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

Setting aside the AI angle, perhaps just recording the assumptions in a way that can be measured would be enough.

Tooling, runtime sampling, or just code review could reveal when the assumptions go awry.




Consider applying for YC's Spring batch! Applications are open till Feb 11.

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

Search: