Hacker News (curated)new | past | comments | ask | show | jobs| show hidden

Why can't a next token prediction machine not predict a train of reasoning?




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact | github