The University in Nottingham England has a great series of videos on many different computer topics and one of my favorites right now is this playlist. In the series, Rob Miles does a tremendous job of explaining why really thinking through the process of AI design is crucial before unforeseen consequences might arise from a general learning AI run amok. When AIs come near to our human level of self-awareness and generalized adaptability it will likely be beyond the point of design consideration and soon past the point of return to stop a system that takes the wrong trajectory (evil like skynet or non-sinister but ignorantly or indifferently non-human preserving). Really interesting and informative!
