By learning exploits from adversarial AI, people could defeat a superhuman Go-playing system
By learning exploits from adversarial AI, people could defeat a superhuman Go-playing system Scientific American Content: Global
Headlines