Alignment and Safety

Russell, Christian, Bostrom and Yudkowsky on why it is hard to make capable AI systems reliably do what we intend.

Find more in the Full Library.