Ethics in Artificial Intelligence—CSC484

Professor Clark Elliott

Artificial Sentient Beings:

Taking this one step further, if we succeed in building artificial sentience into our AI systems (either intentionally or by chance) what are the ethics of...

...turning them off?
...using them as work slaves to do exactly what we tell them to do?
...building severely depressed agents so we can study them?
...achieving a state of Nirvana / rapture / ecstasy / enlightenment in some of our agents: are we obligated to build as many of them as possible?
...one of our sentient AI beings losing its close relationship with a human and experiencing grief over it?
...humans forming what appear to be exclusive attachments with an AI agent that in turn has the same relationship with hundreds of other humans (e.g. as in the movie Her)?