News for hackers. Stuff that matters.
Re:GPT-5 by noosphr on Thursday August 07, 2025 @05:08PM (#44830384)
>This looks like they're not training the single big model but instead have gone off to develop special sub models and attempt to gloss over them with yet another model. That's what you resort to only when doing the end-to-end training has become too expensive for you.

The corollary to the bitter lesson strikes again: any hand crafted system will out perform any general system for the same budget by a wide margin.

Re:GPT-5 by fidotron on Friday August 08, 2025 @07:18AM (#44835731)
That is, at best, wishful thinking.

In practice the whole point is the opposite is the case, which is why this direction by OpenAI is a suspicious indicator.