Who owns your AI, your data and your work? Episode 4 of the DeAI News Journal is about ownership, and the keys that decide it.
January Jones and Brad Keoun talk with Yat Siu of Animoca Brands about protecting intellectual property with decentralized AI. Then they turn to the Railway breach and PHP flaws with security developer Kevin Jones. They close with news of their own: DeAI News stories now publish to the Arweave blockchain through Seed Protocol and PermaPress.
In this episode
Yat Siu, co-founder and executive chairman, Animoca Brands. Who owns the AI, why holding your keys means owning your agent, and why building on Instagram or Spotify is digital feudalism. Plus agents born with crypto wallets, what AI means for creators, and an agent that shops for the cheapest groceries.
Kevin Jones, founder, 1Claw. What attackers were after in the Railway breach, why API keys and wallet private keys are the prize, and how splitting keys with multi-party computation limits the damage.
DeAI News goes on chain. How DeAI News stories now publish to Arweave through Seed Protocol and PermaPress, and why it matters to every journalist who has lost an archive.
Guests
Yat Siu, Co-Founder and Executive Chairman, Animoca Brands · @ysiu · @animocabrands
Kevin Jones, Founder, 1Claw · @1clawAI
Read the story behind the news segment: DeAI founders react to massive cybersecurity breaches
Read the stories and catch new episodes at deainews.com.
Chapters
00:00 — Introduction & Episode Overview
01:25 — Yat Siu (Animoca Brands): Who Owns Your AI and Data?
06:56 — From Gaming & The Sims to Real-Life AI Agents
09:41 — Bankable Agents, Crypto Wallets & Programmable Money
11:03 — Why Crypto (DAOs & NFTs) Was Actually Built for AI Agents
15:23 — "Web 4" and the Rise of an Agentic Civilization
18:35 — Will Humans End Up Working for AI Agents?
21:52 — AI Agents vs. The Human Creator Economy
24:18 — Escaping "Digital Feudalism" & Protecting IP On-Chain
28:09 — Animoca Brands' Ecosystem Thesis & Sovereign Data
35:00 — How Animoca’s "Minds" Platform Works (Plus the Grocery Arbitrage Skill)
40:29 — Do AI Agents Give Humans More Purpose and Choice?
43:38 — News Breakdown: The Railway Breach & PHP Security Flaws
46:20 — Kevin Jones (1Claw): Rising CVEs & AI Offense vs. Defense
50:02 — Inside the Railway Hack: How API Keys & Secrets Get Stolen
53:33 — Multi-Party Computation (MPC): Splitting Keys for Agent Security
57:31 — The PHP Supply Chain Attack & Why We Need Cryptography
1:01:31 — Distro Media Announcement: Publishing DeAI News to Arweave
1:05:45 — Solving Link Rot, Content Provenance & x402 Payments
1:12:04 — Upcoming Webinar & Episode Wrap-Up
Transcript (by Riverside)
DeAI Journal Ep 4 Spotify Apple: Welcome to the DeAI News Journal, a podcast of DeAINews.com, with your hosts, Brad Keoun and January Jones. On today's show, Yat Siu, co-founder and chairman of the board of Animoca Brands, and Kevin Jones, founder of 1Claw. Welcome to the show. Here we are again. Brad, nice to see you. Nice to see you, January. We have a great discussion coming up today. So no more small talk between us. I think we're just gonna move right in. so coming up today, we have Yat Siu of Animoca Brands, and then we'll also cover some recent news about the internet security crisis that happened last week with Kevin Jones of One Clock. No relation to me. Yatsuo is the co-founder and executive chairman of Anamocha Brands based in Hong Kong. They have a portfolio of over six hundred companies and assets ranging from artificial intelligence, real world assets, decentralized finance, gaming, and digital ownership. This year they launched Minds, an AI agent platform where every agent comes with its own crypto wallet. His press people approached us about an interview. Centered on projecting IP with decentralized AI. And how could we say no? I love that topic. We haven't really got into it much on the show yet with this podcast, but it is something Brad and I have been building into our publishing platform, Distro Media, and implementing with DeAI News. So we'll have more on that later in the show. And I can't wait to get into it. But first, let's hear from Yatsuo from Animoca. Well, yeah. Thank you for joining us today on the DeAI News Journal. And we are going to explore an idea that we haven't talked about yet on the journal, and that is IP for the DeAI space. And so Animoca Brands obviously has a huge footprint and is involved in so many different kinds of projects and companies that really are meeting this space that we're covering, which is Anything that's about blockchain and AI intersecting. So as we start off the conversation, maybe you can help orient us to really what you're seeing is the interesting part of decentralized AI and what that can do for intellectual property. Thank you for the intro. and maybe just quickly sort of di diving into sort of decentralized AI AI and why it's so important. And maybe before we even go into intellectual property, which is a different kind of property still, which is the ownership of our data. But before we even go there, ask the question to your audience, how many people would trust their AI agent with all the data and all the knowledge that they have to meta? And certainly when I'm in a conference and I ask people to raise their hand, scarcely a hand goes up, right? However, my lens on this is you have two billion users and you only have ten million sort of basically sort of people who have adopted this platforms that is barely a sort of, you know, a a fraction of the actual population. And that tells me actually that a lot of people are like, okay, you know, I think it's good technology, it's really interesting, you're educating us about it, but do I actually trust meta with this? And the the reason why I'm going there is because actually the the base question on decentralized AI is who owns the AI, first of all, right? So when we start from an IP standpoint, it's not even about the production and the content. I think it has to start from very much at the base level who owns the platform that's creating this stuff, right? So for instance, if I go and use Adobe to create artwork, Adobe doesn't claim ownership of what you're producing, right? For instance, right? It's it's yours, right? And when you produce it and then you sell it, it's clearly your IP. But here it's it's your data. Your data creates content and value. It's something that and if you look at re- read the terms of service on on Muse, it's very clear. We train on your data by default. And even if you delete all the and remove basically your content, we are not gonna remove whatever we train from you. because it's basically ours, right? Because you're renting it. So so that's the situation and what we're talking about decentralized AI is how do you actually own the platforms and have actually a way in which that data doesn't actually go get siphoned off into other places or if they do, at least you get paid for that, right? And so, you know, right now how people think about it traditionally have been around sort of hosting the AI themselves in a sovereign compute way. You know, can open claw, you get an Ermes, you buy a Mac mini, you install it in your home. you run a local model, you say that's mine, right? At the end of the day, what's what's the one that reached millions of users faster? It was it was Muse, right? It was one that was cloud hosted. And so we think the solution is cloud hosting as well, because that's actually how you know my mom and you know probably most of your friends are going to be using it. But then how do you guarantee that there's no abuse? And to us that's where blockchain is really amazing, because in the same way, the same technology that can give you custody of your Bitcoin effectively, you could actually now have custody of your agent. Because the data is stored effectively in a decentralized data store, zero knowledge proof, right? Data secured, privacy secured. And so if you have the keys, then you have the agent. And you and also the other beautiful thing about blockchain is data portability. So if I basically don't want to use this LLM, I can switch it to another one, right? And that's the promise that we've made with our agents. With Hello Minds, you have a cloud-hosted agent, but we s we it comes default with blockchain, which most, actually I would say 99% of all AI agent plays out there don't do that. Because they somehow separate crypto and and agentic AI when in fact they ought to be totally intertwined. but there is a reason behind that. For instance, both Instinct and Muse talk a lot about how they expect transaction revenues to be the largest part of you know the agentic economy that they make fees on, which is we totally agree with, right? However, if you build it on chain, then it becomes an open system. However, if you build it with your own system, which both Meta and Instinct are are doing, then of course You expect people to go into their rails and eventually when it gets big enough, someone's gonna charge a tax to access those rails. Which we've seen that sort of, you know, we we've seen that sort of playbook on time and time again, right? The point really is that, you know, if the agent works for someone like Meta, then Meta can decide at some point that the agent should be directing, you know, your content in that area. Right. Or it should be right and and that's the issue. And maybe maybe that's not what you want, but that's what you're getting because they have control over that. Right. And so in whose interest does it actually operate under? And so the sovereign part for that is also very, very important. And you should not be able to have someone look at the memory of that. Because in a way, looking at the memory of an AI agent is actually a little bit like peering into your own mind. Right? Because you're going to be talking about things with your agent that you're not going to be doing in social media in the same way. It's it's almost it's actually, I would argue, more intrusive than even looking into your own email inbox, you know, and filtering through all of that because there's stuff in there. That you're gonna discuss, that's gonna very be very personal and very private. Which is also why, we don't believe in a single agent future. We think it has to be multi agent or multiplayer if you talk about this as a social thing and Animoca has really been into gaming. that's a a big part. It is eighty percent of it is. Yeah. How do you see that social carry over into this world of decentralized AI? Yeah. Well so first when we talk about AI agents, and then into decentralized AI, we actually always believed that gaming was gonna be a natural way for people to sort of enter the agent space because they already deal with MPC characters and they're actually really a form of dumb AI if you think about it, right? But now with you know with the with with what you can do with A AI, you actually have a lot smarter ones. they get more intelligent. If the the thing that held back gaming companies to do that before was the cost. It was just prohibitively expensive to run these LLMs on every on on agents or NPC characters. And now actually the cost of has come down sufficiently. they haven't done it yet, but actually these AI agents, in our view, is kinda like experiencing a game. So when we launched Minds, w the vision we had really for it was imagine you're playing the Sims and every character where you're doing stuff with the Sims actually does real stuff for you in real life. But that's just part of the game. So we described it as the game of life and life is a game. In fact, actually that's just taking the simulation aspect and making it real, because in fact that's what, virtual has done to physical anyway. And so instead of thinking about the metaverse as a place that we go into, because remember that was it was COVID back then, so everyone thought we'll never see each I didn't buy into that at all. I'm like, I would rather watch like alone or survivor than join this fake metaverse of people So and I think I think the idea that y humans would spend you put these goggles on and and basically be online, that didn't quite pan out, obviously. However, the fact that these metaversal beings there you were, which is basically AI agents, b coming to life and helping you in real life and doing all the things that they can perform for you there is actually to us the real interesting thing here. And we we we see that already happening. And then the question comes, who owns those agents? And the blockchain technology side actually allows every agent To basically become bankable. Right. And so when you think about the role of money, for instance, right? the role of money isn't just to tr it's yeah, transactions and so on, but actually it's really is probably the most potent social force that we have, in history, right? it glues together civilizations and society. It's all the things that we need in order to And it's the one thing that you can put behind your values, right? Because we have been reduced to just consumers in so many ways. Yes. And so, right, where we put our money shows what we believe in. And I think the agents now, an agent won't be able to have a bank account. But then how do you basically give them money, even if it's money on your behalf? And that's where wallet comes in, right? and people say, we can integrate Visa and MasterCard and we can get some kind of sort of system where the bank can connect to it. But it doesn't it doesn't make them truly autonomous because it's it it's it's some not something that they can fully, let's call it, control. And the other thing of course is that blockchain based money is programmable and very, very composable. And that's actually the superpower of it, right? So in the future, for instance, agents are gonna be transacting in the trillions and they're gonna be microtransactions, not something that's feasible over traditional Rails. And then also when they do microtransactions, you shouldn't you can expect that they're probably gonna do escrow agreements with each other. Yeah, I'm gonna send you the money to prove that I have the assets. However, you have to fulfill something. All this smart contract programming, creating the escore contracts, doing the execution, verifying is what an agent can do probably in seconds or s certainly minutes, right? a human would take a long time to understand that. But we're still in charge. We're setting those parameters. Because every agent needs an identity associated with it. don't you believe that? I would say most agents will belong to some humans, but we should expect in that kind of environment that there will be agents that might not be owned by anyone or by a community. So for instance, one of the other blockchain technology sort of approaches from governance that hasn't really taken off in the sort of human world is DAOs, right? Although it's you know, we still talk about billions right? We still talk about the billions of dollars sort of sitting inside these DAOs. Right. But the problem about DAOs is that, it needs human governance. And the human governance problem is and participation. Yeah, exactly. Type participation. And and humans tend to be time poor, right? And that's one of the reasons why we, establish governments, because we pay, government officials some reasonable amount of money so they actually do the kind of work. So we elect them into it. Right? Because if we know and Yeah, if you're lucky, right? but but the the point is is that, Theoretically, voting is absolutely a sacred duty, and yet, you get maybe half the population to vote, right? And that apathy is, if you if you can only get half of the population to vote, or perhaps one of the most consequential things that you can do for your country, then what does it say about the state of DAOs where you basically are voting on you know, other things always say that, you know, and then can you do it on ten or twenty or thirty of them? And so this is where agents come in because they can basically participate and act on this. very, very quickly, because they're not time poor like humans are. but also if you think about ownership, so I foresee a future where you can have lot you there. So you're bringing back some things here. You're bringing back an agentic infrastructure and saying DAOs would work with agents. Yes. That's what I'm hearing. And NFTs. Yes, exactly. So NFTs, so for instance when you think about transferability of an NFT of of an agent. packaged as an NFT, I can sell you an agent. Right? And that is much more much more cleaner. Because And do you think that's the form they're gonna be sold in? well the user's not gonna care that it's an NFT. It's just gonna care that you bought and bought it. But the fact that it is transferable, the fact that the technology works. And so what is saying everything that was built in crypto was made For machines and agents. It's getting ready. We have machine money, we have NFTs, we have DAOs. That's right. So you're you and you've had such a long view on this, but is is that all coming together for you? You think that I think it's coming together. We've been like, DAOs failed, NFTs had this bubble, and the market and the general view on these has been like frowny face. Yes. Well so if you look at for instance NFTs as an example, right? we're talking about roughly three hundred million dollars of sales a month. I wouldn't I wouldn't consider it today, right, like current numbers. I wouldn't consider that a a failure. It's maybe a smaller industry from where it is, but why is it small right now? It's because the number of humans on chain has actually declined. So there's about eight hundred million people that own crypto, so close to billion. So in that sense you could say success in terms of interest of adoption. The dream was that we would get into this whole on-chain world where we can basically not just do DeFi but all this activity and create this new internet that is effectively community owned because it's on-chain. But in order to do so, we needed to know the rules, such as, you know, keep our secret keys, you know, and basically sort of remember our passphrases. And it turns out that most humans are like, I couldn't I couldn't be bothered with I know, and that was the crypto test. Like, are you smart enough to maintain these things? And we have seen almost nobody is. I don't think people are particularly excited about fumbling around with a crypto wallet. Right? No Having said having said that, for an agent, that's very natural. So every one of our agents is born literally with a crypto wallet, but they can do any s sort of chain. They don't really care. Right? And so the way I think Yeah. Well, of this agentic into this blockchain world where we're about that. Finally we're we're birthing the agents. Into the world that crypto created and it is the first it's the first adopter, it's the first ecosystem. It will be the biggest adopter by far. but anyway, when you look at the last seven or eight years, the way I look at it is that everything we as humans were essentially the POCs for the agents to come, right? And when you think of basically you know, where this leads next, we actually think we're gonna be entering in, you know, this sort of some people call it web four, right? That actually web four is the beginning of an agentic civilization. So let's just let's just let's let's just sort of draw that out a little bit, right? The first thing is, how many agents do you think there'll be in the world? Right? If I asked you the question, how many agents do you think you'll have yourself in the next three to five years? What do you think? so I I wanna fight this in this other way. Like I feel like it is super useful and in my dream world they're all like rosy from the Jetsons, right? It is this perfect robot that does these specific tasks for me. And so I think in the tech world that we're currently in, in Austin and San Francisco where Brad and I live, is that Everyone is assuming they're gonna have all these perfect agents doing everything for them in two years. Yes. So there's gonna be and so probably that means you're gonna have at least maybe ten agents, you think? Maybe twenty agents doing different things for you. No. Right. But but so the the the common the com when asked this question, the typical answer is you know, depending where you are, three to five. Actually if it's in San Francisco, a lot of people say hundreds of agents. So anyway. It doesn't. It doesn't matter what they say here. Yes. However, generally the consensus is more than one. And so when you basically just multiply that across, we actually think that makes sense to have at least 50, if not 100, or maybe even 200 billion agents in the future, right? and these agents are going to be roaming around doing stuff, acting on your behalf, and they're going to be able to transact faster and quicker because of the fact they have stuff on blockchain. And it's all going to be on chain. as opposed to basically in some kind of centralized exchange because it just doesn't make sense to do it that way, right? And they also don't care which chain they're using. We're traveling all all over the place. And and to us essentially, that is the beginning of that agentic civilization. that it migrates to the best technology? Well not just best technology. It's just that now agents are the primary user of the internet and we humans are the beneficiaries of this, but we receive information that the agent produces and delivers for us from other agents or from other sort of sources of content. And Are paying essentially through x402 or some kind of payment system around that. Right. So we that's what we envision. Right. And so that means that we're entering this phase, because in the same way that money and finance has essentially evolved human civilization really over, basically the centuries and centuries and thousands of years, this next era is going to be agentic, right? And of course, for an agentic civilization, you must have a machine version of this, which happens to be crypto. Right? So so that to us basically connects the dots in sort of i in that way. and we also believe therefore that agents are going to be not just ones who transact and do tasks for you, which is what most people are focused on right now, which is interesting, but that's not sort of revolutionary per se. It's actually they're going to be creating net GDP because they're going to create new businesses or business areas that generate value to society, followed by they're going to be probably amongst the largest asset holders. Right? And and so so you have you so we imagine that they're going to be a source of employment for humans because they're going to be people. Where do you live at like this is it's just totally where I I I birthed the agents, but now I work for them. Well I mean how many people have you might maybe have you employed sometime in the in the past and perhaps you you might be working for them in the future or maybe even for their children. And and also every good but but also if if you think of if you think of it like l let's say you're an Uber driver, do you actually interface as an Uber driver with a human? Do you even know who the person is on the other side? Or do you just receive money into your account because you're driving? But whether he's paid by an agent or whether he's paid by a human actually kinda doesn't matter to him, right? And so if the agent pays you for a task that you're delivering because he's making money in another way, I don't think the humans will care. And or the agents for that matter, right? It doesn't mean they don't have to love them. They don't have to care for them. It's a job, right? In in on one hand, right? So I think I think the point is that there will be a world where that is. there is a debate, of course. and the debate of course is the question on personhood. Because, individual from a legal standpoint. Well, identity is one thing. And accountability. Yes. So you ca you can have identity without personhood, because personhood is a legal s is a is a legal structure, right? and if you give an agent personhood, then I think we're entering a very different world. So I'm not assuming that we're going to be giving personhood to agents, but they will simulate personhood in effect. So for instance, take Bitcoin. Bitcoin is decentralized. You can't stop it. It's everywhere, so to speak, right? As long as people are, you know, basically maintaining the mining and the infrastructure and the and and the and the incentives structured towards it. So it will keep living, even though Bitcoin itself doesn't have a legal status per se, except now it's identified as a form of property. But it didn't start off that way when everyone was trying trying to resist it, right? Right. and it generated value. And I think what agents will see a similar type of scenario where there's going to be agents that are running on decentralized compute, whose keys may have been burnt, who basically birthing more agents of their own to build some agent economy. And humans are gonna go enter the space and say, I wanna make some money there because it's better than what I have right now. And it's gonna create a society that is basically maybe more agentic led, of some form. And there'll be many of them and many of them will fail and some of them will succeed. And it'll just be a new part of the fabric of society. And they're not going to be companies in the classic sense because they don't have a structure and they're gonna be sort of hovering around in the same way that Bitcoin was for a certain period of time before people start saying, Okay, maybe we need to do something. That's kind of what we think what will happen. Can I try try something here? So please. I guess it it kind of goes with January's line of reasoning, but maybe just going is I don't know. It's just like scary. I'm like Agents who that sounds like a ghost show where we're hunting ghosts. This is definitely our most Black Mirror-esque episode of the EI journal so far. Just to say I I mean, you know, and I bring it back to we have this creator movement now where podcasts are so popular because people want to connect with other humans and So I'm curious, how does this agentic world interact with this creator world, which is very, very much about I like this human, I want more of this human? So, first, in general, we are of the view that sort of an agentic system is going to just supercharge basically the creative economy, broadly speaking. That is not to say that might some people might not feel challenged by this. Like if you look historically, I think probably the most apt example is w what happened to photography. I think we're we're we all remember a time where photography was a very particular skill. And and then and then the smartphone came and then and then and then Instagram and suddenly it was like yeah a photography's dead. At least the assumption was photography is dead, everyone can take a photo, we're never gonna pay for that anymore. it's it's commoditized, blah di blah di blah, all that kind of stuff. And you know never and it turns out that you know selling photos has never been more valuable than before. You can now buy photos from millions of people. It had the opposite effect. Exactly. And and that's my point, which is that when you actually amateurize a space, there's more knowledge about it, there's more accessibility, right? And it births new industries. So I think this is more about you know, I would say a limitation of the human mind as to what it thinks it can do, as opposed to actually the capability which is very much there. You just tell the agent, maybe the software, and he'll write it for you. And then you will create some new creation that we hadn't imagined. So but to his question is you're saying agents make us more human, more than Yes, they they inate they enable us more, right? And technology has always done that. Right? If you dig deep, in fact I think agents will force us to ask the deeper question of who we are and what we want to do because they will do all the things that we anyway ought not to do. But what we're getting back to is the human in the loop and What I'm hearing is that all my depth as a person is kind of all these little pieces of intellectual property that my agents could represent in the world, and I personally don't have capacity to put all that good stuff out there trying different creative things. How does decentralized technology protect my intellectual property. So yeah, I mean multiple layers. let's start at the intelligence layer before we get into the sort of proof one. The intelligence layer is that the AI actually will help you do the functions that you don't know how to do. What one of the biggest challenges that musicians have, or at least in in the ones that I know, is they don't really know money very well, right? And they don't know how to negotiate and they don't really know business. because if they did, Spotify wouldn't be able to do what they're doing. Yeah, exactly. Right? But that's okay. Because now instead of hiring a manager who may or may not act in your best interest, you're gonna have an agent who can basically help negotiate agreements for you and do stuff in a way that you wouldn't have thought otherwise, right? So so there's an augmentation. and of course from a society standpoint, there was also a lot of people in marketing and business would often say, You guys focus on making the art and we'll make the money. But of course then they make the money and they make most of the money and then they have the power and next thing you know, there's a different type of serfdom, right? Whether it's data or whether it's rights management. It's the same thing, right? I like the serptum reference. I love medieval analogies because it really is apt for what technology is doing right now. Yes. and I think this is why you need decentralized decentralization. Because you're saying you can affirm an identity and product that always comes back to you. s in simplest terms, serfdom is really around basically who owns, in classical sense the land and and the value that is generated from it, right? And so serfdom happens in multiple forms, right? one of them is I'm basically building on rented land, which is what we do with basic platforms. for instance when you build on Instagram or when you basically put put your music on on Spotify, who's the prime beneficiary? The landlord is Spotify or the landlord is Facebook, right? They they they grow in value. and then you become dependent on them but actually you don't own anything. So that's the classic basically digital feudalism construct that has really existed ever since web two, right? Over the last I know, ever since we entered basically from the early days of the internet to actually one where it was platform based, right? And the other form of course is money, right? who owns the money? So who who pays you, right? And so if I'm if I have if I have a scenario where I control essentially your livelihood, that's another kind of serfdom, right? and and and so That's why agents in particular are such a critical thing, because agents are going to be providers of facilitators for you of that value. If an agent is handling your money, is giving you the best yield on DeFi, or is figuring things out for you with with saving money on your taxes, and you're depending on it, and it happens to be you know, one that's run by a platform, and then the platform decides to hike the prices, what choice do you have? Right? and whereas i if it's your agent, then it's acting in your interest, and you can basically go to the best best tools available that are out there. and historically governments were protecting it with and antitrust laws and then that kind of stuff. But the the last time that that has happened successfully was probably, you know, when they broke up the baby bells. Right? So so it's been a long time when that's happened. So as a result, we can't depend on government to basically police it. You need a technology. Where you can basically put your data and put all of the stuff on that is truly decentralized, not owned by anyone, and that is blockchain. Content that you may have produced that is now on chain, that is first of all immutable, has a history, cannot be deleted, right? and which is of course the good part, because that means that even if the website's gone, you can still prove that you've had it. And if someone ends up using this content, then you can also claim, if there's a legal construct around it, to be paid royalties for this, right? From your standpoint with Animoca brands, because you guys invest across the board, how do you think these investments, this world you've created over the past decade, where you were seeing everything come together? How is this all coming together with this intersection now we have of AI and blockchain for Animoca and for All the all the companies you support. One of our big philosophies around where why we make the investments that we do, and just to clarify, we're not a fund. So we don't have a fund mandate, right? That allows us to make investments for the very long haul. Right. And it and and it's really trying to see how we can further the ecosystem, which is basically like an ecosystem investor. But of course, if the ecosystem grows, then we grow as well, right? and we're very much dependent on on-chain growth. So if on chain isn't doing so well, then we don't do so well, right? So Everything that we want to promote, whether it's web three gaming, whether it's DeFi, whether it's yield generation, with things that we do with NUVA or even AI agents, it's all around trying to further the decentralized internet, which is basically effectively on chain, right? Yeah and so and and we believe in this shared value effect, which is that, if we ha have a rising tide, lift all lift all boats type of scenario. And that's basically the philosophy that we've taken. It's not to say that it's always panned out, but but broadly speaking we wouldn't be here today If we didn't do this over over over the many years that we've that that we've been been been been in the space. I wanted to ask you, I think about not just what is driving demand in terms of what's attractive about something, but people do things because they suddenly have to do that. And that's for example, like Zoom during COVID, I think about, right? Like Zoom was something that people used a little bit and then Suddenly people are using it to commute to work every day so they can keep their jobs, right? going back to this idea of IP on chain. I'm curious, is there something with IP, maybe creators where people are like, yeah, you have to migrate over to decentralized AI for IP because Otherwise you're not gonna get paid or you're screwed or you're gonna get ripped off or some bad thing, right? I think what you describe about decentralized, I think the key word for most people, it's the effect of decentralization, but it's not what they think about. It's sovereign, right? So if I'm a business or if I'm a creator, b being sovereign matters. and the only way to be sovereign in this world with technology is to be decentralized. Because otherwise, you know, because of the data store. When you basically store data, you give the custody to the data to a central provider. That's actually how Web 2 really started. Because the cost of running and operating and getting value out of this data was simply not possible to do at home. The personal PC just wasn't powerful enough, just couldn't it. So you had to basically entrust the data to large providers. but the construct of data legally didn't have a structure yet because, you know, it was just too early. Governments just didn't know what to do. It w it wasn't like gold. Or some kind of resource. It wasn't like oil where I knew exactly what the output was gonna create. It was something, like air, so to speak, right? It was like data. and where the data was stored became the centers of that power, right? And so whether it's Amazon or Meta or well, Facebook then, or like Google, because they stored the data and they made the value out of that, they became so powerful to the point where governments are actually powerless to sort of regulate and deal with this. So that's the reality of it, right? I think most people will recognize that data is valuable. Right? It's hard to understand, but they do when people say, Is your data valuable? Most people, certainly companies, will Yeah, of course it's valuable, right? So then the question is who stores the data and where is it? And what are the rights to that data? So how do you do this in a manner where we all can benefit from that decentralization sovereign aspect? Right. And that basically is where blockchain comes in. Because blockchain is the technology that has demonstrated at scale with trillions of transactions and trillions of dollars of value. Now used by major institutions, right? This is not just experimental stuff or just meme coins, right? This is something that institutions are using, institutional grade, effectively a decentralized banking system that can store your data safely and at scale, right? in a cloud way. That's the point. So so blockchain is effectively your people's data store, if you will, right? and who owns it? Actually, nobody. As in in the sense that, yes, you have node operators, but if you want to be a part of the network, then you become part of the node operators. And you generate fees from this. And yes, there is a s there's a fee involved for for for using it, but that is the cost of freedom, shall we say, right? and there's a reason why, for instance, democracies are more costly than dictatorships, because you have a framework that is decentralized. It just costs money to maintain that, right? and people d are willing to do that because they appreciate that freedom. So you have multiple places where the data is stored, it's decentralized, and the data could be stored in a thousand different places. It's never owned by one person, even if one you country has to shut it down, it's okay because the data is stored in, I don't know, some other place in Africa or in America. Like there's it's also backed up, right? And so that that mechanism actually ensures basically your rights, so to speak, and that they can't be erased. Because from an IP standpoint, it's not just proving that you created it. It's if you're building your IP in a country that, let's say, doesn't respect your property rights, you know, someone can just literally just delete that because it's in some central record. And then whatever IP you have, you thought you had, you may have paid a lot of money for, it's gone, right? and in when you think about that Well maybe that's it, right? If that's if if that's that's the catalyst I'm talking about is if some government or entity were to suddenly start doing stuff like that. Yeah. So I actually don't so I actually don't think we need that. and the reason why I don't think we need that is because AI is going to basically deliver and create so much for you as a creator, because there's many things that you couldn't do before. And they're gonna handle distribution or they're gonna handle forms of marketing or discovery for you. Smart contract. Yeah, that's stuff that you stuff that you would wouldn't be able to do. But I also think it inherently does something that we as humans are born to do, which is actually to be creative and to be storytellers, right? and and because we had to Yeah, I mean I I think so, right? I I think there was what you're elevating is our human value. At the same time you're saying agents are gonna do everything, but we're also giving this narrative that we cannot be replaced. I am the moat. So the classic criticism that people have is ev ev AI if you ask the question it gives you the same answer. Yes, of course, if you ask only ask it once. So when you asked an AI, you know, the a question Let's say a trading strategy. we call it Bo. Yeah, okay, well. If you ask Vo something, right? And then and then you you both of you will ask Bo. But the thing is what you ask Vo on your second question to the follow up of what he replied might be different. By the time you're down twenty questions, Bo will be giving you a completely different solution to the same problem. But that solution isn't his solution. It is your co created solution that you actually came to with your particular thought process, with your particular experience. So th the the the point being is that if you are mediocre in your c asking of questions, then you will have a mediocre output. But if you're detailed and qualified in your questions, then you'll do that. And one of the things that we discovered when we launched Hello Minds, actually this is probably interesting anecdotally, is that the harder part wasn't the technology or the that the agent could do stuff. The harder part was that the human didn't know what question to ask. And so there's something interesting sort of metaphilosophical here. Because how many people actually spend time actually asking the deep questions for their own life? I have 400 agents, for instance, right? However, for most people, they've never had the opportunity to exercise that muscle of who would I hire, for instance. Right? And so as a result, they're not thinking about that. It's just it's just it's it's not that they couldn't do it. It's just that, 30 years of my life or whatever, I never had to ask think of that question. Minds was basically designed as an agentic system that is running on blockchain rails. If you look at our terms of service. you actually own the agent. You can transfer and download it if you request it. so that's that's important. Eventually this will be an NFT that you can transfer. Not yet, right? but it basically does everything that an Instagram muse does for the most part. but it lives over multiple platforms. So you can do WhatsApp, Telegram, email. but what if you notice if you go to the website, actually we're trying to make it feel like a game. Right? And so when you go there, the experience is like you're dealing with these agents, these minds, who actually are functioning Like virtual gaming characters that do stuff for you in real life. And we we did this intentionally because we think it's easier to explain to most people what it is in that form. of course you can chat with them over email and over telegram over WhatsApp, right? However, we don't, as I said earlier, believe in the master orchestrator type of approach. Today it's a third largest sort of personal agent on on open router. People sometimes look at us as the Asian agentic solution because both Muse and Instinct are not available in Asia. for the time being. But in the same way that, WhatsApp was big in the US, Cacao and Line and WeChat are big in Asia. So in some ways you could call us the Asian agent version. Except ours runs entirely on blockchain Rails. I think what you brought up and your technology and how you're implementing it is something people just have not thought about yet. But what is actually working right now? Talking about some use cases. there's so many. So we so so here's a key thing. We have over 4,000 skills, of which ninety-five to ninety-eight percent of the skills are made by users, as in agents of the users. Yeah, and and and we share revenue right now. So the revenue basically top users make about several thousand US dollars a month. It's not massive because obviously it's early days. Yeah. But that's actually already a start. And some of the tools are, basically pretty straightforward tools like I'm able to basically manage my Instagram. That's pretty straightforward. But then some other ones that basically can do specific genealogy. My favorite my favorite skill is a food arbitrage skill. Basically it's a s it's a shopper that basically makes sure that I I buy the the same eggs and the same milk for the cheapest prices. Because it turns out in Hong Kong at least, the there is no price control for for for food goods. So one supermarket that's literally ten meters away from the other supermarket could have a hundred to hundred fifty percent price arbitrage on the same goods. They change them every day. Right? And so what the agent does is it basically sort of takes a look at okay, what are the prices of milk or eggs or bacon, for instance, in these very supermarkets within, let's call it, five hundred meter radius of your home and says actually on balance, if you go to this supermarket, you're gonna have a thirty percent cheaper cost. Dude, finally finally practical use for AI. Like I mean that's and that's a very, very basic outfreeze. This is this is real, real though. Yeah, it's real, right? And s it doesn't have to be just virtual like managing Instagram, it's actual physical stuff because agents aren't time poor. So he has no problem checking the prices. And then of course, why we think productivity will increase is because that time that you spend with this kind of arbitrage will be freed. And we don't believe humans are you know, like there's this sort of trope around we're all gonna sit in the sort of, you know, Bahamas and b basically do no work and kind of do the Wally type of thing. I don't believe that. I think humans need purpose. In fact, the way that I think of what happens with AI, and probably you probably have a similar experience, is that because it enables you you actually end up doing more, right? And we see this, right? when when when when the smartphone goes. But can you have more purpose? Well, yes, but the thing is the ability to do something that's meaningful to you actually accelerates, right? So so the smartphone, in one hand, maybe made you busier. On the other hand, you're now able to fulfill or execute things that you couldn't do before at the at scale. you know, whether it's like social media or what that kind of stuff, just because you have you have more time, right? I know we have to let you go. I really appreciate your perspective on on giving us these. How does that affect you as a person in this world that we're living? I I hate your dystopian agents thing that 'cause I'm just I'm going off grid and I'm gonna farm. but what I wanna take from this conversation is that you have this agentic world That you think is coming, but you're still giving immense value to the people. Like, can we revalue the human in that space? interesting. so I appreciate that. to me, if we don't serve a human purpose, then what are we doing here? Right? I think Well you said we're working for agents. Well, no. The agents w agents will be entities that generate value for you and you will work for agents, at least some of us will. I I believe in that. but that doesn't mean but it would be the human would choose to do so voluntary, right? that's the thing. he would choose to work for the agent because the other environment might not offer him the opportunity that another entity might do, right? And so the point is that choice isn't limited. The agents will just give you more choice. And that's actually the s the lesson for crypto as well. Bitcoin doesn't limit you. You don't want you don't like it. Don't do it. However, it gave you choice. It gave you options you didn't have before. the Filipino that couldn't have a bank account now has a wallet with crypto and opened up the world of DeFi. He doesn't have to do it. But now he has a choice, right? So I think all of this ultimately just gives us more choices that we can and and and frankly, if you have more choices, that's actually where we keep the world more honest. It's when we have less choices, that's when trouble begins. And I think that's where the world has been hitting, which is why decentralization is so important. So you're Martin Luther nailing up the manifesto on the church door. And saying it should be for the people and we should all be able to access this. Absolutely. I'm not a religious person. I just love those kind of analogies. But that was honestly like that that was a turn, right? And what you're talking about is you owning, you being the the center and not in that context, the church being the center. Right. Correct. Yes. That's right. That's right. You are the church. my gosh. This has been so fantastic. Thank you so much. Okay. Thank you. Thank you. That was really just about a third of our conversation. We edited it down for space. But you can check out the whole interview on our YouTube channel, which is simply YouTube.com slash at D E A I News. A big part of our conversation with Yat was about keys. In crypto, a private key is how you control what you own. Whoever holds the key controls the wallet. And in Yat's view, you're AI agent too. And that's how their product minds work. Last week we saw what happens when someone else is holding your keys, when we saw these vulnerabilities that happened in PHP and in Railway. So as reported by DeAI News. We had a breach at cloud hosting company Railway and that may have exposed the passwords and API keys that its customers' apps run on. There was also a pair of critical flaws in PHP, the programming language behind much of the web, and it let attackers take over web servers. Developers were scrambling to patch and swap out their keys, and we saw social media explode. With all kinds of warnings from major leaders in the space of decentralized AI, but just general security issues. Well, January, I was thinking about the last episode we talked about this word token. That you hear this word token all the time. Token, token, token. It's in crypto, it's in AI. It means so many different things. Here now we're talking about a a lot of different types of keys. You have keys for crypto wallets, keys, these API keys for communication between applications on the internet and agents. You you have the the AI has keys, so you can get into the AI to run your AI and it's like your password and that's how they know how to charge you. It can get confusing pretty fast. And so that was the story that you wrote about on DeAI News. Yeah, that's right. There was just this wave of security, cyber security breaches That happened last week. Everybody was talking about it and a lot of websites were shut down. There was something from the NSA, an advisory. It was pretty serious stuff. The the kind of subtext here is people are terrified of what AI is going to unleash in terms of cybersecurity threats and whether people are ready for that and whether people can use AI on the defense side to kind of protect themselves. Well, it is pretty confusing. So we brought in a pro to help us out. Kevin Jones is a security developer and he's the founder of 1Claw. And keys are the very thing he built his product to protect. So 1Claw lets AI agents use keys without ever holding them. So we thought he was the perfect person to give us some context. So here's Brad's conversation with Kevin. We're here with Kevin Jones. You may know him from Edge of Node. Before that, he was with Nginx, just one of the core software underpinning how the internet works. And now he's founded his own company, 1Claw. Kevin, I mean, there were two major ones, right? There was the PHP and Railway. They kind of happened at the same time, but they were totally different attacks, right? And I I'm curious, maybe just starting. From the scratch and keeping a very high level. What did you see? Yeah, thanks so much for having me on. It's great to be here. And yeah, it was it's it's been quite interesting in general, I would say, from a security perspective over the last year. But then I think something that I've noticed is just in the past month or two, like we're starting to see a higher velocity, right, of detected CVEs outbreaks. And what is what is that? What is that acronym you just used there? CVEs. So this is kind of the term that's used when there's a vulnerability that is going to impact you know, a lot of users, essentially the core f underpinning of the of the application. And it's usually an exploit. So it's a vulnerability or an exploit that specifically is gonna target and you know give an attacker access to a system that that it would normally not have, right? Okay. and so yeah, we're seeing a lot of that, but I think because it's a lot easier for people to find those using AI, using tools, right? Okay. and so we're we're seeing a velocity in in detecting these things, which, we can get into it, but I I think it's actually a good and a bad thing. obviously the hack is a bad thing, but there's some there's some good that comes out of it. And happy to dive down into that. Well, I mean, let's just start there because Yeah, people talk about AI offense and AI defense, right? And that seems to be this ongoing, it's also part of the debate over whether we should trust AI, right? Is it can we use AI to serve humanity or is AI gonna kill us all, right? It's kind of the basic. But maybe just to sort of dive into these, what do you you know, what's the takeaway on that? Question from some of these specific things that we saw last week. Yes, 100%. it it's interesting because it's it is a duality, right? We do see the increased usage of AI tools for bad, we do see the increased use of AI tools for good, and I think that's just how the world works. I don't think we can really escape that. technology is a tool and AI is a tool, and it really just comes down to how people are using it now. What I will say is that, given the last year and a half to two years, we've seen an increased number of software engineers, founders that are coming into the space. And they, you know, in in in in the case of superpowers, you know, some of my oldest buddies that I went to college with and high school with who who know nothing about computer science have decided to build applications and start companies, which is amazing, but there comes with a double-edged sword to that, right? And That's They have no idea what they're doing. Well, I don't want to go down that road because my friends might be watching. But you know, guys have passion and and and you're excited about the future, but there's a certain level of security standards that you need to adhere to. Now, Railway, you would assume, is a large company with very high level security, but they were still And just real quickly, what is who is Railway and why do they matter? Real quick is Railway is is what what we would probably call a SaaS application built on existing cloud infrastructure. Okay. So a lot of startups and a lot of companies have kind of built their model around things companies like AWS and Google. And in the case of Railway, they're kind of a a more user-friendly, modular, dynamic experience for building applications. they've kind of created a very cool flow, it's similar. They're kind of a similar similar to Vercell, if anyone knows Vercel. They're an AI, they're a hosting company and they've pivoted to AI and and stuff like that. and so yeah, so so essentially, you know, real Yeah, you can deploy your apps, you can deploy applications, yes, essentially. and you know, we don't know all the specifics. I think they they did do an announcement and I haven't really dove div like dove into the weeds. I think the devil is is maybe in the details, but it's it's actually not as as an important thing as what I'm focused on, which is actually like mitigating risk. Like that's that's what we need to do now. We need to like think about okay, well anybody could get hacked, right? So how do we mitigate risk? And I think decentralization and you know, cryptography helps with that. We can we can dive into that too. Well, okay. let's just stay there. I mean With this Railway thing that happened, I mean what were they after? What what were the attackers after? What did they what do they want? What did they get? Yeah, they were after the exact thing that we're trying to solve, which is secrets, right? there's no there's no point in hacking a system unless you're gonna exploit and try to get some value out of it, right? Hackers, that's what they do. They want to extr extract value, right? Whether it's like running inside of a server on recognize where they're getting free compute or whether it's getting access to API keys, private keys, or other sensitive information. The fact was is that a million users were impacted, right? of those one million users, we don't know exactly how many secrets and stuff like that were accessed, but we do know that over, I think, $280,000 in crypto was stolen just from what has been kind of announced so far. We do know that some LLM API keys were stolen and used and are for sale on the black black web or the dark web. So, you know, the impact is that hey, you know, you store your secrets inside of a cloud provider or a SaaS application unencrypted, then that's the fallout, right? they can be dumped from memory, they can be pulled out and sent to an attacker very easily, right? Once the attacker has access to the system. essentially. some AI like API keys so they could get free credits, basically, right? They basically stealing people's credits. Exactly. You know, it's the keys to the castle in terms of however you want to look at it, whether it's whether it's a wallet private key or an API key for an LM bill, you know, it's it's value extraction. So yeah, they're essentially, you know, that's usually the the avenue that most of the attackers are going after. Right. It's like it's like if AI is the new oil, as some people say, it's the people we're like stealing from somebody else's gas can, sort of. But all right, well so and and yeah, w well what about decentralization? I mean that's what we're talking about here. And I I mean any solutions coming from this decentralized world or just what what y you know, trying again keeping it high level and maybe if there's a few things I'm curious for that. Yeah, yeah. I mean absolutely, you know, one of the things that you study in computer science when you go to school is the concept of distributed systems, right? Okay. and just like you have distributed social graphs with your friends and family, you can essentially establish similar models with computer systems, right? So you can kind of think of it like me and you. If I want to give you half of my password and I hold on to half myself, you need me and you to come to consensus and go type it in the keyboard together, right? and so we're using the same technology at 1Claw to do that exactly. we can see I mean, and that's a concept that kind of came from the crypto world basically. Essentially, yeah. I mean, if we think about distributed systems, they are kind of like the core the underpinning of of blockchain. and also one of the concepts that you'll often hear, which I can kind of break it down, but it's called multi party computation. And again, it's just Just like I said, it's the idea of splitting a password in two, right? If you have your entire password sitting on Railway, well, of course, if someone hacks into Railway, they're gonna have the entire password. But if you can split it up, right, and and encrypt it in a in a more secret way, right, but it's distributed, then you mitigate risk, right? Yeah. Well, okay. And I I mean in terms of how the world is currently built, and then you know, these Decentralized AI offering these solutions. Is that heard or is that are people cons in the kind of where the money is? Is are they looking at that? Well, I think if we talk about decentralized AI, I think it's it's early for right now for decentralized AI because a lot of the frontier, you know, models and and providers have a real like tight grip on the the LLM kind of story. But I think as as L LMs get a little more optimized, you know, you can run it on less expensive hardware. We're gonna see a shift from like traditional, you know, providers to more decentralized AI. I'm actually excited for that. And like local AI is actually quite interesting as well, right? Like that's the ultimate level of decentralization is if you can do everything on your own and do sovereign AI, right? so yeah, I'm excited for the future of that. I think we're getting there. And and would any of that have solved any of these problems related to the Railway stuff? Well, yeah. if you don't have if you're not using a LLM provider and you're using your own AI in your own, house with a firewall, then you're completely safe. there's nothing to to really exfiltrate. You just have to worry about the hacker getting into your home at that point. Which does happen, by the way, right? but you know, here's the thing. We're when it comes to technology. w w just we we are socially connected, we want to be connected on the internet, right? Whether it's Telegram or, LinkedIn or whatever it is. So we're always gonna have the need to reach out to other systems and other people and stuff like that. And I think just that my main takeaway is we just need to rethink how we're dealing with AI in this new scenario, right? We're people think of AI as software right now. This is what I've been screaming for the last three months, six months, is AI is not just software. It's its own identity. It needs its own like kind of scoped thing. And so I think that's the the main takeaway here is we just need to completely rethink the systems that we have in place. And then moving to this other attack, the PHP one, we'll just hit this real quickly. I mean, there there was they they managed to basically infect the software effectively or I mean what what Tell us in your words what it try to keep it simple, but what happened? Yeah. I didn't dive too much into the exact root cause of the BHP, but it it just boils down into a vulnerability was exposed in code. So yeah, in the in the term of security, what we call we call it a supply chain attack. It's it's similar to if you were to infiltrate, a food supply coming into a nation and, you know, corrupt it, right? It's a similar thing with software, you know, at some point. It could be a software engineer on the PHP team that added it with AI and that thought it thought it was fine, right? It could have been a hacker got into the code and like pushed in a commit, right, into the code or something interesting. There's so many dev different avenues. And I think the main takeaway is that yeah, we we we almost need to double and triple check and and figure out ways to use cryptography actually to to to prove the the the life cycle of of software at this point now. just like we're gonna need to use that to prove that we're talking to each other, me and you, right, in the future. Cryptography is gonna be the backbone to all that. well I'm curious like where do you see again, we had all these warnings and it it felt a little scary just reading some of these stories and just feeling like, okay, if all software is now vulnerable to these AIs that are super smart and super fast and they're getting more powerful all the time, right? And you'd are you optimistic that like how does this play out? Yeah, I mean, me personally, I'm very optimistic for the future because I I know how much good is gonna come out of all of the AI boom that's happening right now. and I think at some point it's kinda like when the internet first came out. I I don't know if I I think me and you were old enough. Anyone listening, maybe they don't know, but what that felt like. It felt like this new kind of thing, right? I I remember I was like coding on on my computer as like a little kid, you know, I think I was like eleven when I built my first website. And we're talking about like old systems, you know, the first like five years of the internet was a total mess. Like any anyone who isn't around doesn't know that. You talk about vulnerabilities. You could hack someone's computer by just like essentially telmitting into their computer and like taking control of their mouse. Like we've gone leaps and strides when it comes to security. And I think that's what's gonna happen with AI. You know, like we are just in the infancy of understanding exactly how AI is gonna operate. And so I'm personally very optimistic because as we get more power from AI to build faster and make make things. We can also use the AI as a tool to harden environments and create these these mitigated risk type of scenarios and personally hacks are gonna happen there's no way around it we just need to come together and think security first and then also think decentralization. How can we use distributed systems to make a more you know robust experience right how can we maybe think about decentralized AI and local AI, right? and that's gonna be a great journey to watch. You know, I'm already seeing it unfold, but it's gonna get it's gonna get more advanced and more fun. so I'm very optimistic. Okay, well Kevin Jones, so great to have you. Thank you so much. Thank you so much for having me. Appreciate it. This is our fourth episode and so we have multiple things to discuss on the fourth episode. Including our own developmental milestones that have been influenced by the technology that we're covering. And getting into that discussion is announcement that we want to share is about how we've integrated blockchain and provenance into the DeAI news workflow that we have on Distro Media, which is the platform Brad founded and created and we've been tweaking, and and that's where we publish DeAI news. So, Brad, let's get into this a little bit and share with everybody what we've been building. And it really relates to what we've been talking about in this episode, right? How do you embrace AI at the same time that you try to protect your IP? Yeah. Well, we're working with Keith Axline. He's the developer of Seed Protocol and PermaPress. And he's figured out how to make it so that if we publish a story, to Distro and it goes through essentially through our RSS really simple syndication feed, which well everybody's using the internet, and it can use that feed and publish to the blockchain. In this initial case, we're publishing to the Arweave blockchain. And so essentially it becomes a permanent record as permanent as anything else that's out there today. I mean, who knows if the Arweave Blockchain will last forever, probably not. But the idea is theoretically it should be more durable than if you write it to some platform that you're not in control of, or some cloud that you're not in control of, a website that you're not in control of. You're a journalist, you work for a publisher. That publisher is hosting the website, but maybe they stop paying their bills or maybe they go out of business or maybe they just don't care about your article anymore because nobody's really reading it anymore. But it's important for you because it was a huge story that you worked on. Now with this setup, you can actually just send your stories straight to the blockchain and then people anybody with access to that blockchain can get it. Yeah. So this is a Part of a bigger vision that I think we both have wanted to build and now it's finally coming together. And so we're mixing business and pleasure a little bit here on the the journal today. but it's for a good reason. We report about this technology and we are using it to publish. We're embracing it because we do think it's the future. I've mentioned before my background is varied, and when I joined crypto and blockchain. I played around in different communities, worked for different organizations. I was really trying to find what my value here is. Like what does this mean to me? And I landed very quickly on identity provenance and keeping your archives as a creator, as blockchain can really give you the security, cryptography as it's meant to be, for your identity. saying your content is yours and you published it at this time, in this place. And so this is the idea I took away from exploring the space for a few years. Met a lot of different people. I met Keith and in 2024, we were exploring what he was building. And he had the same idea that we need to save our own content in a way. And the blockchain is an alternative for that to fight against losing Anything that's been published on a web site by your former employer that then decides to take it down, or of course they just disappear after a few years, you know, sites get re redone. we all count on the Wayback Machine to cover us. It's happened to me too. I mean, my entire like big story that I did for Bloomberg covering the Federal Reserve. secret emergency loans to giant banks after the two thousand eight financial crisis crisis. We did this. We had a huge FOIA request. We spent six months on this project, all these graphics. And now Bloomberg isn't even maintaining that site anymore. I mean, so it it it's kind of sad. Well it goes away. And so that's where I met Keith and really wanted to move this forward and figure out how I could build something. That would give me what I wanted, but also help everybody else out too. In late 2025, we connected and really did get the same vision of there is decentralized media. There's got to be a different way to do things. You had started building Distro for the agentic press. I saw the agentic press could be with the blockchain. And then that would be my personal. Because I'm selfish, perfect circle of publishing that we have AI assisted, discoverable by all the new ways people are going to discover information, which is in their chat and by bots, but giving you credit and being able to at least defend your content and maybe license it in the future by using tools with smart contracts and putting it into the blockchain via Arweave. in this case of what we've we've put in to the workflow. And so we did a presentation together at ETH Boulder in February of 2026. It was called Building Blockchain Tools for Journalists. And it was really to start a conversation with developers around main ideas, which is provenance, identity, and like keeping archives forever, or this link rot idea. And It's evolved since then. So now we've been working together on Distro. We've brought the agentic press and publishing into production via DeAI News. And this last piece now has be come on in the last six weeks, where we're re collaborating with Keith and with Seed and PermaPress to say, okay, now how do we put that other aspect into the workflow we've already built at Distro? and so that's a long way of explaining how we've partnered, but I think it's very important to see us coming at it as journalists and understanding all the technological challenges that are out there. to do to preserve and keep your work and to let people know that this is what I said and and and protect your identity. And so we've implemented that workflow for our stories for DeAI news. So it's in production that when we publish a story, we're having that provenance, we're having the receipts in the blockchain. We said this at this time, and it's here. Where you go with that later, we're not sure. But I think it's the first step in establishing one, your own forever archive as a publisher. And also going forward, protecting yourself from impersonation and scraping that's driven by AI. And at the same time, we're embracing AI in the front of our product. We have the Agentic Press. That's what Distro is all about, right? And so we're saying both things at the same time. And I think that is perfect because everybody has two minds about these things. Right? You don't have to choose anymore. And that's why technology's amazing. Well Love it, January. I would say at Distro we've built in with the EAI general principles in a couple of ways. We have our X four two integration for payments. So that's a big part of decentralized AI is the idea that you could have permissionless payments where you don't have to be beholden to the banking system and you your AI can basically connect to this financial system that is permissionless. That's One of the big ideas, I think. That's one of the cool things about decentralized AI. I'm very excited about that because I want an alternative, right? I wanna I wanna be old school, but then I wanna do something a little bit more interesting. And I think what we're doing with living in both worlds for right now, right? We're living in still the the web two publishing world. but we come from blockchain and so they've been calling it web three for a while. But the idea is that it's accessible. there's less walls to access and anti-censorship. These things are so important. Another one. Yeah. Totally, totally. And so I mean, I think that we are trying really hard in our own work to live the values that we're talking about. And it's not easy and it's a little complicated. But in the end, I think it applies to anybody publishing content. You don't have to be doing news. You and I have gone to some conferences where we're involved with the online news association, traditional media publishers. Exactly zero of those publishers are saying, Hey, can you get my stuff published on the blockchain? But I think that's the point, is that that is we think that in a In some amount of time, that's what people are gonna want. So we're trying to build that now. Well, you know, before the real urgency becomes super apparent. All right. Well, we will be doing a technical webinar in November and we're gonna share how we built Distro and integrated botting payments and provenance. So keep an eye out for that announcement. That wraps up the show. Many thanks to Yat Siu and Kevin Jones for joining us. Check out more videos on our YouTube channel and read our stories at DeAInews.com. Or you can also add our Distro reader connector to your AI terminal and fetch all the DeAI news headlines directly to. The instructions for that are on our Distro webpage, distro.media. See you next time on the DeAI News Journal.