@tess The risk has never been a self-aware malicious AI, it's AI being applied towards malicious acts with no intervention. The opaque bureaucracy of "computer says no" destroying lives and automating institutional bias to a murderous efficiency.