Metis Reasoning Aims to Transform AI with AlphaGo's Principles

Thore Graepel, a former DeepMind researcher, aims to apply AlphaGo's principles through his startup, Metis Reasoning. Instead of confining knowledge within model weights, an explicit knowledge ledger would allow every conclusion to be verified against evidence. Currently, there is no functioning system or established benchmarks for this innovative approach. Graepel's goal is to tackle AI hallucinations by providing a more transparent method for information verification. This innovation could transform how we interact with artificial intelligence.