3 ms·
Key sentence for me: > It might be better to say that everything GPT does is a hallucination, since a state of non-hallucination, of checking the validity of s
by einpoklum 3y ago
Key sentence for me:
> It might be better to say that everything GPT does is a hallucination, since a state of non-hallucination, of checking the validity of something against some external perception, is absent from these models.
I try to explain this to people who are obsessed with using ChatGPT to tell them things. So far I've been telling them something like: "it does not attempt to provide you valid information, it's optimizing for what would read like a reasonable continuation of the conversation, which is really not the same thing."
- TeMPOraL 3y ago> "it does not attempt to provide you valid information, it's optimizing for what would read like a reasonable continuation of the conversation, which is really not the same thing." No, but it's quite close - because "reasonable" is positively correlated with "valid, correct information".
- d0mine 3y agoThe same can be said about our brains too. What we think we see is quite different from what our eyes register (e.g., blind spot, hollow-face, colors on contrast backgrounds, etcchttps://m.youtube.com/watch?v=mf5otGNbkuc https://m.youtube.com/watch?v=mf5otGNbkuc
- pixl97 3y agoAlso one to add to your watch list. PBS: Perception Deception https://www.youtube.com/watch?v=HU6LfXNeQM4 https://www.youtube.com/watch?v=HU6LfXNeQM4