Earlier quoted context omitted.
> non-trivial coding tasks I’ve come back to the idea LLMs are super search engines. If you ask it a narrow, specific question, with one answer, you may well get the answer. For the “non-trivial” questions, there always will be multiple answers, and you’ll get from the LLM all of these depending on the precise words you use to prompt it. You won’t get the best answer, and in a complex scenario requiring highly recurs…
I wish LLMs were good at search. I've tried to evaluate them many times for their quality at answering research questions for astrophysics (specifically numerical relativity). If they were good at answering questions, I'd use them in a heartbeat Without exception, every technical question I've ever asked an LLM that I know the answer to, has been substantially wrong in some fashion. This makes it just.. absolutely us…
It's funny for me to read this. They don't exhibit "confidence". You are just getting the most accurate text that it can produce. Of course, the training data doesn't contain "I don't know" for questions, that would be really bad training data! If you are getting "attitudes", it would be because you are triggering some kind of dialogue-esque data with your prompts (or the system prompt might be doing that).
Expecting the LLM to say "sorry I don't know" would be like expecting google search to return "we found some pages but deemed them wrong, so we won't show you any".