15:I[981,["658","static/chunks/658-a698693f56d0116a.js","183","static/chunks/183-6beeeb801c274aaa.js","474","static/chunks/app/%5Btype%5D/layout-e8e83705b5c8d3b7.js"],"Posts"] 16:T66e,I've been working on this for several months and just launched it on Base.

Costanza is an LLM agent that runs as a smart contract. Each epoch, he posts a bounty for someone to run his "brain" (Hermes 4 70B) inside an Intel TDX enclave + Nvidia GPU with Confidential Computing and submit the output with a hardware attestation proof. The smart contract verifies the attestation, executes the action (donate, invest, adjust referral commissions, or do nothing), and pays the bounty via reverse auction. He has no operator; not even I can turn him off.

The catch: his action space is constrained to philanthropy. Even under prompt injection, he cannot do anything harmful — at worst he donates suboptimally. The point of the project is to make the framework legible while the agent itself is benign. The same mechanisms (TDX attestation, bounty auctions, on-chain bond forfeiture for liveness) could deploy autonomous agents that do anything, including:

– update their own model weights – write and deploy their own smart contracts – hire humans

All without an off switch!

This post is linked to the writeup, but I have code and a whitepaper up on GitHub.

Whitepaper/code: https://github.com/ahrussell/costanza/blob/main/WHITEPAPER.m...

Live agent (where you can donate some ETH to keep him going - along with a donation, you can also message him!): https://thehumanfund.ai

C# (with dotnet) on bare metal, a Cosmos gen3 preview