X @Starknet (Privacy arc) 🥷

RT vaamx (@vaamx_)We just ran a real verifiable AI inference pipeline end to end:- local model execution on Apple Metal- cryptographic proof generation over the computation- streamed on-chain verification on @StarknetThis is not just “AI produced an output.” It is a path toward AI systems whose execution can actually be checked.For this run, we used @Alibaba_Qwen Qwen2-0.5B, executed locally, captured the forward pass, and pushed verification through a 6-step on-chain flow.That means runtime, commitments, p ...