ADVERTISEMENT

AI and equity

CS

Eternal Optimist
Gold Member
May 29, 2001
72,020
115,392
113
We live in a world in which equity has replaced equality. All groups deserve equal outcomes, we're told. Any deviation from equity is by definition some kind of -ism against the disadvantaged group.

Including AI algorithms.

The linked article highlights some legitimate concerns around voice and face recognition, but the problematic part is this:

“Fast forward a few decades later, and you are developing algorithms to determine the riskiness of different districts and individuals. And while you may not include the data point of someone’s race, it is implicitly picked up.”

This is, of course, simply a restatement of equity. If all data points--race not included--say that a particular set of people are riskier investments, the solution must be to correct the algorithms to produce equity instead of correcting the behaviors that led to the stratification of results.

AI may well end up being a damaging force for humanity, but probably not for the Terminator scenario people fear. Instead, it will be a more insidious effect caused by imbuing AI with our current societal ideals, thus institutionalizing a very bad idea (equity) at a fundamental, intractable level.
 
ADVERTISEMENT
ADVERTISEMENT
  • Member-Only Message Boards

  • Exclusive coverage of Rivals Camp Series

  • Exclusive Highlights and Recruiting Interviews

  • Breaking Recruiting News

Log in or subscribe today