> The reason those predictions are so skewed is still unknown
Well, I bet the reason are historical statistics of recidivism, which presumably were used to train these algorithms.
But hey, this is NYT, they cannot simply acknowledge the existence of differences in crime statistics between races, so 'the reason remains unknown'. Hilarious.
Well, I bet the reason are historical statistics of recidivism, which presumably were used to train these algorithms.
But hey, this is NYT, they cannot simply acknowledge the existence of differences in crime statistics between races, so 'the reason remains unknown'. Hilarious.