News for hackers. Stuff that matters.
Re:The Myth of AI Omniscience: AI's Epistemological Limits by visarga on Saturday August 05, 2023 @12:54PM (#37013979)
You are not fundamentally wrong, but there are degrees. LLMs can do some symbolic manipulation and a few reasoning steps. Maybe they are just the result of pattern matching, but I have a hunch that most of the time humans do the same. We fake our understanding as much as we can, we use all shortcuts.

There are no comments on this comment yet.