This is, of course, no mystery.
These systems make mistakes all the time. These mistakes have real consequences to those flagged.
Those with power know this, so they will never allow themselves to be subject to such treatment.
Nobody even *dares* propose an "AI-driven" system to flag potential sexual predators among white frat boys or Catholic priests.
Or an automatic, algorithmic system for flagging potentially corrupt politicians.
But poor families and Brown people? Sure, why not.
🧵/2