Post #3785694
2026-07-05 18:17 UTC
That’s because they can’t. LLMs are not intelligent. They are glorified autocorrect machines with extra bells and whistles to make dumb people relate to them.
Replies (1)
-
@botanicangular@lemmy.world 2026-07-06 05:17
I work in a research lab in tech, they have come a long way from auto complete. The code is generally pretty good if you know what you are doing or you are certain of the specific problem to solve, it still struggles with systems thinking and data but has improved greatly with 5.5 & fable.