newgalfix

• newgalfix 26 Sep 2025 4:10 UTC

  1 point

  0

  in reply to: Vladimir_Nesov’s comment on: OpenAI Shows Us The Money

  As a naive follow-up: let’s say GPT-6 could be trained in 3 months on a 3GW cluster. Could I instead train it in 9 months on a 1GW cluster?

• newgalfix 26 Sep 2025 4:01 UTC

  1 point

  0

  in reply to: Vladimir_Nesov’s comment on: OpenAI Shows Us The Money

  Depending on OpenAI growth, this is more of a soft upper bound on what gets built.

  I’m confused: the announcement indicates that the $400B has been committed, and is not dependent on OpenAI’s growth (although perhaps you’re implying that there’s no way they actually spend the $400B unless OpenAI revenue continues to rapidly grow)?

  Also, why would this $400B /​ 7GW be an upper bound? A recent WSJ article suggests they are planning to surpass that, although details are super light.

• newgalfix 16 Sep 2025 22:41 UTC

  1 point

  0

  in reply to: Felix Moses’s comment on: Ja­cob_Hil­ton’s Shortform

  Has Terence Tao publicly indicated that he’s shifted his main focus to applying AI to math?

• newgalfix 6 Mar 2023 1:37 UTC

  1 point

  0

  on: Models Don’t “Get Re­ward”

  Thank you for writing this post—it really improved my understanding of how RL works.