CyberneticsCybernetics

During World War II, MIT mathematician Norbert Wiener worked on a hard problem: how to aim an anti-aircraft gun at a plane that will have moved by the time the shell arrives. The gun had to predict the pilot's future position from past behavior, then correct itself as new observations came in.

Wiener realized this loop of act, observe error, adjust, act again was not just an artillery trick. It described thermostats, steam engine governors, a hand reaching for a cup, and the way animals keep their balance. The common thread was feedback: systems that steer themselves by consuming information about their own mistakes.

In 1948 he published 'Cybernetics: Or Control and Communication in the Animal and the Machine', naming the new field after the Greek word for steersman. The book became an unlikely bestseller, dense with mathematics yet read by biologists, engineers, psychologists, and the general public.

Cybernetics gave the pioneers of AI their first shared vocabulary. The idea that intelligence might be a matter of information, feedback, and control, implementable in flesh or metal alike, directly shaped the Macy Conferences, early neural network research, and the generation that convened at Dartmouth in 1956.

Wiener himself grew uneasy about where it all pointed. In later books he warned about automation displacing workers and about delegating decisions to machines whose objectives we specify badly, concerns that read as strikingly current.

The word cybernetics faded as AI, control theory, and neuroscience split into separate disciplines. But its core insight, that behavior can be understood as feedback loops processing information, sits underneath nearly everything that followed.