Why AI Agents Need Blockchain Identity? - Yat Siu | ATC #610

Yeah, I mean I think most people obviously know, at least on your show, who we are and what we do. But I would say, as I told a lot of people, my background: I grew up in Europe, and maybe not a lot of people necessarily know that actually German is my mother tongue. So that's probably something that sometimes surprises people a little bit, just because I don't look particularly German, but that's just because that's what we speak in Austria. But I think the other thing also is I've been quite involved in building schools, supporting the construction of some of these in some of the less fortunate places. But you know, I'm on the board of the Dalton School. We started something where we were really trying to develop essentially divergent and innovative thinking. And that's also one of the reasons why when we launched Open Campus, which was basically a blockchain education program with EDU, there was a little bit of that ambition around, let's try to bring the benefits and the power of education to everyone. So that's kind of probably something that most people might not know in terms of the background. And maybe the other thing, I have three kids and two of them are in the US right now, basically having, interestingly enough, a pretty good time, despite all the controversies around, right? And then my youngest is still in Hong Kong. And I think we're very fortunate to have very dedicated and happy kids. One thing I will also say, one of the reasons we started getting into education is because I think the world of learning is changing rapidly, and especially what's happening with AI, which we'll probably get into as well. The purpose of school, what it was meant to be, has completely shifted, but it's not moving quickly enough. And there is a big question that I certainly have, which is, is it really worth getting into that kind of debt for an education that's probably not going to give you what you're looking for in comparison to what it was maybe 20 years ago? I'm not saying you shouldn't go to school, but you need to ask yourself the question why you need to go to school. And maybe the one example is my oldest was working first before he went to school. He was working for 3 years and then he went to school, and it's an entirely different reason because he's not worried about getting a job in the future because he's basically marketable because of his work experience. So he's going to school to really learn. And one of the interesting insights he shared with me was he takes really, really hard courses and he doesn't care if he gets a lower grade because he's learning something new and that excites him because he's not worried about having a certain GPA that will help him get a job. And while that makes a lot of sense, it sort of struck a different light bulb in the thinking, which is, well, actually that should be true for, I mean, if you're hiring someone, do you want to hire the kid that got the perfect GPA or do you want to hire the kid that challenged himself and took the hardest courses and got an okay grade as opposed to the best grade? Because he really tried something that was difficult and then eventually excelled in it. That's much more valuable because that's how work is, that's how life is, as opposed to trying to sort of, let's call it, game your way into the perfect score, where you've never really failed because you never let yourself fail because you want to have the perfect GPA to get into the so-called perfect job. But I would argue that kind of person is less employable, but he has no choice when that's how head hunters or recruiters look at that, right?

Exactly. And in a world that was like that maybe even just 10 years ago, and certainly the decades before that, where that worked because the kind of person you hired was not a person who was daring and challenging and experimental, but rather was someone who was predictable, shall we say, and working more like a machine to do what he was told, that made sense. But we're not at that stage anymore. And the future is changing rapidly, especially what you see with everything with AI and agentic AI, and of course everything happening in the blockchain world, which I'm sure we'll get into. You can't expect that. And so that to me is something we really have to think about for the future of our kids.

So I actually wrote an article about this around where I thought the future was going in terms of jobs. But the summary of this really is that in fact, classic software coders, they're not safe from a job. I mean, if you think about all the layoffs that happened recently, actually it was particularly the software departments that were affected. And one of the ironies of course is that what programmers are, are the ones who translate. When you think about it, programmers and software engineers for the most part were translators for the machine economy. They were the ones who were able to translate human ideas into basically computer language that a computer could interpret and do stuff with, and then voila, you had the software. But now basically computers are better at translating it than humans are. And so when you look at things like what's happening with Anthropic or with things like Replit or so forth, all these things are basically just making it so much more easier and you're getting to the point where basically a lot of the code that's written by machines is better than the code that humans have written in and of itself. So what is it that basically distinguishes ourselves? And just stepping back a little bit, people used to say that about the industrial revolution as well. Like, okay, well, we go from horses to cars, we go from farming to machines, right? And then the machines basically take away all our jobs. But actually when you think about it, there's more jobs and there's more abundance and there's more opportunity because these machines even back 100 years ago produced more food, produced more materials, produced more opportunity. And say what you will, society today despite all the issues is significantly better than it was 100 years ago or 150 years ago or 200 years ago. You look at the data, it's very, very clear. And we're reaching a similar type of stage, but now the abundance doesn't come in the physical mechanical space. It's now coming essentially in the virtual software space. And what that means is that you've basically got an entire machine economy that's going to accelerate and crypto is going to play a huge part in this because that's basically where the agentic commerce will take place. And so what is it that distinguishes us as humans? It's really our creativity and it is basically our ability to connect the dots in a way that a machine cannot. And I just want to dig into that a little bit because a lot of people are like, oh, machine is super intelligent, whatever. So, if you've used things like ChatGPT for instance, and you go into the context window and you speak to them about 20 different topics, he'll start to hallucinate and it'll get really weird on you. Which is why every time you start a new session with ChatGPT, you start a new session, it seems very good because essentially for that search, he's excellent. But actually what happens is that when you get to deeper context like we see with Claude, what happens is you have lots of agents running on the back and they're specialized agents. Like if you basically use Claude or Perplexity and you ask a question, they call many agents on the back, the legal expert, the Canada expert, whatever experts. They're there only for a moment, they answer the question, and because they're experts in this area and they don't have this long context window, they don't hallucinate. And that's basically how people harness or connect these agents together. What I mean by that is it's essentially orchestration. And so how to orchestrate well means that you have to know how to connect the dots. And so how do you connect the dots? Well, that means you need to know about life, about society, about really the humanities. And so the irony of course is that the most important skill I think people need to learn now is the humanities more than it is how to code per se because a computer can code well. So you need to understand how it interacts and how it comes together from a human perspective. So I would say liberal arts are important, philosophy is important, history is important. These are things that you might say are not as practical, but when you come to a point where the actual skill of doing the software isn't that important, then these things matter much more. So the creativity aspect, and I want to leave the audience with one important bullet point here, which is in the 60s, George Land and Beth Jarvis basically had a study and they were running, they were actually running and helping essentially the NASA recruitment program back in the 50s and 60s because they needed to really hire very divergent and creative thinking people. So these people, NASA scientists, you send someone into space and it happens, you better be pretty creative because if you don't know what to do, you're out of luck because there's no guidebook or manual that says when in space there's a problem, do this. So you have to find really divergent people. So they created this test and this became a pretty famous study that became the body of work of people like the late Sir Ken Robinson, where from the 60s they basically followed a bunch of kids from the age of five until 15 and they basically gave them a test that they designed sort of inspired by the NASA test to see how creative they were and then they gave that same test to adults who were 21 years old. And so, take a guess how creatively the kids scored. I mean, I probably already kind of gave you the answer, but take a guess as to how creative the kids were that were 5 years old.

Right. But let's give you the numbers for the five-year-olds. What do you think it was?

Yeah.

Yeah. So it was actually 91%. 91%, right. And then by the time they reached 15, it dropped to something like 8%, I think.

And the 21-year-olds basically scored 2%. Now, so that seems pretty devastating, but there was another point which of course was not picked up, but an observation that I then made, which was that, well, it's interesting that it's 2%. Because what is the entrepreneurial index of the world today?

It's roughly low single digits, about 2%, right? So the same, so that creativity, or basically the people who are the ones who tend to be more entrepreneurial in a disruptive or incremental kind of way. And the point on this, at least my personal observation on this study, is that if we preserve the creativity of our kids at the young age, right? Even if we just preserved it in a way where it goes from 2 to 4%, okay, which doesn't seem like a big number, we've basically doubled our global creativity GDP as it were. And so, from those numbers, if you basically get to preserve more of it, actually, you will basically create this entrepreneurial explosion. And that's kind of what we think about startup culture. Because when I started my startups back in the early 90s, late 80s, which is obviously a long time ago, it was not cool to be a startup. Startups were for people who were unemployed.

Exactly, or unemployed or not employable, simply speaking. The world was not ready for the Elon Musks, they just would have been like, you're weird, I can't deal with you. And one of the reasons why they all started a lot in technology, those disruptive entrepreneurs, is because it was such a new field that there was no incumbent, there was nobody to work for, so to speak. And so that allowed people to basically just go unshackled and attract, as they call it, the revenge of the nerds, they're like the weird guys back in the day. And now today, being a startup is kind of cool and there's a lot of them out there, but it's still a low number compared to the rest of the employment. So if you just track the number of entrepreneurs, it's on the way up, and it's going to go even higher up. And so you talk to me about the future of jobs. Well, the future of jobs is very likely that I think the probability for your child is that the chances that he's going to be an entrepreneur is significantly higher than is today. So then what skills does an entrepreneur need? And it needs that enterprise and that grit and that creativity and that divergent thinking, which is not what you learn in school traditionally because schools don't train you for that. They train you for compliance, they train you for working for someone, they train you for order in society, which is everything that an entrepreneur is not. So that's kind of a long-winded answer to your question.

Well, okay, so first of all, there's a little bit of this thing, you know, I think they say it in English as well, but there's the classic German expression of Schadenfreude, which is basically taking joy in other people's misery. And so anytime when an industry is not doing well or business or whatever, the people who don't like it are usually the first ones who just come out and say, 'Told you so. This is rubbish. This is crap.' And crypto has a lot of them comparatively speaking because you think about, for those who don't like crypto, traditional finance people or people disrupted over the space. But what's changed now is that the institutions have all started to adopt crypto as well. So BlackRock and Fidelity, Standard Chartered, all these groups are basically adopting it, selling products in this area, seeing the opportunity. So that's changed. But historically that's kind of usually been the case where when an industry is growing and it risks to disrupt a whole bunch of people, particularly themselves and their industries, then they tend not to like it. And we see the cycle through and through. I mean, in the 90s, actually when the dot-com bubble burst, I think this was frankly around 2001, there was this famous article that went out, 'The internet was a fad.' And it obviously one of the absolutely worst predictions ever made. But it became a meme in itself, but that just shows how people are not really able to see this because it all comes from your own biased lens. And if you are generally, it's hard for people to think exponentially, right? Just broadly speaking, I think it's easier today than it was before because we have many examples, but back then it was kind of hard to do that. So it's not that you would fault them, it's just they simply didn't have the imagination to do that. And then the kind of people who had the kind of imagination to see where the world would come and go were really frankly a lot of the science fiction writers. I mean, if you think about Asimov or even George Orwell, in the more doomsday type stuff, they were basically writing about the future and they thought deeply about it and there was a certain element of that ability to do so. I mean, even if you look at shows like Star Trek, for instance, if you took the number of gadgets and ideas and concepts that basically came true today in slightly different ways from the creation of Star Trek, you would go, back then you'd go like crazy and now it's like instant translator. Oh, wait, hold on. That kind of works, right? Like the iPad. So you have all of that. So to me, I don't view it really as, you know, like everyone talks about this is working backward, but the reality is, and by the way, it does twist and turn, but the general reality is, is this better for people? And let's go to the example of the instant translator, which sounds like a crazy idea, but isn't it a cool thing to have? Yeah, of course. Wouldn't society be better because of it? Yes. So whether it takes 10 years, 20 years, 50 years or 100 years to get there, it gets there because it's just a better thing. And you have even more philosophical constructs that became political like freedom, property rights, democracy. I mean these by the way in the arc of human history were very small. This concept of property rights is effectively a 16th, 17th century popular thinking. I mean it existed older of course but only a certain privileged few had it. So we have all gotten used to capitalism and property rights and we take it for granted which of course laid the foundation for today's democracies. But thousands and thousands of human history pre let's call it 17th century effectively it didn't apply for 99% of the world and they never even knew that. In fact, most of us didn't even have a last name because we were someone's slaves. So this is how short that is. But if people back then, which was an idea, said, 'Oh, but you should have civil liberties and you should have justice and you should have your own rights to your property,' everyone goes like, 'Oh, yeah, actually we should.' And of course it took a long time and unfortunately a lot of blood to get there. But it was inevitable because it was better. So therefore you go into things like the internet. You could poo-poo the internet all you like. But is more information better for people? Is more sharing of knowledge better? Of course it's better. And then we talk about crypto. Is more distribution of wealth better? Is more ability to have the rights of your digital property yours better? Should money basically be yours and not owned by someone else? And should it be something that is permissionless and transferable anytime you want? Yes. Right. And so yes, there will be problems and there will be things that obviously happen in between and like with any growth of an industry, there's going to be unfortunately bad actors as well. But broadly speaking, if it's better for people, it will succeed. And that's definitely the case here.

Yeah. So, I mean, the granddaddy of digital ownership really, I would say, is probably Bitcoin, right? Because it made it possible to actually own your digital property in a sovereign way. Because you could argue that you had a kind of partial ownership type of relationship with your gaming items in the past. Like I bought my very first gaming item unofficially effectively because it was a black market in a multi-user dungeon I think it was in '89 or 1990. And the way that I did that transaction was I basically bought a sword and the player basically, I had to literally send him a check, like a physical check, that had to clear. So that took like two weeks and then after it cleared he said okay I got the money. And so it's just like all sort of, I didn't, he didn't even meet this guy, right? Just imagine the trust concept. And then he, but the trust of course comes from the fact that I knew him from playing the game with him, so there was the element of that and then there was a social network there as well. Anyway, so he went to this, I don't know, like the equivalent of some safe bar in the game and he would literally drop the sword so that I could pick it up. That was because there was no mechanism in which he could transfer the sword.

But so many people did that and it just demonstrates also partially the human ingenuity that comes in space. Like I want that sword, I'll figure out a way to get it. And there was a big marketplace around it even if the game never designed it for that purpose. People wanted to have the ownership of that. And a lot of people think of ownership first in terms of its utility, that is true, but ultimately what ownership represents is identity. So when you own your house or when you own your title to things or when frankly ownership in your name, these are all part of your identity. I mean you're Canadian, there's an ownership in that. You are as a Canadian. You stake a claim in the ownership of Canada. There is a patriotism, a nationalism sort of implied with it that forms part of your identity. So the idea of ownership is so critical and is so tied to freedom because if you can't own anything then you can't have freedom. It's as simple as that. And I think George Washington back in the day said it best. So let's take it into the digital world. In a digital world basically we don't have ownership of anything because the most valuable construct is really the raw material, which is data. Data powers AI and Facebook and Google and the whole advertising industry which is close to a trillion dollars a year. So it's pretty massive. And also is basically generating the trillions of dollars of value. I mean Facebook represents, their revenues is over 90% advertising and for Google it's 77%. So just sort of, and we're talking about companies that are worth trillions. I think it's like somewhere number three, number five biggest companies in the world are basically just advertising companies. And how do they make the advertising? From our data. And we generate this value because we spend time on these networks creating network effects and creating basically data that they use and we don't get paid any for it because we don't have any rights to that. And part of the reason is because it sits on their databases. So what is a system that is essentially open and permissionless that allows you to have ownership of the data or at least the value sharing of that data? And that's basically blockchain. And that's why when we saw blockchain really with CryptoKitties back in the day with NFTs, which was one of our first projects we got involved in, that's when we realized, wait, hold on a second, we can actually really own a digital property because prior to that we lacked a system. Because the enforcer of our physical property rights is the government normally. But in that digital world we don't have a government that enforces it. So we need to have a different form of enforcement or at least verification. And that basically became blockchain. And while the government...

Right. Well, it's also more important because we're more digital.

Right. As in the time that we spend and we will spend more time digitally as well. But you still need the government to enforce, right? So what the digital property rights do is what the physical property rights do as well, which is your proof of ownership and when there's a loss as in a theft of some sort, you can prove it and then the physical side of the enforcement can say, 'Ah, right.' So at the end of the day, if someone steals your Bitcoin, you're still going to go to the police and say, 'Can you help me?' Right? You're not going to go sort of and try some kind of online version because that doesn't exist. There's no online enforcement for that per se. It still has to be a physical form for that. So I think that still has to exist. But in terms of the journey of Animoca, so we started off with gaming because we're a game studio originally. And we realized that we could really own our digital items but of course we didn't fully comprehend at the time that DeFi was so critical for that. So we made all these investments in OpenSea back in 2018, 2019 and acquired Sandbox and Axie Infinity and many of these companies became really, really big in '21. And that's kind of how Animoca suddenly became really famous as well because it wasn't just the projects we were building. It was also the fact that our portfolios all suddenly became basically unicorns and these billion-dollar enterprises. And so a lot of people think of it as like overnight success but obviously we worked on that really for many, many years before.

Four years ago, same thing, same thing. But I think as we developed that, we really came from the thesis point that people really wanted to own their digital property rights. And when you think about the number one reason why people play blockchain games, it consistently year after year ownership comes out as the number one reason. Not so much the, obviously things like play-to-earn or speculation or whatever, that's one part of it as well, but the number one reason is ownership. And you just have to look at games like Roblox or Apex Legends or whatever, not just the items and the value but how people crave them and think of them as something that is important to their status. Basically just illustrates that example. So that's how it started. But then came DeFi and with DeFi it really exploded because then we realized how the monetary aspect came connected and we became more of a financial organization because we realize that basically property rights and finances really are very much tied together. And so we became more institutional. Today we have obviously RWAs. In fact, literally just a couple of hours ago our joint venture, Anchorpoint, which is a joint venture between us, Standard Chartered, and Hong Kong Telecom, just announced essentially the receipt of the stablecoin license in Hong Kong. So that was basically one such development as well. And of course a partnership with Ondo on RWAs. So that development of course has taken place as well. And then of course there's tokens as well. All the token elements are basically these digital representations. Moving on to the next thing that of course we think are going to really drive the blockchain, which is basically the agentic web, which we like to call web 4, which if you have time we can talk about that as well. Because the property rights aspect, where our head is at right now, is that the property rights aspect is important for humans but it's even more important for agents. So agents are the ones who need digital identity more than humans do because there's going to be billions of agents swarming on the web. How do I know who's real? How do I know what their reputation is? How do I know who's behind them? Digital identity, blockchain, right? And how do I know who's going to do all the on-chain commerce transactions? It's probably going to be an agent. And so forth. So we really think that AI agents are going to be basically the mass adoption use case, serving humans to be clear. But essentially they're the ones who are going to be doing all the on-chain stuff because they're not going to be the ones who have bank accounts. They're not going to be the ones who are trading on centralized exchanges per se. I mean they could but really the benefits are basically on-chain and all the things they could leverage from that because for them it's their native language. If you think about blockchain, just as I had mentioned earlier about programming and code being essentially the language of the machine, blockchain is effectively a platform that is built for machines.

Maybe the framing of the question is perhaps the problem as well when people, because I think the best founders don't think about cashing in per se. So of course money is part of the success and there's a reward of course if you do a good job. However, if you're building it for the purposes of cashing in, there are many other ways that you can cash in frankly speaking. And what that means is that persistence and grit is very low. And I don't mean to say that you can't be persistent trying to make money. I'm just saying that the journey of what you're trying to build, if you don't find purpose in it, if you don't think it's meaningful for you, you're not going to do it. Because it's tiring or you hate going to work because of that. And so it's just not going to work. Like when we back, we have over 600 investments, when we back a founder, we try, and it's not that we always get it right, but we try to find founders who have a long-term vision, whose identity is staked on the project, and who really believe that what they're building is meaningful, impactful. And of course, many of them might not be as meaningful and impactful as it is, but they will give their very best to make it happen. Whereas if your purpose is, hey, I just want to cash in and make a quick buck, which unfortunately in crypto world there's quite a few of them, then you're going to have the same effect that we've seen in the past, which is lots of short-termism. And not building for the longer term. And when we talk about our vision, our vision is around digital property rights, it's very clear. We've always been saying that how we get there though might meander a little bit. It changes because the markets change and we also learn new things. I give this example often, there's many examples but this is probably a famous one, which is most people will remember Amazon was a book seller with a very clear vision around serving their customers relentlessly to whatever they wanted. Of course today Amazon is a lot more than a book seller. They still sell books by the way, but you don't think of Amazon as a book seller. But the mission is essentially customer-centric and basically serving the customers to fulfill whatever they want. And they've just fulfilled delivering the stuff that the customers want. Becoming not just Amazon the marketplace but AWS, Alexa, whatever it is that they're delivering as a result. So the north star I would...

The argument for Amazon has never really changed. It's just that the markets adapted and they adapted towards it to fulfill that vision. Amazon could certainly be the biggest book seller in the world and nothing else, but it wouldn't be the powerhouse that it is today. And that comes with the vision as well. And that's certainly how we see things with Animoca. And if you're building a company that you think should be here for basically decades and decades to come, then what's hot this year or next year is something you navigate but it doesn't define you, if you know what I mean. And a lot of people would like to sort of frame us and say, 'Oh, web3 gaming company.' Well, it's still a part of what we do, just how Amazon still sells books. And the difference is that we don't shy away from it. We don't disavow and say, 'Oh, gaming, whatever,' because we think it's still very important. But gaming is a much smaller part of our business because we've grown, of course. And we had to sort of navigate the market. I mean, if we didn't, then we wouldn't be around anymore.

Absolutely. And so one of the things that we're pushing heavily now, which is similar in the same vein of digital ownership, is AI with Animoca Minds. It's the same thing. The only difference is that, let's call it a pivot, but the only difference is that we think basically the ownership of digital property is going to be even more relevant for an AI agent because it has to prove that ownership so that it could basically do agentic commerce with another agent as opposed to humans. And that's not to say that humans aren't going to be players in the space as well. It's just to say that the ratio of agents to humans is going to be significantly more agents than humans. And that's the world that we're basically constructing right now. And that's basically to us an inevitability.

So there's multiple parts. I think what you're addressing in terms of identity is super critical. But let's talk about numbers first, because I think that helps then frame the opportunity and why it's so important. So if I was to ask you, and the audience can consider thinking of that question as well, like how many agents do you think you'll have in the next 3 to 5 years?