How the World Got Wired · Why Hinton Went North

Why This Story Matters

I want you to hold a few things in your mind before we move to the next chapter.

The first thing is that Geoffrey Hinton’s career is a counter-example to a story you have heard, in various forms, throughout this book. The story is: foundational technology gets paid for by governments, and then private companies build on top of it and become rich. That story is mostly accurate. Hinton’s career has a different shape. He chose, deliberately, to be funded by a less wealthy government, on smaller grants, in exchange for the freedom to work on what he wanted, without the implicit assumption that his work would be put to military use. The bargain was: less money, more conscience.

The second thing is that the country that benefited most from his career, in the end, was not Canada. Or at least, not Canada alone. The students he trained in Toronto did not all stay there. The breakthroughs he eventually had were quickly commercialized by American companies, who paid him and his students enormous amounts of money to come work for them. The patient Canadian investment of decades produced a technological revolution that has, by and large, been captured by American firms. (You will see, in the next chapter, exactly how this happened.) Whether Canada should have done more to keep the benefits within its borders is a real question that Canadians ask themselves, with some frustration, today.

The third thing is something I want to ask you to think about, because the choice you are about to read about Hinton making in 2023 — to leave Google, to speak out about the dangers of artificial intelligence — flows directly from the choice he made in 1987 to leave the United States in the first place. The same person who decided, when he was forty years old, that he did not want his research to go into weapons, decided, when he was seventy-five years old, that the world needed to slow down and think about what artificial intelligence was becoming. There is a thread of conscience that runs through his life. It is, in its own quiet way, in the same family tradition that produced George Boole, and Joan Hinton, and the rest of them.

The fourth thing is the harder one. The work Hinton did, when it finally broke through, would not stay in his hands. It would be picked up by enormous companies, funded with enormous capital, accelerated by enormous compute. The patient quiet of the long Canadian decades would be replaced by an industrial race so frantic that, by the time this book is being written, it has consumed the attention of essentially every major government in the world.

That moment of breakthrough, when it came, happened in a way nobody quite expected. It happened in a single afternoon. It happened because of an annual competition, run by an academic, on a dataset of pictures.

That’s Chapter 21.

All 24 chapters of How the World Got Wired · Library · All courses