AI safety via debate

8 years ago 10
Add to circle
We’re proposing an AI safety technique which trains agents to debate topics with one another, using a human to judge who wins.
Read Entire Article
  1. Homepage
  2. Technology
  3. AI safety via debate

Related

xAI Is Dead. Long Live SpaceXAI

xAI Is Dead. Long Live SpaceXAI

20 minutes ago 0
Syntiant, which develops low-power AI processors, files for a US IPO, reporting a $20.9M net loss on $64.5M in revenue for the three months ended March 31 (Pragyan Kalita/Reuters)

Syntiant, which develops low-power AI processors, files for ...

30 minutes ago 0
Ternlight – 7 MB embedding model that runs in browser (WASM)

Ternlight – 7 MB embedding model that runs in browser (WASM)...

58 minutes ago 0