News for hackers. Stuff that matters.

massicerro

Of course, the 'presumption' of increased complexity or the 'subjective perception' of a drop in performance might be the cause. But we are missing the real point here: the 'false plateau.' Regardless of user perception, is it possible that a 'false plateau' exists that keeps us away from a major leap in performance? The risk is that the simple 'perception of having taken the wrong path' by researchers or companies would lead them to ignore the possibility of such a 'false plateau'...

Ask HN: What if the AI scaling plateau is just a "false dip"?

First of all, I’m Italian, and since I don’t feel confident enough to write this post in English myself, I used Gemini to translate my thoughts into the text below.The Premise: There has been a lot of talk lately about the possibility that AI development (as we currently know it) is approaching a plateau. While I don't personally agree with this hypothesis, it is undeniably a common...

( Read More... | 3 comments | on HN )