Hacker News new | ask | show | jobs
by bleuchase 1188 days ago
I’m sure someone is working on adding support for human-like errors in LLM outputs.