Artificial intelligence technology that works well for the world’s billions of people of color was a business imperative for Google when the company asked sociologist Ellis Monk to help make its AI products more inclusive.
Machine Learning is only as good as the dataset on which it is trained, which is why an internet chatbot trained on most of what passes for discourse on X turns into a really nasty racist in no time flat. By going after programs that reduce bias in ML models they’re almost guaranteeing skewed outcomes as law enforcement embraces ML in policing. This is gonna suuuuuuuck.
Machine Learning is only as good as the dataset on which it is trained, which is why an internet chatbot trained on most of what passes for discourse on X turns into a really nasty racist in no time flat. By going after programs that reduce bias in ML models they’re almost guaranteeing skewed outcomes as law enforcement embraces ML in policing. This is gonna suuuuuuuck.