University of Helsinki free MOOC on the Ethics of AI

Re:University of Helsinki free MOOC on the Ethics of AI by ganzuul on Thursday December 24, 2020 @03:11PM (#25530978)

I have been wondering if it would be possible to collect examples of bias, the same way we collect other datasets, and teach NNs to de-bias themselves. The reason for why this is hard to decide is that bias is kind of the opposite of relevant information. The data would be patterns to avoid rather than follow.

Assembling a database for the purpose of de-biasing might also prove unfeasible because of inductive bias.

[ Parent | on HN ]

Re:University of Helsinki free MOOC on the Ethics of AI by SpicyLemonZest on Thursday December 24, 2020 @06:12PM (#25532430)

The fundamental problem is that deciding what's biased is extremely subjective and context dependent. If an AI says "crime is often a problem in lower income neighborhoods", is it delivering a statistical fact or expressing bias against the poor? Depends entirely on how we think people are going to use the results.

[ Parent | on HN ]

Re:University of Helsinki free MOOC on the Ethics of AI by ganzuul on Friday December 25, 2020 @10:47AM (#25537071)

Good point. Inductive bias takes many forms.

[ Parent | on HN ]