Logic Theorist, the first artificial intelligence program
Newell, Simon, and Shaw built the Logic Theorist at RAND, a program that proved theorems in symbolic logic and is widely called the first AI program.
The MIT symposium often called the birthday of cognitive science
On 11 September 1956 an MIT information-theory symposium brought together the papers George Miller later called the birth of cognitive science.
Bellman formalizes dynamic programming
Richard Bellman's 1957 book Dynamic Programming gave reinforcement learning its mathematical backbone, including the Bellman equation.
The first program to play a full game of chess
Alex Bernstein's IBM 704 program of 1957 was the first to play a complete game of chess from start to finish.
ARPA is founded after Sputnik
On February 7, 1958, the US created ARPA, later DARPA, the agency that would fund foundational computing and AI research.
The perceptron: a machine that learns from examples
Frank Rosenblatt's perceptron was an early trainable neural network that adjusted its own connections to classify patterns.
Mechanisation of Thought Processes symposium
A 1958 symposium at the UK National Physical Laboratory gathered early AI work, including McCarthy's advice taker and Selfridge's Pandemonium.
Arthur Samuel's checkers program
Arthur Samuel's self-improving checkers program popularized the term machine learning and showed computers could learn from experience.
Selfridge's Pandemonium model of learning
Selfridge's 1959 Pandemonium described pattern recognition as layers of parallel feature detectors, a conceptual ancestor of neural networks.
General Problem Solver and means-ends analysis
GPS, a 1959 program by Newell, Shaw, and Simon, tried to solve any problem by repeatedly reducing the gap to the goal - a method called means-ends analysis.
PLATO, the first generalized computer-based education system
Donald Bitzer's PLATO system at the University of Illinois pioneered computer-based teaching and the online community in 1960.
McCarthy defines Lisp
John McCarthy's 1960 paper defined Lisp, the list-processing language that became the standard tool of AI research for the next thirty years.
Wiener's Some Moral and Technical Consequences of Automation
Norbert Wiener's 1960 Science article warned that a learning machine pursuing a literal goal could become impossible to stop in time.
Michie's MENACE learns noughts and crosses from matchboxes
In 1961 Donald Michie built MENACE, a stack of about 300 matchboxes and beads that learned to play noughts and crosses by reinforcement, with no computer.
Unimate, the first industrial robot, goes to work at GM
George Devol's patented programmable manipulator, built by Unimation, became the first industrial robot on a factory floor at General Motors.
Hubel and Wiesel map the visual cortex
Hubel and Wiesel's 1962 paper showed the visual cortex detects edges with simple and complex cells, inspiring later convolutional networks.
The IBM Shoebox recognizes spoken numbers and commands
IBM demonstrates the Shoebox at the 1962 Seattle World's Fair, a device that recognizes 16 spoken words and does arithmetic by voice.
The Association for Computational Linguistics is founded
Founded in 1962 to study machine translation, the ACL became the leading scientific society for computational linguistics and NLP.
Roberts' Machine Perception of Three-Dimensional Solids
Larry Roberts' 1963 MIT PhD thesis, often called the first computer vision dissertation, recovered 3D shape from 2D line drawings of blocks.
Woodrow Bledsoe builds the first facial recognition system
In 1964-65 Woody Bledsoe's team built the first computer facial recognition system, funded by an unnamed intelligence agency.
DENDRAL, the first expert system
DENDRAL was a Stanford program that reasoned like a human chemist to identify molecules from instrument data, and is widely regarded as the first expert system.
I.J. Good and the Intelligence Explosion
Mathematician I.J. Good's 1965 paper introduced the idea of an ultraintelligent machine that designs ever-better machines, producing an intelligence explosion.
Gordon Moore predicts the doubling of components on a chip
Gordon Moore's 1965 article observed that components on a chip were doubling roughly every year, the forecast that became Moore's Law.
1966 ALPAC report ends a decade of machine-translation funding
The 1966 ALPAC report found no near-term prospect of useful machine translation and urged cutting funding, collapsing US support for roughly two decades.
ELIZA, the first famous chatbot
Joseph Weizenbaum's ELIZA was an early conversational program that mimicked a psychotherapist using simple pattern matching.
Shakey, the first mobile robot that could reason about its actions
SRI's Shakey combined perception, planning, and movement in one machine, and produced lasting AI techniques including A* search and STRIPS planning.
Mac Hack VI beats a human in tournament play
Richard Greenblatt's Mac Hack VI was the first chess program to enter human tournaments, earn a rating, and beat a person in rated play.
Peter Toma founds SYSTRAN, the long-lived machine translation company
SYSTRAN began in 1968 translating Russian into English for the US Air Force and later powered the Apollo-Soyuz space mission.
Perceptrons and the first neural network winter
Minsky and Papert's book exposed the limits of single-layer perceptrons, helping freeze neural network funding for years.
The Antikythera Mechanism, an ancient Greek geared computer
A Greek bronze geared device from around the 2nd century BC modeled the heavens and predicted eclipses - the oldest known analog computer.
Winston's Arch-Learning Program
Patrick Winston's 1970 program learned structural concepts like 'arch' from blocks-world examples and near-misses, an early model of concept learning.
SHRDLU understands English in a world of blocks
Terry Winograd's SHRDLU let a person hold a typed English conversation with a computer that moved blocks in a simulated world - early language understanding.
Colmerauer and Roussel create Prolog
In Marseille in 1972, Alain Colmerauer and Philippe Roussel created Prolog, the logic programming language that let you state facts and rules and query them.
PARRY, the chatbot that simulated paranoia
Kenneth Colby's PARRY modeled a paranoid patient so convincingly that psychiatrists could not reliably tell it from a human.
PLATO, the first computer-based education system
PLATO, begun at the University of Illinois in 1960, was the first generalized computer-assisted instruction system.
Harold Cohen builds AARON, a program that draws on its own
From the early 1970s Harold Cohen developed AARON, a rule-based program that produced original drawings autonomously decades before neural image generators.
The Lighthill Report triggers the first AI winter in the UK
Sir James Lighthill's survey for the UK Science Research Council judged AI a disappointment, cutting British AI funding for a decade.
WABOT-1, the world's first full-scale humanoid robot
Waseda University completed WABOT-1 in 1973, the first full-scale anthropomorphic robot that walked, gripped objects, and spoke Japanese.
HACKER, a Program That Learned From Its Bugs
Gerald Sussman's 1973 HACKER wrote blocks-world plans, watched them fail, and patched its own procedures, an early model of learning by debugging.
First World Computer Chess Championship
The first World Computer Chess Championship, held in Stockholm in 1974, became a 50-year proving ground for game-playing AI; the Soviet Kaissa won.
Werbos describes backpropagation in his Harvard PhD thesis
Paul Werbos's 1974 Harvard dissertation set out the method later known as backpropagation for training multilayer networks.
Harpy understands a thousand words of connected speech
Carnegie Mellon's Harpy, built under DARPA's Speech Understanding Research program, recognizes a 1,011-word vocabulary using beam search.
MYCIN, the medical expert system
MYCIN was a Stanford expert system that advised doctors on diagnosing and treating blood infections using a few hundred rules and a way of handling uncertainty.
METEO, the first machine translation system in daily operational use
Canada's METEO system began translating weather forecasts between English and French in 1977, the first MT in routine daily use.
PROSPECTOR, the Mineral-Exploration Expert System
SRI's PROSPECTOR encoded geologists' expertise in inference networks and famously flagged a real, previously unknown molybdenum deposit.
Godel, Escher, Bach: An Eternal Golden Braid
Hofstadter's 1979 book on strange loops and minds won the Pulitzer and drew a generation toward computer science and AI.
The Stanford Cart crosses a room on its own
In 1979 Hans Moravec's Stanford Cart used stereo vision to autonomously cross a chair-filled room, an early milestone in mobile robotics.
Fukushima's Neocognitron, an ancestor of the convolutional network
Kunihiko Fukushima's Neocognitron introduced a layered, shift-tolerant network for pattern recognition that prefigured modern CNNs.