Re:LLMs are terrible at designing AI systems
by uain on Tuesday November 17, 2026 @03:51AM
(#50529047)
How did you distinguish contradictions from facts changing over time? An old address and a new address can both be valid records. I'd want examples of those cases in the evaluation before deciding whether the checker should be symbolic or model-based.
Re:LLMs are terrible at designing AI systems
by calf on Tuesday November 17, 2026 @05:38AM
(#50529724)
Just thinking aloud but isn't that equal to detecting a self-contradictory sentence and thus equal to solving the halting problem thus computationally undecidable in general (& in principle, etc.)?
Are we heading towards a future where LLMs learn from their own messy output and make code quality progressively worse?