Hacker News new | past | comments | ask | show | jobs | submit login

Also easy to recognize that it was a typo and interpret it as intended.



Indeed even gpt-3.5-turbo recognizes the typo and gives a proper explanation. Can't wait until everyone has an LLM baked into their device that makes submitting stupid content a two step verification.




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

Search: