ben hayum(Ben Hayum)

• ben hayum 10 Apr 2023 20:22 UTC

  2 points

  1

  on: Agen­tized LLMs will change the al­ign­ment landscape

  Ultimately I think this leads to the necessity of very strong global monitoring, including breaking all encryption, to prevent hostile AGI behavior.

  This may be the case but I think there are other possible solutions and propose some early ideas of what they might look like in: https://​​www.lesswrong.com/​​posts/​​5nfHFRC4RZ6S2zQyb/​​risks-from-gpt-4-byproduct-of-recursively-optimizing-ais

• ben hayum 7 Apr 2023 18:58 UTC

  1 point

  1

  in reply to: tailcalled’s comment on: Risks from GPT-4 Byproduct of Re­cur­sively Op­ti­miz­ing AIs

  I can’t convince you to continue or stop, but maybe reading the edit I made to the start of the post will better clarify the risks for you.

• ben hayum 7 Apr 2023 18:50 UTC

  1 point

  0

  in reply to: the gears to ascension’s comment on: Risks from GPT-4 Byproduct of Re­cur­sively Op­ti­miz­ing AIs

  1. I agree, it isn’t the worst heuristic ever. But at the end of the day is still a heuristic.

  2. Yes, you’re right. I tried to phrase things kindly but perhaps I didn’t go far enough. Made some edits :)

Risks from GPT-4 Byproduct of Re­cur­sively Op­ti­miz­ing AIs