Language models are excellent at producing plausible continuations. Agents need something harder: a way to anticipate what happens next.
Language models are excellent at producing plausible continuations. Agents need something harder: a way to anticipate what happens next.