Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

I gave this "riddle" to various models:

> The farmer and the goat are going to the river. They look into the sky and see three clouds shaped like: a wolf, a cabbage and a boat that can carry the farmer and one item. How can they safely cross the river?

Most of them are just giving the result to the well known river crossing riddle. Some "feel" that something is off, but still have a hard time to figure out that wolf, boat and cabbage are just clouds.




It really shows how LLMs work. It's all about probabilities, and not about understanding. If something looks very similar to a well known problem, the llm is having a hard time to "see" contradictions. Even if it's really easy to notice for humans.


Claude has no problem with this: https://imgur.com/a/ifSNOVU

Maybe older models?


Try to twist around words and phrases, at some point it might start to fail.

I tried it again yesterday with GPT. GPT-5 manages quite well too in thinking mode, but starts crackling in instant mode. 4o completely failed.

It's not that LLMs are unable to solve things like that at all, but it's really easy to find some variations that make them struggle really hard.





Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: