Skip to content

The CCS Interview

XYO's Markus Levin: Robots should be owned by people, not just Tesla

Markus Levin, co-founder of XYO Network, talks about how verifiable, trustless data could underpin AI and robotics, and why he thinks robots and cars should be owned by individuals rather than a few corporations. The conversation, recorded in February 2026, also covers weather data, privacy and the role of the XYO Layer 1 network (XL1) as Levin described them at the time.

Interview by Ashton Addison Recorded February 2026 6 min read

Key takeaways

  • XYO Network functions as a reality engine that proves location and validates data from IoT devices, mobile phones, and sensors to connect the physical world with blockchain.
  • Decentralized autonomous systems allow individual owners to earn money from their robots and vehicles rather than having centralized corporations capture that value.
  • Trustless data verification through cryptographic proof is essential infrastructure for AI and robotics operating in the physical world to ensure safety and reliability.
  • Decentralized physical infrastructure networks enable peer-to-peer coordination and payment for autonomous work without intermediaries taking a percentage of earnings.
  • Current centralized systems like Tesla's autonomous network and healthcare insurance extract significant value through intermediaries, whereas blockchain-based alternatives could reduce overhead costs.

From proof of location to a reality engine

Ashton Addison: For anyone who missed earlier episodes, what is XYO's vision and how has the network grown?▶ 00:25

Markus Levin: XYO is a reality engine connecting the real world with blockchain. We started with proof of location, proving the location of devices or people. From there we moved to data: collecting it, proving it is true and correct, and connecting it with people on the Web2 and Web3 side. We began with IoT devices like Bluetooth beacons and GPS trackers, and today we collect a lot of data through mobile phones and other ways too. That's useful for smart cities, self-driving cars and drones, and for AI companies and others that need lots of data.

You are the sucker if you let the big corporations take some of your money while you could earn more doing it yourself.

Who should own the robots and cars?

Ashton: XYO started in 2018 as a decentralized physical infrastructure network, before the AI and robotics boom. How does that vision fit with autonomous vehicles and robots?▶ 02:06

Markus: A robot is two things: a node that witnesses things, and something that does work someone gets paid for. In DePIN, you put up devices and get paid. With Helium you create a cell network. In our case you put up our devices, collect and validate data, and earn XYO. There are many others too, for AI compute, data storage and energy. The way Elon Musk envisions robots, they're owned by Tesla, part of a centralized company. We're talking about decentralized physical AI, where robots are owned by you and me and many other people, and you can participate in the value they create. In America, cars are used about 3% of the time. If your car were self-driving and could go somewhere to charge itself, you could use it 90 or 95% of the time, and it would earn money while you're not using it. It's the same with robots: send yours to do garden work for your neighbor and get paid for it. That's where we fit in. We connect all these devices, prove that something is where it's supposed to be, and witness what's happening. It's proof of work for the robot: did it really do the garden work? The data gets written on chain, and payment rails pay you or your robot automatically.

Ashton: Tesla owners may soon be able to switch on a robotaxi function. What do you miss out on by not putting your car on a decentralized network?▶ 05:31

Markus: Someone else earns off you. It's the middleman we try to avoid in blockchain. Everything should be peer-to-peer and borderless, and we should be able to interact without intermediaries, because intermediaries cost money and take value away from you. Look at US healthcare. Of the premium you pay, only 6% goes into care, and 94% is wasted in the system. Imagine a peer-to-peer insurance company run on blockchain. The same goes for a fleet of cars or a team of robots run in a decentralized, open-source way. You can plug your robot into a platform, take it out and put it into another one. I think that's a fairer world. Humans are profit-seeking, and you are the sucker if you let the big corporations take some of your money while you could earn more doing it yourself. It's about control and being able to monetize better.

Ashton: Centralized providers already have network effects. How can decentralized networks compete and grow?▶ 10:57

Markus: Centralized networks are often quite efficient, so you have to make onboarding very easy. Crypto often fails at that, especially in DePIN. You need to buy a device, install something on your computer, connect it to your Wi-Fi, then create a wallet and worry about losing your keys. You lose people at every step. The best way is a website or app where you download it or log in and everything happens automatically, with preset settings. It runs on crypto rails because that's more efficient, but that's in the background. On cost, there's so much excess capacity, like the car I mentioned, or your computer, which could do data storage or compute for AI training. Many people don't do it because it's complex, but I think these networks will grow over time and be very successful.

Verified data against hallucinations and forecasting gaps

Ashton: Could AI agents tie into XYO's data verification the way physical robots would?▶ 13:49

Markus: AI agents complement AIs by doing specialized functions, including data collection and validation. An AI might get feedback that it hallucinates, and an agent might go and collect another piece of data. In a DePIN network like ours, that agent can be an AI or a human. Say you want to know whether it's raining somewhere. The request comes into the network, and someone takes a picture of the clouds and says it's sunny. That data is validated with proof of location, and with what we call proof of origin we can prove where it was generated, on the phone for example. It's saved into an IPFS database, put into a smart contract and run through the XL1 network, so we can say where it originated. That obviously helps against AI hallucination.

Ashton: Weather was one of the earliest DePIN use cases. Where does verified data help most?▶ 15:51

Markus: We have a partner called WeatherXM, and it's a great use case. In a lot of places you don't have the temperature and other sensors you need for good forecasting, which matters for industries like farming and electricity generation. In the US that's not a problem. But in parts of Southeast Asia, in Africa or on remote islands, you don't have proper forecasting, and without it you might make the wrong decision about when to sow seeds. With DePIN you can put up temperature sensors, or take pictures of the sky and the ground, and feed all of that into a farming AI. Even if only 3% of the farmers in a region participate, you get a wider net. The AI can then give farming advice, and you increase your yield. That's DePIN really changing the world.

The Layer 1 creates immutable data that can't be changed later, so no one can say the robot didn't do that work.

Privacy, the Layer 1 and what comes next

Ashton: As robots move into homes, what are the threats around unverifiable data and hacking, and where does DePIN fit?▶ 21:35

Markus: There are definitely threats. A few years ago there was a scandal where Tesla was recording people in their garages, and that footage was shared inside the company. That's the problem when big corporations have visibility and decide what's private and what isn't. It feels like you've lost your right to privacy. With blockchain you can build things like zero-knowledge proofs, where you can know that data is true and correct without tying it to an identity. DePIN works there, because a node can sit somewhere without the data being tied to an identity. That's very important, especially as robots might be in our bedrooms. Take elder care: there's a shortage of people for it. If robots can help the wonderful nurses and helpers, they might feel less stretched and can focus on what matters, like the emotional side. But you want to make sure you have your privacy.

Ashton: Why does verifiable data need its own Layer 1 rather than using Ethereum or another chain with a bigger network effect?▶ 25:46

Markus: Our chain is built specifically for data. Ethereum and others are made for value transfers, where you send one ETH to someone else. You can do that on ours too, but it's made for data certainty. You can prove the data was generated on a particular robot, say one used for elder care. You can add a zero-knowledge proof so no one can identify the person, then store the data elsewhere and hash it on the Layer 1. The Layer 1 creates immutable data that can't be changed later, so no one can say the robot didn't do that work and so you don't need to pay. It also helps ensure there are no hallucinations. It connects the real world, this reality engine I mentioned, with AI and other virtual worlds.

Ashton: How do you see AI and crypto merging, and what could drive adoption of verifiable data networks beyond the blockchain echo chamber?▶ 31:04

Markus: AI is hungry for three things: data storage, compute and data. It gets them any way it can, and right now those systems don't care whether they're decentralized or centralized. I think blockchain can be an awesome part of serving AI. In the future, our Layer 1 and other technologies can make sure the AI is doing what it's supposed to do and doesn't hallucinate. A lot is coming in the next few years. Quantum computing, for example: they say the first commercial machines arrive in 2027, and that will accelerate AI and its capabilities. We'll need even more data, and we'll get better robots, even space-faring robots, doing healthcare and maybe helping solve diseases. For all of that you still need compute, data storage and data, and blockchain fits very nicely into that.

This interview has been edited for length and clarity. Recorded February 2026.

Markus Levin
Markus LevinMarkus Levin is a co-founder of XYO, a decentralized physical infrastructure network (DePIN) that collects and verifies real-world data such as location through a network of consumer devices. He works alongside co-founder and CEO Arie Trouw and has long served as the project's main public… Full profile →
Share this interviewXLinkedIn
▶ Watching the interview