Envisioning Existential Risks of Artificial Intelligence


I wrote this as a final paper for Olin's social science and humanities foundation course, What Is "I"?, which combines philosophy and cognitive science in an interdisciplinary quest to understand the nature of consciousness, intelligence, and the self.


My work uses the film I, Robot as a springboard for discussing the realistic chances that a human-level artificial intelligence could be created, what the potential damaging impacts could be, and how we might act in the present to mitigate risk.  My analysis draws from Kurzweil's technology-grounded predictions for strong AI by 2030, Asimov's fictional experiments with his "Three Laws" as a top-down control mechanism, and finally from Eliezer Yudkowsky's recent work on Friendly AI.


PDF available for viewing and download below.
Mike Hughes,
Jan 15, 2009, 12:13 PM