News for hackers. Stuff that matters.
Re:Microsoft lays off an ethical AI team as it doubles down ... by jpgvm on Tuesday March 14, 2023 @05:12AM (#35148947)
Would an AI that doesn't understand racial slurs be more useful?

I don't think we should hamper the usefulness of a tool to please someones sensibilities. For most people in most of the world such words aren't relevant.

I get it, these models are made the the the US and they think these models should preserve and further their values but the rest of the world doesn't care. Most likely each of them have their own batch of words they would prefer the AI didn't understand for whatever uncomfortable reason. Reality is that censoring it on any axis won't please everyone.

Re:Microsoft lays off an ethical AI team as it doubles down ... by zirgs on Tuesday March 14, 2023 @05:46AM (#35149190)
Well, if entering "n-word" and "black person" leads to the same result then we could drop the "n-word". But yeah - censorship isn't going to stop there.