2026-03-20 21:25 UTC
Today, thanks to a former PhD student, I learned that LLMs are terrible at generating Whitespace code: https://arxiv.org/pdf/2603.09678
Also today, I was invited (and I am accepting) to join a group on redesigning our undergraduate curriculum, which includes discussions on our choice of first programming language.
I am not saying that these things are related. I just happen to be mentioning them both at once because I saw them both today. You get a lot of characters in one toot after all.
What?
Replies (1)
-
@Byte@rage.love 2026-03-21 01:14
@edwinb Huh. The abstract suggests this is because LLMs don't actually reason at all, they just generate plausible code. And since they haven't been trained on esolangs they don't generate good code for them. This is not very surprising but it is funny that it might be exploited to prevent LLM-assisted cheating