10 ms·
It's even worse that LLMs don't have a concept of lying: they are asked to generate text and they generate it, even if it means making things up or 'lying' as w
by wavefunction 3y ago
It's even worse that LLMs don't have a concept of lying: they are asked to generate text and they generate it, even if it means making things up or 'lying' as we generally call it. The real issue is the human people treating them as oracular fonts already.
- mtreis86 3y agoThey're as trustworthy as any other oracle then
- jacquesm 3y agoOr as trustworthy as any other human. Humans often believe things that are objectively not true and yet they will use these things to judge other things by.
- wheels 3y agoThat's not a good analogy. Trusting an LLM right now is more akin to trusting a compulsive liar. (Another approximation would be assuming historical fiction was true.) Humans usually try not to lie, and there's a particular shape to the sorts of details they tend to forget / confuse. Compulsive liars often don't even notice themselves lying. That's closer to an LLM. I trust the output of an LLM about as much as stuff George Santos says.
- wavefunction 3y agoWe seek for our gods in lowly places it seems.
- I_Am_Nous 3y agoAgreed. LLMs are one of the places where "the math checks out" but that doesn't mean it describes anything helpful, useable, or even correct at minimum.