(shchegrikovich.substack.com)

379 points shchegrikovich | 3 comments | 13 Oct 24 21:22 UTC | HN request time: 0.694s | source

1. johnisgood ◴[18 Oct 24 09:05 UTC] No.41877585[source]▶

ChatGPT seems to do awful with Prolog. Do you guys have any experiences with using an LLM to write Prolog?

2. tannhaeuser ◴[18 Oct 24 10:07 UTC] No.41877867[source]▶

See my top-level comment for an example. Though different, there's also casual experimentation described in [1] (and additional posts linking to academic research on the SWI Prolog forum).

Do you mind sharing your experience with ChatGPT (which version)?

[1]: https://swi-prolog.discourse.group/t/chatgpt-prompts-prolog-...

replies(1): >>41878273 #

3. johnisgood ◴[18 Oct 24 11:09 UTC] No.41878273[source]▶

>>41877867 #

I do not think that I have the conversation / chat anymore, but I was trying to get it to make a monthly schedule for N workers with specific constraints[1]. I thought Prolog is suitable for this (is it?), but the generated code got stuck / hung up.

[1] https://news.ycombinator.com/item?id=41756679

↑

Use Prolog to improve LLM's reasoning