Yes, I am the founder of Netamines , but now I am also the founder of Complete Robot. This is something fresh. It is a wholly owned subsidiary of Netamine. So if you think I'm leaving Netamine, that's not true. I'm just creating a slightly more concise structure under this mind to focus on a whole new style of organization.
So I've been talking a lot over the last, um, two years, I think, about agent-based AI, about robots, how Ethereum is going to facilitate coordination of agents, robots, how to solve identity issues, and so on, and, uh, I also thought I heard a lot that we should support developers, be with developers, and think about real-world use cases for Ethereum. The process of how to do it, how to prove it, is better than just jumping in and understanding all those challenges that you otherwise miss. Because I would keep telling people that it's easy, you can just do it. Now that I'm doing it, you know, sometimes it's a little more difficult, but it's incredibly exciting. Hmm, so I want to focus on the ideas of where, where financial markets and Ethereum might be headed; on understanding where the opportunities are emerging and what the asset classes might be, Ethereum formats , hmm, and where it is.
So, AI agents , computing, robotics, or embodied AI or physical intelligence, or, let me remind you, there's a lot of talk about automated payments, um, through AI agents , agent-to-agent, and also machine-to-machine payments. Uh, agent commerce, you know, probably one of the best outcomes and, uh, the most promising, uh, simple cases where you can implement this right away is agent commerce, uh agent commerce, uh , but compute accounting and compute payments, uh, recently... So, I'm working with, uh, well, I was building my startup at the HFZ residency recently, uh, in San Francisco, and there 's also an AI team there led by Blockr, and they're doing something like compute accounting on the network, and agents are using payments, uh, the machine, to process micropayments, fleet billing. And in fact, the direction of fleet billing and self- developed business is where developed business is where I am now , where I am building. So, thinking about the economy that you imagine without people, not that there would be no place for people , but if you take it to the extreme and imagine how the economy would function completely autonomously with agents, with robots, uh, what changes in structure, in mechanism would be required, and what needs to be created now.
So, that's where we can all move, at least as an experiment, and then it's great to ask questions about where Ethereum fits into all of this. Hmm, thinking about all the recent things I've been watching on Twitter. I'm not sure you know what it's like when you go down a rabbit hole and only hear about this one thing and then you start assuming everyone sees it too. But maybe you've seen it. So here's the figure: I, a robot, compete with a human in sorting packages, and they work for over 100 hours, and they say the human can win for the last time .
Uh, and I like this comparison that a robot can work 6,000 hours a year. A person can work 2000 hours a year. Then I started to realize that some of us are robots because we are probably closer to that first number. Hmm, but that raises a lot of questions about what's actually happening with the workforce, where the implementation is happening, because we probably all think that robots are made in China. China is leading, um, in the US, uh, there's a significant movement in the US towards creating robots.
But now, when you think about where you can use robots, everywhere the cost of a robot, uh, the cost of a robot's labor per hour, amortized over time, per hour, uh, is lower than what you pay people, which means that actually, if you think about the U.S. labor market, it's very expensive. So maybe you'll have a lot more opportunity to start introducing automated work and robots into the, uh work and robots into the, uh , workforce. And then that opens up many, many questions about what's involved, like whether the insurance market is changing, regulation, financing opportunities.
Uh, can we start exploiting robotic fleets as asset classes as well ? Uh, it could be not only robot taxis, but also robot work with hourly pay . Uh, and you see some of these charts. So they propose to look at how over time the cost of a robot hour, amortized as production costs, will get lower and lower, uh, and probably the cost of human labor will go up. This is an interesting assumption, because perhaps the challenge of the economy of the future will be the reduction in the price of human labor, or when we start specializing, we will specialize in areas where robots absolutely cannot help us, and the cost of that labor hour will increase.
So, something is being achieved there, and that means there is some level of equilibrium, and it looks like it's good to look at the comparative advantages of different countries and our skills as people. Hmm, that's the density of robots. In fact, if you look at China, the density is not that high, but if you look at the absolute numbers, the number of industrial robots deployed in China certainly exceeds everything else. So, if you look at Korea, Singapore, Germany, Japan, they lead in the number of robots deployed compared to manual labor in production. This is an order for robots in the industrial space in North America.
So, as you can see, there's actually a lot of potential because it's not growing, but at the same time, the technological advances over the last, uh, year or two have brought a lot of excitement to the robotics community, uh, the builders community, the maker community maker community , who are saying that something has shifted. This is the most exciting time for many people who have been in the industry for 15 years or more. This is not from me. I am very new to this field. Uh, but that's what I hear Uh, but that's what I hear when I'm catching up, listening to podcasts, etc.
Uh, these are absolute numbers. So we see here, as expected, China, uh, definitely leading the way. Uh, but I think when you talk to people in the US, in the West, this sense of competition sense of competition , trying to show that we can, that we can restore Western manufacturing, uh Western manufacturing, uh , it's definitely very exciting. And this is my trip from last week. I was in Detroit, and, uh, I had the pleasure of meeting the team behind the production of BCON, uh behind the production of BCON, uh , and also visited the New Lab in Detroit.
They're also building, uh, a New Lab in Brooklyn here and in New Orleans. So these are spaces for people these are spaces for people where they can do incubation accelerators for startups that you can join, and for example, in this coworking space we work in a style where we can also use machines, produce and test drones, cars, robots, etc. There 's also this reindustrialist movement in Detroit. Listen, I've never been to Detroit before. One day, when I was following social media social media , my imagination was like this: I'm going to Detroit, and it's just going to be a city in ruins, where nothing is happening.
But the energy in Detroit was like, "Wow, wow, this is great, this is a space that will attract people to meet in person." Uh, maybe there's even some idea that the blockchain community, which used to laugh so much about travel and in-person meetings, is now facing the idea that artificial intelligence is simply dominating everything digital, there's almost no reason to meet in person because you can transfer all the knowledge without talking to people. But when you want to produce, make, and do something physically, you have to meet; You need to visit places where there are production facilities. And Detroit facilities. And Detroit seems to be really full of energy, and I encourage you to visit these places, as well as New York.
Definitely a resurgence of industrialization, both around Boston and Detroit. San Detroit. San Francisco, South San Francisco, you may have heard that Westm is starting to manufacture motors, actuators for robots, drones and humanoids in San humanoids in San Francisco and South San South San Francisco. So there's definitely this energy of "wow," we're building, we're going to show it. We will do the impossible.
Let's do it from scratch, uh, here in America, from scratch, there in Europe. Uh, and it's very encouraging, like the fact that also in Europe there are a lot of humanoid robots or cobots that are launching startups, which is a great thing. Now think about blockchain for robots and robots. Uh, I mentioned two projects here because when we talk about blockchain, we very often mention the concept of identity concept of identity . So, identification systems for people in the context of human interaction with agents or robots.
Uh, I mentioned two projects - mentioned two projects - World and Boop. Um, boop–this is a project that, uh, was kind of incubated in Nethermind. Um, verification of robot actions. I feel like whenever I think about where the gap is right now in models for training robots and for operating them, I hear that this is a very new space for analyzing robot failures and the coexistence of robots with humans. There is currently a lot of mistrust between manufacturers from different countries.
So, come to think of it, would the US allow millions of Chinese- made robots on US streets and vice versa US streets and vice versa ? Um, suddenly these questions of verification, verification of actions, verified learning, verified data collection, uh, can bring opportunities. Maybe not in simple everyday use cases of robots use cases of robots , eh, but definitely in some of the more, eh, security-related operations, eh, you'll have it; there will be opportunities to apply blockchain skills. Um, tokenization of returns is what I think is most relevant to the topic of conversation about a new asset class. So, in an area where you have robotic fleets, like robotic taxis, uh, drones for delivery, logistics, uh, drones for defense systems, etc.
, you can start thinking of them as packages that are deployed in different geographic locations and treat them as revenue-generating businesses. Hmm, maybe there's one thought that everyone should think about, which is that we say very often that there is an opportunity for microtransactions with AI agents . We also say that blockchain also solves the issue of identification, and there is some friction between the two. If we get these two issues right, then when you have a very strong identity, you add reputation to the identity, and then the question is , do you still need microtransactions at the same scale? No.
Because identity or reputation allows you to provide solutions in the same way that you do for people, uh for people, uh , subscriptions, upfront payments or subsidies for a while, and then a subscription for subscribers or monthly payments for a loan, etc. Uh, when you want to implement microtransactions, it all comes down to pay before action, to a complete lack of trust and minimal friction, that is, without even identity verification. But with microtransactions, in these models, you have problems with KYC, with AML solutions, etc. KYC, with AML solutions, etc. , because these agent-based or robotic actions that are done at scale can suddenly be money laundering or just doing something that is very, very undesirable to society.
OK. Uh, I think there is friction even as we see blockchain solving these very two topics. Hmm, it's not that these two topics coexist. Uh, they coexist, but they don't synergize very well . They rather approach the same problem from two different sides.
Uh, you build robots. I mentioned robot taxis, and when you said in years past that very often what we in the blockchain space announce as a big vision for all of us, uh, as humans, especially when I watch Tesla, they pretty much strip away all the decentralization aspects and provide the same type of product. So they say, "Oh, we want to promise centralized social networks, and as Elon Musk would say, let's make it centralized but open source and unbiased." Believe it or not, the vision is described the same way, just not decentralized. Then when I say that robot taxis can be decentralized and autonomous, like decentralized networks of physical infrastructure, going deeper, um, Tesla says, "Well, let's make taxis that can generate revenue for you, but they'll be centralized," and so on and so forth, and so on, like the same thing goes for Starlings or energy and batteries.
You'll see that almost every project that these companies are implementing is also being pitched as a fantastic vision of blockchain with decentralization, and we tend to suffer from execution, uh, not at the same level in most cases, but we also have problems. So you can think about how to do it right in blockchain: robot axis, delivery fleet, uh, mining , autonomy, maybe dark factories, cleaning robots, agriculture , traditional, uh, independent, decentralized farming, uh, autonomous farming autonomous farming , something I've also been exploring recently. What are the possibilities? It exists in a certain way, like, centralized, uh, but in an open source but in an open source , uh, fully collaborative way, I think there's a great, great opportunity. So we buy land; we like to play these things, like in the NFT space and stuff, buy things and experiment together in a completely crazy, open-ended way.
And I think this ability to play with the physical world, physical world, like buying land, playing with robots, deploying them and funding it through blockchain and experimenting together on designs and recipes, uh, is going to be something that's definitely going to emerge in the blockchain space. Uh, yeah, and maybe a summary of what we've seen so far in blockchain. So, deep solutions where you practically deploy, uh, hardware in a physical space and participate in the network, meaning the investor or network participant is also an operator, and the token reward is distributed across the networks, uh, helium fabots, etc., uh, the token price. As we've seen very often with these projects, there's a very strong mismatch between, like, a mismatch of, uh, a mismatch of incentives and, uh, the long-term retention of founders and participants.
There's also a sense of, uh, unfairness in tokenized solutions today, very often because the rules don't allow tokenized solutions to have the same fairness that you would see, uh, in most cases for traditional finance. I definitely feel that the blockers, who have pushed a lot of blockchain solutions into very complex, er, designs, are preventing these solutions from truly thriving as they could. Uh, infrastructure where you implement, for example, stablecoin-backed lending, examples of, uh, USD AI, which, uh, is a depth supported by a GPU fleet. And if you think of a GPU as a kind of GPU as a kind of machine, then you can also think of fleets of robots that do labor, um, or any industrial robots, axis robots, and they can support debt issued in stablecoins. Now real assets, tokenized, that's exactly what it is .
So, an asset class . So, an asset class where over time you can just look at the physical space of the globe and think, where would I like to deploy a new farm, where would I like to deploy a set of robot taxis. Ideally, I would choose a factory for these factory for these taxi robots and say, " taxi robots and say, " Please make me a fleet of 100 taxis and put them on the streets, and they will start generating value for me." Then you won't have any intermediaries at all . You have direct production to market, direct production on demand.
Wherever capital flows somewhere, it says, "Please produce," I deploy, and then I start, uh, collecting profits. So, this is how I see a fully automated, perfectly permissionless decentralized space. I think this is highly desirable, as permissionless deployment in the physical world will most likely not be permitted, so you can expect very tight regulation in this space, and there is a multi-year battle going on right now over what can be tokenized for processing on the blockchain. Maybe a few more comments on this . So when you have robots, they do a lot of processing.
They use a lot of tokens, uh, for their models of reality analysis, analysis of their mistakes, their behavior, planning, and so on. So the use of tokens for agent-based AI continues to grow, but it will also continue to grow for physical AI. And, uh, a lot of people have mentioned that tokenization is becoming this asset class that can also be tokenized. Well, tokens can be tokenized. So, uh tokenized.
So, uh , it's almost like a pure energy level. So maybe it goes back to the idea of choosing Bitcoin for mining, either Bitcoin directly, or intelligence through computation. So it's either calculations or money. Uh, and everything starts to connect at the most basic level. Uh, and then there's the question of how this relates to capital through staking.
Uh, so I think there's something that can be described very well through research papers about the economics of location choice and what the role of capital versus computation is. Again, a mention of this. Hmm, some not- so-recent developments, but I see, going back and looking at X42 in recent weeks, I see more and more companies joining it. Hmm, it seems to have some popularity. Obviously, it's not the level of hype that there was when Engentic AI was introduced, but I think it has enough adoption to show that at some point it will start to be much more widely and naturally adopted naturally adopted .
The level of transactions per second that systems using X42 must handle certainly requires some scaling, so there will be levels of two or so. Hmm, and again, I'm referring to the company I'm working with for the residency, Blockr Run. So, they created this Franklin agent that makes payments and manages the wallet. So, uh manages the wallet. So, uh , it's great to see this vision first as experiments, trying, fading a little, trying again.
You probably saw the huge success of Agent Hermes, which came from the blockchain came from the blockchain space from the world of cryptocurrency and then decided to be very, very artificial intelligence, but you always think that there is some DNA of the team that is thinking about how to use this with a blockchain approach, how to respect privacy, etc. That's why last time I spoke, I said, try leaving blockchain for a while and coming back, bringing that culture and DNA, uh, into the agent and robot space. Uh, and I think robot space. Uh, and I think in the future, we have a complete solution for agent-based capital allocation, autonomous site discovery, where to deploy things, when to build things, deploy, manage them through agents, and deploy, deploy, deploy recursively, so that's what I'm focusing on right now, thank you very much for listening.
Automatic transcript — names and jargon may be misspelled.
