Comment Comment by zenlikethat Nothing, but imagine using LLMs for a classification taskPeople out there are so resigned to the models being unreliable that they are really doing things like hallucinating deliberately, and then matching the hallucinations to embeddings -https://softwaredoug.com/blog/2026/08/10/hypothetical-classi...You could do that or you could just... use a model that will never produce unreliable outputs in the first place. parent zenlikethat 1 month ago
People out there are so resigned to the models being unreliable that they are really doing things like hallucinating deliberately, and then matching the hallucinations to embeddings -
https://softwaredoug.com/blog/2026/08/10/hypothetical-classi...
You could do that or you could just... use a model that will never produce unreliable outputs in the first place.