logoalt Hacker News

perching_aixtoday at 12:09 PM0 repliesview on HN

It generally helps when one is not surrounded by tactless buffoons.

I had this happen to me a few times, kindly produced my own LLM output screenshots in response, and the issue resolved itself. I was lucky: I got the kind who - mistakenly - thought they were being helpful. They weren't, got the hint, and buggered off with this. I wasn't really asking them questions though per se, so maybe a bit of a different situation.

Maybe worth trying if you have not. Obviously, if you have a hard-on against LLMs this won't be easy though.

Though I will say, some colleagues of mine are visibly absolutely terrible in using LLMs, so with them it does make sense to prompt on their behalf. Definitely wouldn't lead with the LLM output like this though, not the least because it's always a mountain of prose.