Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

You're totally right on the last point: if they stopped training models and fully focused on inference efficiency, then they could totally get to profitability (but presumably would be obsolete in the medium-term unless all progress stopped). But right now, inference is very expensive, and your wife's case is the exception not the rule.


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

Search: