AI trained for treachery becomes the perfect agent
go.theregister.com
Opinion: We’re blind to malicious AI until it hits. We can still open our eyes to stopping it