The Decoder· Jonathan Kemper·· 4 h ago
Former Deepmind researcher thinks AlphaGo's playbook can cure AI hallucinations
Former Deepmind researcher thinks AlphaGo's playbook can cure AI hallucinations
AI summary
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.
Selection record
Threshold 76Media and individualsFirst 47Second 38
Not admittedSum of both 85 < twice the threshold 152
- Source tier
- Media and individuals; this tier's threshold is 76
- Pre-filter
- passed：前DeepMind研究员提出用AlphaGo思路减少AI幻觉
A model scores each item twice, independently, against one written standard, out of 100. An item is admitted only when the two scores add up to twice the threshold. The threshold is set per source tier.
Source: The Decoder · the-decoder.com