Marvin Minsky

Marvin Minsky had a question that followed him his whole life: how can something as simple as a neuron produce something as complex as a thought? A mathematician trained at Harvard and Princeton, he joined MIT in 1958 and never left. A year later, alongside John McCarthy, he co-founded the MIT Artificial Intelligence Laboratory, which became one of the most influential AI research centers in the world. And before that, in 1956, he was one of four organizers of the Dartmouth Conference, the event that gave artificial intelligence its name and its shape as a field.

In 1974 he published "A Framework for Representing Knowledge," introducing the concept of "frames": data structures that represent typical situations with default values, an idea that changed how computers could organize and use knowledge. But his most ambitious work came in 1986 with "The Society of Mind": a book proposing that the mind is not a single thing but the result of hundreds of simple processes (which he called "agents") interacting with each other. None of those agents is intelligent on its own. Intelligence, according to Minsky, emerges from the conversation among all of them. He had received the Turing Award in 1969, the equivalent of the Nobel Prize in computing, in recognition of that pioneering work.

What set Minsky apart was that his curiosity had no disciplinary boundaries. He invented the confocal scanning microscope in 1956 (today a fundamental instrument in biology), built one of the first automatic learning machines based on neural networks in 1951 (the SNARC, which simulated a rat in a maze using 40 vacuum tubes), and in 1961 wrote "Steps Toward Artificial Intelligence," a text that mapped the field for decades. He also received the Japan Prize in 1990 and the Benjamin Franklin Medal in 2001.

The lesson Minsky leaves me with is that intelligence, yours included, doesn't come from a single powerful place: it comes from many small pieces that cooperate. When you use Claude to solve a difficult problem, what's happening under the hood isn't all that different from what Minsky imagined forty years ago: a society of agents, each doing its part. He didn't have the tools we have today, but he had the right question.