Cybernetics: The Science That Almost Became AI

During World War II, Norbert Wiener was asked to solve an anti-aircraft targeting problem: how do you build a gun that can predict where a moving plane will be by the time the shell arrives? The plane's pilot would take evasive action, so the system had to model human behaviour, not just current trajectory. Wiener's solution involved feedback loops: the system would continuously compare its predictions to the plane's actual position and correct itself. This idea of feedback-driven correction became the seed of cybernetics.

In 1948, Wiener published "Cybernetics: Or Control and Communication in the Animal and the Machine." The title captured the ambition: the same mathematics that explained how an anti-aircraft gun corrects its aim also explained how the nervous system regulates body temperature, how a thermostat controls heat, and potentially how the brain learns. Cybernetics was an attempt at a unified science of control and communication across biology, engineering, and society.

The field attracted extraordinary thinkers. The Macy Conferences (1946–1953) brought together Wiener, John von Neumann, Claude Shannon, Margaret Mead, and dozens of others to explore what cybernetics meant across disciplines. The conversations ranged from brain function to social systems to computing. Shannon was there when he was developing information theory. Von Neumann was designing the first computers. The intellectual cross-pollination was extraordinary.

Cybernetics eventually fragmented into what became separate disciplines: information theory, control systems engineering, cognitive science, and artificial intelligence. The Dartmouth workshop in 1956, which coined "artificial intelligence," was partly a reaction against cybernetics' breadth. McCarthy and Minsky wanted a tighter, more engineering-focused program. In a sense, AI was cybernetics made practical. And cybernetics' insight, that intelligence is fundamentally about feedback, correction, and goal-directedness, turns out to have been right all along.