News for hackers. Stuff that matters.

A comment on Ask HN: SWEs how do you future-proof your career in light of LLMs?. Parent comment

Re:Ask HN: SWEs how do you future-proof your career in light... by eschaton on Monday December 16, 2024 @02:19PM (#42434314)
They shouldn’t be expected to improve in accuracy because of what they are and how they work. Contrary to what the average HackerNews seems to believe, LLMs don’t “think,” they just predict. And there’s nothing in them that will constrain their token prediction in a way that improves accuracy.
[ Parent | on HN ]
Re:Ask HN: SWEs how do you future-proof your career in light... by cozzyd on Monday December 16, 2024 @02:50PM (#42434649)
If anything, they may regress due to being trained on lower-quality input.
[ Parent | on HN ]
Re:Ask HN: SWEs how do you future-proof your career in light... by rybosworld on Monday December 16, 2024 @02:23PM (#42434353)
> Contrary to what the average HackerNews seems to believe, LLMs don’t “think,” they just predict.

Anecdotally, I can't recall ever seeing someone on HackerNews accuse LLM's of thinking. This site is probably one of the most educated corners of the internet on the topic.

> They shouldn’t be expected to improve in accuracy because of what they are and how they work.

> And there’s nothing in them that will constrain their token prediction in a way that improves accuracy.

These are both incorrect. LLM's are already quite better today than they were in 2022.

[ Parent | on HN ]
Re:Ask HN: SWEs how do you future-proof your career in light... by reshlo on Tuesday December 17, 2024 @05:15AM (#42440064)
> I can’t recall ever seeing someone on HackerNews accuse LLMs of thinking.

There are definitely people here who think LLMs are conscious.

https://news.ycombinator.com/item?id=41959163

[ Parent | on HN ]