Because people see machines and AI as coldly logical, they don’t see the biases built in by the human programmers deliberately or indirectly through data set selection and training.
Microsoft Forced to Apologize for Racist Twitter AI chatbot
Microsoft Confirms Its Chinese-Language Chatbot Filters Certain Topics
Looks Like Microsofts New Chatbot, Zo, Is Already Turning Into a Right-Wing Misanthrope
Prediction: Progressives will link artificial intelligence to their wolf cries of racism and bigotry
An Artificial Intelligence Developed Its Own Non-Human Language [Ruh-Roh!]