7 ms·

I'm not surprised. People really thought the models just kept getting better and better?

• guerrilla 1y ago
  Maybe. How would I know?
• jMyles 1y ago
  ...even if the agent did "cheat", I think that having the capacity to figure out that it was being evaluated, find the repo containing the logic of that evaluation, and find the expected solution to the problem it faced... is "better" than anything that the models were able to do a couple years ago.
• segmondy 1y ago
  The models are getting better and better.
  • giveita 1y ago
    That's expected. No one will release a worse model.
    • sodality2 1y ago
      Not a cheaper one, or better in some ways, or lower latency, etc?
      • giveita 1y ago
        They do that too but right now it is an arms race as well.