But at the end of the day, humans are still in charge, not algorithms. If they can't adequately control the algorithms, perhaps they're not ready to be deployed.
To see this a little more clearly, if a military deployed autonomous killer robots that kept accidentally killing civilian children, would you then argue the military isn't responsible?
To see this a little more clearly, if a military deployed autonomous killer robots that kept accidentally killing civilian children, would you then argue the military isn't responsible?