> You don’t think AI is a “threat to democracy”? First, we have to define what "AI" is. In the current context, it might be Machine Learning or a collection o

> You don’t think AI is a “threat to democracy”? First, we have to define what "AI" is. In the current context, it might be Machine Learning or a collection of algorithms defining what to display to a give user, for example. Both ML and algorithms are completely neutral by themselves, it is their use that makes them "dangerous". You can use it for decades, e.g. collecting data from sensors and classifying them, an nothing bad happens. But the moment a human being connects a software system to an actuator - news feed, an engine, a rocket, an x-ray machine - they explicitly take responsibility for the functioning of such a connected system. If it fails, putting the blame on "AI" is ridiculous.