If Anyone Builds It, Everyone Dies: Why Superhuman AI Would Kill Us All

Eliezer Yudkowsky & Nate Soares. If Anyone Builds It, Everyone Dies: Why Superhuman AI Would Kill Us All
Artificial intelligence is advancing at an unprecedented pace. Eliezer Yudkowsky and Nate Soares argue that the development of superhuman AI poses an existential threat to humanity—and that once such systems are created, controlling them may prove impossible.
Rejecting the assumption that more intelligent machines will naturally share human values, the authors explain why advanced AI could pursue goals fundamentally incompatible with human survival, even without malicious intent. Drawing on computer science, decision theory, and AI alignment research, they argue that the challenge of building safe superintelligent systems is far more difficult than commonly assumed.
Direct, provocative, and uncompromising, If Anyone Builds It, Everyone Dies presents the case for treating AI safety as one of the most urgent scientific and political challenges of the twenty-first century.