly digital, our transactions become, you know, natively digital, how do we actually, you know, keep good records, which is where blockchains come in, but also, you know, create very highly automated and trustworthy, you know, compute-based processes around those transactions, which is where Trubit comes in. Can you also talk about, of course, in this space, new technologies, new paradigm shift, new jargons keep coming up, we talk about JetGPT a lot of these days, Genetic AI, earlier we were talking a lot about Web3 and there are a lot of misconceptions about Web3 as well. Talk about what are some of the emerging technologies that are of your interest for your industry, for your space? The blockchain industry itself, right, or the Web3 industry more broadly, which includes blockchains, but is broader than blockchains, obviously has been a hot topic in itself, but I think what we're really interested in is now the convergence between this concept of decentralized trust and new technologies like Genetic AI. Where we see this convergence taking place is primarily, you know, think of the world of the enterprise, right, that's going to increasingly be, you know, building critical core business processes on top of AI technology that they have no control over. There's really only a handful of companies that have the compute resources, that can even procure those resources to begin with, that have the models trained, that have the expertise to build and run these models. So that kind of relatively small group of AI providers needs to be trusted by a huge group of enterprises and what's the ability to trust? How do you know that the data that's being fed into a model that might be specific to your business? How do you know that that data was actually used in the model? How do you know that the requests you're making of that model are being processed accurately and the results you're being provided back, that you have proof, you know, a chain of custody, if you will, for that entire transaction, particularly if you're in a regulated industry. So this emergence of generative AI, which is, you know, again gonna be outsourced by most organizations to other partners, has a lot of different components, if you think of it as a supply chain in its own line, a lot of different components of data being supplied and computing supplied, that intersects well with what we've been working on in the Web3 world, which is how to build trust amongst different parties that really don't otherwise have a basis to trust each other. When we look at, once again, modern infrastructure environments, workloads, I mean, depends on what somebody is doing, of course, we talk a lot about cloud. We talk about a lot of, you know, technologies that are underneath, but we don't talk much about the computer itself, so talk a bit about what kind of technology you're seeing when it comes to kind of, once again, emergence, evolution of the whole computer and we can also look at from the Web3 perspective. I think in, you know, in cloud-based compute, what we're seeing really top of mind with the developers that we talk to are a couple of, you know, really, I think key trends that, you know, depending on where you say it, you might, you know, be skeptical of them or you might really, really be embracing. The first trend in cloud-based compute is serverless compute. Serverless compute, right, is this notion that you can reduce your code down, abstract it down to kind of its most atomic function level and then deploy that code as a function on someone else's server. Typically, you know, these are servers within the large cloud computing companies that we all know, but then there's an extension of serverless compute that's equally interesting that's going on, which I think some might say is, you know, is it an evolution beyond cloud-based compute and that's edge-based compute, which takes this notion of serverless compute and extends it even further and says, well, maybe you don't need these function codes to run inside, you know, an Amazon or Microsoft cloud. Maybe this function code can run on edge servers at the very, very closest point of contact with, you know, with the consumer that's using those compute services. These two trends, which I think are interrelated, really tee up the, you know, the compute world for what we think is an evolution completely beyond cloud compute, which is an evolution to decentralized compute. So once you've made your code compact enough and small enough that it can run anywhere on any one's server, and once you've deployed that code in many, many different places so that it runs as close as possible to, you know, to the client that's gonna consume the output of that code, the next question is how can I get that code not to just run on, you know, Amazon or Microsoft or maybe, you know, CloudFront or Faptly's Edge, how can I truly get that code to run anywhere? And then once I wanna do that, once I wanna say there literally could be, you know, not just tens of thousands of edge points of presence but millions of them, that's when we get to this question of how do you secure that? How do you make that code trustworthy? And we think from the security and compute side alone, that evolution is what takes us into this questions of decentralized compute that Web3 is solving. And then we talked a minute ago about the business evolution that's taking place about needing to really more and more trust someone else to write key code that's running your business. Those two things we see kind of merging together to be the perfect storm for Web3. When we do look at all the adoption of Web3 technologies, what kind of major challenges that developers face and how you folks kind of help them tackle with some of those challenges where we can also talk about the new announcement that you folks are coming out with. So it's a great time to talk about Trubit Verify. So Trubit Verify is a brand new platform that we're announcing now. Trubit Verify is a verified computing platform as the name might indicate. So what verified computing does is it takes these compute functions that you've built much the same way you would build a serverless function that you might deploy on the edge or you might want to deploy anywhere. And it makes that function trustworthy by verifying the code that's executing, verifying the data that goes into the function and verifying the outputs of that code. So what we do in Trubit Verify is have a platform that manages that verification process, manages the mechanics of here's some code that's served up anywhere really on the internet, if you will, but here's verification that's going on. We're gonna have multiple compute nodes run the code to see if they agree. If they don't agree underneath the hood, we do low level verification. This is Jason Twitch's big invention, right? Something called the verification game, which is using game theory to incentivize the people that are running the code to behave properly, but when they don't to be able to point the finger and say, okay, I've gone down to the machine code level and I can prove step by step who's right and who's wrong when this code is executing. And then I can come back if someone has a disagreeing result and say, okay, I know this node was right, this node was wrong. And as a result of all that, I can bundle up all of those results, provide you back with the correct answer as the person that's running that code and also do something really important, which is give you proof. I give you a certificate that says your code was run correctly. These were the inputs. These were the outputs. And that certificate is what allows you then to convey trust to whoever is dependent on your code and your data. Can you also talk a bit about the benefits of kind of decentralization for developer? Of course, when we look at decentralization, we can look it from a different perspective, decentralize workload, decentralize environments, decentralize architecture. What kind of decentralization you are looking at and what kind of benefit you see there are for developers? What we find is the number one reason developers move into Web 3 and move towards decentralization is they're looking for transparency. Again, this transparency might be because they are running transactions or have a business that in of itself has no underlying basis for trust. So they need to be very, very transparent about the transactions that took place for that business and about any of the code, which is where Trubit Verify comes in, any of the code that contributes to those transactions. The API calls that you might need to make to an upstream service or to move data or move a transaction downstream. The movement of data from one place to another, including across blockchains. And then those backend algorithms and AI models that really, really impact how a transaction takes place. The transparency around that is what brings folks to Web 3. What they find when they get to Web 3, though, are some really important other things. Web 3, decentralization as an entire concept, is extraordinarily resilient to failure. There's literally not a single point of failure in a widely deployed blockchain or verified compute network such as Trubit's building because there are generally tens of thousands, if not more, independent parties running that network. And a lot of different ways to get around even network routing issues, things that can even plague the largest in most successful cloud companies. And then the final point is efficiency. Again, this is one of the drivers perhaps for serverless, but it gets extended even further in the Web 3 world. Why pay to have your code hosted all the time, particularly if it's something that it was only occasionally being used when it can be ready and waiting and just called on demand? So the efficiency aspect brings developers to Web 3 as well. And then obviously, Web 3 as its own phenomenon has some interesting businesses that have emerged there, like the Metaverse, et cetera, that have a lot of people really interested in that as well on the consumer side. So one of the things, as we were building Trubit, verifying that we did was really spend a lot of time talking to not just folks that are interested in blockchains, but folks that are actively building Web 3 applications. And we defined that as applications that leverage, either directly leverage a blockchain ledger or leverage decentralized compute as part of their design. And what we found are that there actually are all sorts of really hard things that developers have to do in order to get an application up and running. And they have to make kind of a lot of compromises and take some shortcuts to do that. And so that's what we really wanted to focus on when we were building Trubit Verify is how to bring transparency into these areas starting within Web 3 alone, where developers weren't able to actually make their application as transparent as they wanted. Maybe they are using a blockchain and the code that they're writing is too complicated or too expensive to run on a blockchain. Trubit Verify allows them to run that code transparently and then have that certified proof that they can then attach to the blockchain. Talked about API calls before. The other thing that we really found is the way that Web 3 developers are told that they should be securing the code is by writing these really complicated things called smart contracts, Trubit Verify allows you to write your code in JavaScript or in Rust and these languages are most likely used to working with on your own and tries to help you simplify this process of getting transparency into your code. And that's where I want to head next, which is just from Trubit's perspective, what kind of use cases are there? Of course you can or cannot name some of the companies that are leveraging your technologies. If not, then just give us a glimpse of the use cases that you folks kind of cater to. Yeah, absolutely. So Trubit Verify, announcing this product, our first live use case will be coming online towards the end of this year, which is a supply chain based use case. So there's a supply chain consortium in Europe that's built around the luxury goods industry. This industry is one of the first industries to actually adopt blockchain technologies. You mentioned hyper ledger as a technology that's been around for a while on the enterprise side of blockchain. It's one of the ledgers that's being used in this project, but there are other ledgers as well. And so what this project needed was a way to move data securely from one ledger to another. And Trubit Verify provides that. We provide an auditable kind of certificate that allows any of the participants in this consortium to see data that may have originated on hyper ledger move to any of the other blockchains that are part of the consortium project. So that's one use case. We've got other kind of earlier stage projects in the entertainment industry where we're working with a group of major recording artists who aren't trying to completely change the way the industry operates at this point, but they are trying to make the royalties that they distribute for their songs much more transparent so that all of the parties that they work with that are part of those royalties can see exactly what money came in and how it was split up and have an auditable record of getting paid. That project's working with us to do all of those calculations to show the calculations, have that proof of compute to show the source of the data to prove that the data wasn't changed when they receive it from API calls that they make and then to show proof of payment that when they finish the calculations and made calls to their payment APIs, those calls are also recorded with a transcript and all of that is available for inspection. I want to hear from you a general high-level overview of market industry, not as specific to your industry, where the technical trends are heading and then you see, hey, you know what? These are some of the things that creates a lot of opportunity for us at the same time. These are the challenging areas that we are looking at at the same. We talked about serverless and edge compute before, obviously Web 3 as well. This, I think at the moment, I think a lot of us are fixated on the changes that are coming with the shift to generative AI. I mean, the top to bottom, if you haven't contemplated how generative AI might change your job, you've probably been busy, very, very busy doing something else because it's very top of mind. So in the world of generative AI, the challenge is when you set aside maybe surface concerns that the code will write itself and none of us developers will be a code ache, and I think nobody's truly hopefully worried about that. What we do see underneath that hood though is this ability to scale up, the ability to do so much more with the help of AI that it puts a lot of pressure on our supporting processes, on our processes like security. A lot of organizations spend an inordinate amount of time, rightfully so, vetting any piece of code that they're gonna bring in or any library that they're gonna bring in or any partnership that they're going to establish around the compute that's going on with the organization for security reasons. You think under the hood, these security checklists and compliance procedures, really in many cases, yes, they're somewhat automated around the edges, but they're very, very manual processes. How does those processes scale? How do you know whether it's AI that's written your code or it's written somebody else's code or a model that you're gonna be calling into, how do you know that that's secure? And how do you do that when that code's being generated that maybe a thousand times the pace is what we're seeing right now as some of the AI really hits stride? The second challenge within AI then is the data. There's well-documented cases of garbage in, garbage out in terms of the data that's fed into models, just being regurgitated back and being very inaccurate. Maybe in some of the use cases we see for chat GPT as consumers where we're just having chat GPT write a letter for us or something along those lines, it's not a big deal, because we can see those inaccuracies, but as you see AI being really used in industries like medical science where AI is doing diagnoses or where AI is being used to make critical financial decisions, that question of is the data good is another area where trust becomes very, very important. Do you see where the data came from? Do you see how it was applied, et cetera? So there's just a couple of things kind of relative to that trend that we're thinking about. Lane, thank you so much for taking time out today and share these great insight, the announcement of the new product now. So this whole market where it's heading, what challenges there are for developers and how you folks are helping solve them. Thanks for all those insights and I would love to chat with you again soon. Thank you. Thank you so much Swap. Really been a great conversation. Thank you.