As part of my fellowship at Harvard, I’ve been thinking hard about artificial intelligence (AI). Start with the physics of the problem: AI is what lawyers call an unavoidably unsafe product. No matter how carefully we build it, it is … Continue reading →