A macro investor who trades sovereign debt just made a bet that AI agents will eat tokenized assets for breakfast.
The Summary
- Veteran investor Jordi Visser is buying Ether, betting that AI agents will drive demand for tokenized real-world assets on Ethereum.
- Visser's thesis: AI agents need "food," and that food is tokens, not compute or data, but programmable value units they can transact with autonomously.
- This convergence could reshape financial systems, making Ethereum the settlement layer for machine-to-machine commerce, but may concentrate wealth among early crypto holders.
The Signal
Jordi Visser doesn't chase memecoins. He's a macro investor who built his career reading central bank tea leaves and trading government bonds. When someone with that background starts accumulating ETH, it's not because he believes in decentralization philosophy. It's because he sees the plumbing of finance changing.
His bet is simple: as AI agents proliferate, they'll need a way to transact value autonomously. Not through human-mediated bank transfers or credit card rails. They need tokens. Programmable, divisible, instantly settleable units of value that can move between agents without asking permission from a payments processor.
"AI agents need food, and that food is tokens."
This isn't speculative. We're already seeing it. AI agents managing DeFi positions, routing transactions, optimizing yield. The next step is agents that hold tokenized real-world assets: shares of companies, fractions of real estate, commodities, bonds. Visser sees Ethereum becoming the settlement layer for this machine economy because it's the only blockchain with the liquidity, developer tooling, and institutional adoption to handle tokenized versions of traditional assets at scale.
The convergence matters because it solves both sides of a chicken-and-egg problem. Tokenization advocates have been waiting for a killer use case beyond "faster settlement." AI agents are that use case. They don't care about 24/7 markets for philosophical reasons. They care because they're always on, and they need rails that match their uptime. Meanwhile, AI builders have been wrestling with payment infrastructure. How does an agent pay for API calls, data, compute? Tokens solve that without requiring each agent to have a bank account and compliance officer.
But there's a darker edge to this. If AI agents drive tokenization demand and Ethereum becomes the settlement layer, the people who already hold ETH and tokenized assets will capture enormous value. The wealth concentration in crypto, already severe, could calcify as agents transact primarily with other agents, compounding returns for whoever got in early. The machine economy might be efficient, but it won't be equitable unless we design it that way.
The Implication
Watch where institutional money starts building tokenization infrastructure. If Visser is right, the next 18 months will see a wave of real-world assets moving on-chain, not because regulators blessed it, but because AI agents made it economically necessary. For builders: the tooling layer between agents and tokenized assets is wide open. For everyone else: this is the moment to understand how agents will reshape markets, because by the time it's obvious, the positions will already be taken.