Quote:
Originally Posted by Crowbob
Yes. You are. I’ve decided.
Per the article, Google banned two developers because they published a study that was in conflict with the notion that, if the goal is to use AI to perform as or to predict human behavior, bias should not be eliminated from AI, that looking for bias in the algorithms to weed it out is counter productive.
IMO, I think minimizing bias to almost zero makes AI much more dangerous than simply tolerating it and excising the extremes.
I think at some point AI is gonna get out of the lab like COVID did. Such a thing is scary of it doesn’t have some kind of ‘feelings’ one way or the other.
|
I don't think one has to be biased in order to be aware of and even understand bias.
To say a machine has to be biased in order to detect it and predict behavior is illogical.
Best
Les