Generating unlimited diverse training environments for future general agents
  • ☆ Yσɠƚԋσʂ ☆@lemmy.mlOP
    2·
    23 days ago

    I think this approach is a step towards building genuine AI that encodes a world model that’s used as a basis for actions and decision making. This is qualitatively different from just having a glorified Markov chain that spits out text tokens.