A lot of weird things in this comment, like implying that it is OK to use llms to search for surface information. That is just as biased as searching for more complicated info, and the tone of the llm response white-washes it all away.
If you ask any big LLM a mundane question, it’s likely to get it right. And as you put it, it’s one of the way people get sucked in. But it does work, very inefficiently, for that purpose. Saying it does not is lying to yourself.
A lot of weird things in this comment, like implying that it is OK to use llms to search for surface information. That is just as biased as searching for more complicated info, and the tone of the llm response white-washes it all away.
I don’t know where you think I implied it was ok for surface search, but it’s the exact opposite of what I wanted to say.
This implies that it is reliable for “surface informations” which i disagree with, though a lot of people get sucked in that way
If you ask any big LLM a mundane question, it’s likely to get it right. And as you put it, it’s one of the way people get sucked in. But it does work, very inefficiently, for that purpose. Saying it does not is lying to yourself.
I disagree. A big LLM might get a mundane question right, but it might not. To call it reliable is laughable.
What it is is confident, and if you don’t know the answer yourself, it will seem plausible.