Enacting contestation of AI was quite inspiring, too. I wonder if AI will ever enjoy counterfactual trust. I guess there is only one technology so reliable, it doesn't even need trust, abd thats telling machines...
Discussing bias is always duscussing proxy and population subjects - which can be thought provoking. But since the debate seems to revolve around that in the end, the issue is stuck deeply in capitalism (whatever that is rn(, we might consider regulate AI not by dedicated contrstation bjt occasional contestation. Your get sued by a rejected applicant for sexism? Treat it as a human case, investigate, contest, let consequences follow. Within status quo, this might add up.