I think that the broader concept of instilling desired ethics into an AI is part of the friendly AI problem, which is very real and serious — and possibly not reasonably solvable. So while I don’t really think that Cortana 2045 running around raping humans or something like that is very high on my likely risk list, I think that the broader problem that contains that particular issue probably is something that we’ll need to deal with.
I think that the broader concept of instilling desired ethics into an AI is part of the friendly AI problem, which is very real and serious — and possibly not reasonably solvable. So while I don’t really think that Cortana 2045 running around raping humans or something like that is very high on my likely risk list, I think that the broader problem that contains that particular issue probably is something that we’ll need to deal with.