Hacker News new | ask | show | jobs
by kakuri 257 days ago
Keep in mind that many (most? all?) LLMs are not trained strictly on factual data, and are probably not trained to differentiate between factual and non-factual information. If you ask an absurd question you will likely get a (delightfully) absurd answer. If you ask a question somewhere in the borders between reality and fiction... results may vary.