3 ms·

While I agree with the authors that large language models only trained on text lack the ability to distinguish "possible worlds" from reality, I think there is

While I agree with the authors that large language models only trained on text lack the ability to distinguish "possible worlds" from reality, I think there is a path ahead. Large language models might be excellent candidates for evolutionary methods and RL. They need to learn from solving language problems on a massive scale. But problem solving could be the medicine that cures GPT-3's fuzziness, a bit of symbolic exactness injected into the connectionist system. For example: "Evolution through Large Models" https://arxiv.org/abs/2206.08896 https://arxiv.org/abs/2206.08896 They need learning from validation to complement learning from imitation.

zwaps 4y ago
This is what ChatGPT does, for instance.
- tbalsam 4y ago
  In a sense, I thought it was more human-in-the-loop rather than an explicit RL objective (i.e. potentially somewhat limited reward surface, even if there's a reward model trained from it)