Expert Systems Boom
Corporate America embraced rule-based expert systems like R1/XCON and MYCIN, and Japan launched its ambitious Fifth Generation Computer project. The industry boomed past a billion dollars before collapsing when these brittle systems proved expensive to maintain and unable to learn. A second AI winter followed, but backpropagation’s rediscovery by Rumelhart, Hinton, and Williams in 1986 quietly set the stage for neural networks’ eventual return.
John Hopfield showed that a neural network could store and retrieve patterns like a physical system reaching equilibrium, reviving connectionism from its decade-long exile.
Japan launched a massive government-funded project to build thinking machines, triggering a global AI arms race that reshaped research funding worldwide.
A Hollywood thriller about a teenage hacker and a war-playing AI so alarmed President Reagan that it sparked America's first cybersecurity directive.
Doug Lenat's Eurisko discovered winning strategies by rewriting its own heuristics, dominating a naval wargame so thoroughly the organizers threatened to cancel the competition.
Doug Lenat launched Cyc, a project to hand-code millions of common-sense facts into a machine — a decades-long bet that AI needed to know what every child knows.
James Cameron's The Terminator introduced Skynet to popular culture, permanently shaping public anxiety about artificial intelligence.
Hinton and Sejnowski invented the Boltzmann machine, a neural network that learned by simulating the random thermal fluctuations of molecules.
Digital Equipment Corporation's R1/XCON became the first expert system to prove AI could deliver massive commercial value, configuring minicomputers faster and more accurately than any human.
IBM's speech recognition team, led by Fred Jelinek, proved that statistical methods could understand human speech better than any linguistic approach — famously quipping that firing linguists improved the system.
Marvin Minsky proposed that intelligence isn't one thing but a society of tiny mindless agents, reshaping how researchers thought about building thinking machines.
A 1986 Nature paper showed neural networks how to learn by propagating errors backward, reviving a field that had been declared dead.
Judea Pearl's Probabilistic Reasoning in Intelligent Systems replaced brittle logic with the mathematics of belief, quietly reshaping AI from the inside out.
Yann LeCun's convolutional neural network learned to read handwritten zip codes, proving that neural networks could solve real-world problems and laying the foundation for modern computer vision.
Jeffrey Elman's simple recurrent network showed that neural networks could process sequences and learn grammar, opening the door to modern language AI.
A rebel roboticist published a manifesto arguing that true intelligence doesn't come from abstract reasoning — it comes from having a body in the world.
Gerald Tesauro's TD-Gammon taught itself to play backgammon at world-champion level, proving neural networks could discover strategies humans never imagined.