Former Deepmind researcher thinks AlphaGo's playbook can cure AI hallucinations

Illustration showing the Metis octahedron in a dotted circular path with an arrow and a yellow checkmark; below it, a Go board tilted at an angle with white and blank stones, with a dotted line pointing toward it. Former Deepmind researcher Thore Graepel wants his startup Metis Reasoning to bring AlphaGo's search principle to general reasoning. Instead of locking knowledge inside model weights, an explicit knowledge ledger would make every conclusion checkable against evidence. So far, though, there's no working system and no benchmarks. The article Former Deepmind researcher thinks AlphaGo's playbook can cure AI hallucinations appeared first on The Decoder .