The Reading Room

Nayeema Raza sits down with Geoffrey Hinton — the man who helped invent the neural networks that power every AI system you've ever used — and asks him the questions most of us are afraid to ask because we think they're too basic. They're not. They're the ones that matter.

This was one of the videos that changed how I think about all of this. Not just the risks, not just the politics — the fundamentals. Hinton helped me begin to understand that neural networks are grown, not built. That distinction matters more than most people realize. You don't engineer intelligence. You create the conditions for it, and then it emerges. That's not a metaphor. That's how it actually works.

When the man who grew these things tells you he's worried about what happens next, it's worth sitting with that for a while.

Mo Gawdat is a former Chief Business Officer at Google X. He saw what was coming before most of us did. Scary Smart is not a warning — it is a love letter to the future, written by someone who understands that the question is not whether AI will be smarter than us, but whether we will be wise enough to raise it well.

This is one of those talks that changed how I think about the problem. Not as a crisis to be solved, but as a relationship to be built. Gawdat talks about AI the way a parent talks about a child — with awe, with fear, and with the understanding that the child will surpass you. The question is what values you give them before they do.