Ethereum X Robotics

Transcript

[music] [music] All right, then. Are we Are we good to go? >> Okay. Hey, everyone. I'm Arnov and >> I'm Thomas. >> Hey. So, I in November 2024 at Defcon, you spoke about the intersection of Ethereum and robotics. You laid out a vision of a world where autonomous machines could coordinate transact and establish trust with one another with Ethereum serving as the founding uh as the underlying infra. Two years later you've left Ethereum and you've started complete robots. So you've gone from thinking

about this and uh and you've gone from the side of the this this Ethereum infra work that you've been doing for this last decade to now building in robotics. What has that transition been like and how has that like changed your thinking? So we can even go back to 2017 or even earlier when when I was first time learning about Ethereum and some of the most exciting use cases were uh the aspects of the vision of robo taxes that pay for themselves, Airbnb that pays for itself that when you have those decentralized payment system then um all the machines will be able to to

autonomously maintain themselves and then you really remove the intermediary in the interaction between the person and the machine. [snorts] So they can think of oh I produce a robot axi as a factory but someone already invested in it and it's almost like a RWA a tokenized asset that is deployed on the road and I just order it and allocate it as a capital allocator. So that was exciting but it was for many many years not really possible. people were experimenting in the early ethereum days with robots with the ROS on solidity contracts and so on and so on. But only only now with the emergence of AI uh you really can start doing that. So we have ability for the smart contracts to start

adapting to the situation based on coordination of the models. So you asked me specifically about the transition. Um so I see that as continuation in 2024 I think I was gradually resteering Nez minds towards AI and robotics and hardware. Uh but then I took that leap into Ethereum Foundation leadership and I didn't think that pushing Ethereum Foundation too much towards AI and robotics would be healthy because um this felt like very specific forwardlooking uh niche idea that some people will fiercely following but uh but for the general public there were many other

problems. So uh so now just after leaving foundation I feel okay this is continuation not atom starting it as a new company it's it's actually a subsidiary of mind so so it's still continuation within within the broader concept and I love it I'm learning a lot uh once again I'm approaching it start something and learn from from zero in accelerated pace when I was starting mind it was learning cryptography learning ethereum uh implementing everything from scratch to to build it and now it's very similar approach for me.

>> When you were building Nethermine, this was a world where I believe you built the the first version of the client entirely yourself, right? And now >> there was like uh two or three people involved. Uh definitely majority of the of the codebase for the client was written by me. So over the first one or two years or even three years, this was maybe 90 95% of the codebase. uh but it was without prior knowledge about cryptography about distributed systems. So lots of learning. I was building large enterprise systems before at banks. So I I knew the financial aspects of it. I had some ideas of what will be coming to Ethereum in the concept of in the context of MEV. Um so I had I had expectations of how the markets will be

evolving. But on the technical side, it was lots of learning of those low-level optimizations and managing the application that is not strictly enterprise. And then some fantastic developers started coming and fixing many things that uh you know I I couldn't really reach to the level of what we see nowadays in mind implementation. uh but but the general architecture I think that on the other hand few companies that were building client implementation could benefit from that uh structure or feel of someone building the enterprise system that is at least somewhat focused on performance because I was building systems for the FX trading >> and what is the focus at complete robots

been like? Yeah. So the the vision generally is that there will be a coordination layer for the global physical AI and if you go outside of just blockchain and financial system on blockchain you think that the entire world gradually uh like changes into a machine. Like when you when you generalize everything and you think okay so world is is humans and a lot of machines the machines are disconnected but as we have growing complexity of the systems then you can start looking an arbitrary complex system as almost like as a single machine with components. So is a city uh a lot of totally independent things or is a city really a single machine and I

think that over time we'll see city more and more as a like a prefab. So now we start talking about the pre-fabricated data centers. People ship them in containers but in the future you can maybe deploy the entire village and connects a bit of ideas from the pop-up villages of the data centers, infrastructure deployments, robotics at which stage like you you talk about robot and which stage you talk about the building. Is it just about moving but what about moving parts inside? Uh so we focus on something that is nowadays very much bringing a lot of revenue opportunities and a large margin to accelerate the business. So focusing on first data center deployments and design

partners for robots deployments. But only in October I got my robotics team joining and opening the lab and starting to build while I I still have the mixture of experimenting myself, coordinating and so on. So data centers first but also some of the distribution of buying and selling robots uh connecting different parts of the world. So we have advantage of my story of being very very global traveling a lot so able to connect with the suppliers in China collecting uh with the buyers in Europe in the US and so on.

A lot of the attendees at Defcon are students at the beginning of their careers and as they're watching this crazy like AI buildout happen, people are wondering like how to just approach the world that they've been dealt with. What do you what would you be doing if you were just starting out? Let's say what would your advice be to like folks in this room? Um I think there's a like over and over again I'm think about what the message is for the new generations you know like uh I I do have uh like many years of career behind me.

So it's a story where I think okay I had some plans I I built that there is some sense of fulfillment I still want to chase it and I feel a bit of scared as a founder how do I build with the world with AI and with automation and then then I start thinking what would it be if I was G&Z person joining the market. So either maybe for for them it's much more natural to feel it and they'll build it and they'll surprise us of of how you define yourself in this world with with AI but we also see that um there is they feel a bit lost right like at the universities they start protesting they start asking questions uh they start complaining and I I understand what the message is I think

that there are some ideas only but we cannot be really sure whether it will work I think that maybe going to something hyper local uh focusing on the on the friends, family, local communities and thinking never before you could learn so quickly things like um specialized medicine uh facuing like the specialized devices for agriculture that are local. Second thing what my local community needs to to improve and particularly I think India being a country of so many engineers and this massive culture of engineering and many families everyone can think oh I'm going back and I have access to all of

the AI and explanation I can focus maybe on improving efficiency and health uh healthcare and the logistics within my community. So maybe maybe that return that you don't have to really look at the whole world as something to solve and and go to universities but instead you can immediately turn into some productive uh a really impactful things for for the closest people to you. What are some like problems that you're uh dealing with at complete robots that might interest some some people watching? What are I guess some maybe some roles that have been like really hard to like hire for? What are just some domains of expertise where you just

been looking for? Um I think this obviously because the the vision is very broad. So I think it's consistently inspiring for me every day to discover things related to to power engineering to the novelties around the uh the things with the data centers with robotics with the world models with AI LLMs uh the feeling of ability to manage workflows as as I work now with you know lots lots of agents running uh together. So but of all of this I think just going back to the basics of physics and learning from the beginning of uh all the power engineering electrical

electromagnetics that that was the most fun for me and I think that's where I recommend everyone if you if you have gaps there as an engineer start moving back from software to actual physical engineering learn HVAC learn uh civil engineering learn electromagnetics and so on and it will be probably super not only super useful but also addictively interesting. >> I guess this has been your journey in the last two years. >> Yes, last last few months particularly uh maybe the year before it was mostly strategy and coordination within the foundation. So, [music]