DonJefe19
Don Juan
- Joined
- May 25, 2025
- Messages
- 60
- Reaction score
- 36
- Age
- 40
LLMs’ “simulated reasoning” abilities are a “brittle mirage,” researchers find
Chain-of-thought AI "degrades significantly" when asked to generalize beyond training.
arstechnica.com
It's one of the reasons why I find it a bit ridiculous that people will refer to ChatGPT's prompt responses as another person in a conversation. Even the idea of it being used for therapy sessions seems stupid. It's one of the reasons why I don't trust it.
