Quote:
Originally Posted by Quoth
LLM doesn't "work".
|
This is the vital point that many commentators are missing. I've spent some time asking ChatGPT questions to which I know the answers, and almost always it returns something that sounds convincing and plausible, but is just plain wrong (often ludicrously so).
The danger is that most people asking ChatGPT something will not know the answer (that's why they're asking), so will accept the response as truth.