R. George Wright (Indiana University Robert H. McKinney School of Law) has posted Some Concerns About Generally Well-Behaved Artificial Superintelligence on SSRN. Here is the abstract:
In a remarkable 1930 essay, John Maynard Keynes declared that “mankind is solving its economic problem.” One possible path to this result could be through the rise of a generally well-behaved Artificial Superintelligence (ASI). Such an ASI would, plainly, and inevitably, generate a range of important legal problems. The law already affords glimmerings of some of the legal problems that a generally benign ASI might bring about. I pursue herein several such possible problems. For convenience, though, I set aside all scenarios in which ASI either comes to little, or else leads to some utter catastrophe.