
GottaStopScrolling
4.4K posts





On-chain AI inference was thought to be impossible. Today I'm publishing the proof that it's not. arxiv.org/submit/7405003… Thread on what we proved and why it matters. 👇

The future of AI is Private $ARC 0x672fdba7055bddfa8fd6bd45b1455ce5eb97f499

On-chain AI inference was thought to be impossible. Today I'm publishing the proof that it's not. arxiv.org/submit/7405003… Thread on what we proved and why it matters. 👇

On-chain AI inference was thought to be impossible. Today I'm publishing the proof that it's not. arxiv.org/submit/7405003… Thread on what we proved and why it matters. 👇

We just achieved deterministic on-chain inference with multi-node consensus. 4 nodes across 3 continents independently run Llama-2-7B and produce bitwise identical outputs. Every inference is recorded on-chain through DAG consensus. Same model. Same input. Same output. Every node. Every time.














