PERSON
James McClelland
The cognitive scientist who co-built the theoretical framework that became deep learning, won the argument that intelligence emerges from distributed connection-learning rather than symbolic rules—and then, having won, insisted most honestly on what his framework still cannot explain.
James McClelland is the most consequential AI theorist that most people discussing AI have never heard of. The systems now reshaping the economy descend, through an unbroken intellectual line, from the research program he co-led in the 1980s with David Rumelhart:
Parallel Distributed Processing, the idea that intelligence need not be a set of rules executed by a central processor but can emerge from the cooperation of many simple units adjusting the strength of their connections in response to experience. That idea was dismissed in 1986 by much of the field; it is now the engine of everything called AI. McClelland is the right lens for this moment for two reasons that run in opposite directions. He proved, against fierce resistance, that the connectionist gamble was correct—that the mechanism of mind is distributed, statistical, and rule-free at bottom, not symbolic. And then, having proved it more completely than he expected, he became the most credible voice insisting