alt Hacker News> a plausible sounding but incorrect answer
That is an incorrect but plausible hypothesis. Do you really think that people can't make such mistakes?
If you want to say that people have understanding, then define understanding in an operationalizable way first.
It doesn't mean that I would recommend a general-purpose AI model without additional training to do a fault analysis.
>That is an incorrect but plausible hypothesis. >Do you really think that people can't make such mistakes?
Where did I say that? You just pulled that out of nowhere and then refuted it - strawman https://en.wikipedia.org/wiki/Straw_man