News for hackers. Stuff that matters.
Re:The Program 2025 annual review: How much money does an au... by didgeoridoo on Thursday December 25, 2025 @07:21PM (#46387996)
LLMs have no knowledge (really “knowledge-like weights and biases”) outside their training set and system prompt. That plausible justification is just that — a bunch of words that make sense when strung together. Whether you’d like to give that any epistemic weight is up to you.
Re:The Program 2025 annual review: How much money does an au... by eschaton on Friday December 26, 2025 @10:07PM (#46398706)
More like whether one is correct in giving it any epistemic weight. Not everything is opinion, some things really are clearly right and clearly wrong and attributing thought, reasoning, and analysis to an LLM is one of those things that’s clearly wrong.