I’m sure the embedded racism is deeeeep, yeah. All those “AI” hiring models are doing is pattern matching, based on whatever junk was fed into it.
So any discriminatory patterns are probably not only embedded, they’re impossible to remove without some hypothetically non-discriminatory training data. Not sure where they would even look for that.
I wonder if AIs also look at a resume from a guy named LaQuan and immediately turns it down because “black name?”
Most assuredly so. Those AIs are trained on human input, which is practically guaranteed to contain some level of bias.
I’m sure the embedded racism is deeeeep, yeah. All those “AI” hiring models are doing is pattern matching, based on whatever junk was fed into it.
So any discriminatory patterns are probably not only embedded, they’re impossible to remove without some hypothetically non-discriminatory training data. Not sure where they would even look for that.