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

A truly Bayesian machine never eliminates the old priors. It updates them, but they remain in there. If somebody starts behind, they stay behind, forever.

And while there are some who are fair-minded, many are not. So such a person is in a lose-lose situation. Those with bias will discriminate against them -- and "fair minded individuals" will use take their poor relative performance to inform their prior. Try to compensate for that, and those fair minded individuals use that, too, against them.

So "fair-mindedness" leads us to perpetuate discrimination. Which doesn't seem very fair.




Join us for AI Startup School this June 16-17 in San Francisco!

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

Search: