Superintelligence

Bostrom, a philosopher at Oxford, works through what may be the most consequential question of the century: what happens if we build machines smarter than ourselves. He maps the possible paths to superintelligence, the ways it could slip beyond human control, and the narrow set of strategies that might keep it aligned with human values. Dense and carefully argued, the book turned AI risk from a fringe worry into a mainstream debate, and it is the title Elon Musk pointed to when he began warning about the technology.