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

It would only move the drama from the real time judging to the determination of the training data set.

I see it playing out in a different way. The AI judge would never be officially acknowledged. Instead, it would be this dark judge that people turn to when their favorite dog is misjudged. “See, the computer said he was the winner!”. It would be a while before this group has the sophistication to blame the data set.

Maybe having 3 human judge and one AI judge would be something acceptable from the public. That is an interesting discussion anyway.

That's not necessarily a bad thing. It could lead to more people learning and understanding machine learning models, and improvements from the constructive criticism.

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