Hacker News new | ask | show | jobs
by iamkoch 796 days ago
LLMs cannot lie insofar as they cannot tell the truth. They're remarkably good at predicting what token comes next given a bunch of tokens, but nothing else.
1 comments

Yes, but it's also generative, so at each time step it will be basing those predictions off of its own recent behavior so it's also chaotically, unpredictably performant in the quality of its predictions, but nothing else.