| Date | Development |
|---|
| Antiquity | Greek myths of Hephaestus and Pygmalion incorporated the idea of intelligent automata (such as Talos) and artificial beings (such as Galatea and Pandora). |
|---|
| Sacred mechanical statues built in Egypt and Greece were believed to be capable of wisdom and emotion. Hermes Trismegistus would write "they have sensus and spiritus ... by discovering the true nature of the gods, man has been able to reproduce it." |
| 10th century BC | Yan Shi presented King Mu of Zhou with mechanical men which were capable of moving their bodies independently. |
|---|
| 384 BC–322 BC | Aristotle described the syllogism, a method of formal, mechanical thought in the Organon.[1] [2] Aristotle also described means-ends analysis (an algorithm for planning) in Nicomachean Ethics, the same algorithm used by Newell and Simon's General Problem Solver (1959). |
|---|
| 3rd century BC | Ctesibius invents a mechanical water clock with an alarm. This was the first example of a feedback mechanism. |
|---|
| 1st century | Hero of Alexandria created mechanical men and other automatons. He produced what may have been "the world's first practical programmable machine:" an automatic theatre. |
|---|
| 260 | Porphyry wrote Isagogê which categorized knowledge and logic, including a drawing of what would later be called a "semantic net". |
|---|
| ~800 | Jabir ibn Hayyan developed the Arabic alchemical theory of Takwin, the artificial creation of life in the laboratory, up to and including human life. |
|---|
| 9th Century | The Banū Mūsā brothers created a programmable music automaton described in their Book of Ingenious Devices: a steam-driven flute controlled by a program represented by pins on a revolving cylinder.[3] This was "perhaps the first machine with a stored program". |
|---|
| al-Khwārizmī wrote textbooks with precise step-by-step methods for arithmetic and algebra, used in Islam, India and Europe until the 16th century. The word "algorithm" is derived from his name. |
| 1206 | Ismail al-Jazari created a programmable orchestra of mechanical human beings.[4] |
|---|
| 1275 | Ramon Llull, Mallorcan theologian, invents the Ars Magna, a tool for combining concepts mechanically based on an Arabic astrological tool, the Zairja. Llull described his machines as mechanical entities that could combine basic truth and facts to produce advanced knowledge. The method would be developed further by Gottfried Leibniz in the 17th century. |
|---|
| ~1500 | Paracelsus claimed to have created an artificial man out of magnetism, sperm and alchemy. |
|---|
| ~1580 | Rabbi Judah Loew ben Bezalel of Prague is said to have invented the Golem, a clay man brought to life.[5] | |
|---|
| Date | Development |
|---|
| 1620 | Francis Bacon developed empirical theory of knowledge and introduced inductive logic in his work Novum Organum, a play on Aristotle's title Organon.[6] [7] |
|---|
| 1623 | Wilhelm Schickard drew a calculating clock on a letter to Kepler. This will be the first of five unsuccessful attempts at designing a direct entry calculating clock in the 17th century (including the designs of Tito Burattini, Samuel Morland and René Grillet). |
|---|
| 1641 | Thomas Hobbes published Leviathan and presented a mechanical, combinatorial theory of cognition. He wrote "...for reason is nothing but reckoning". |
|---|
| 1642 | Blaise Pascal invented a mechanical calculator, the first digital calculating machine. |
|---|
| 1647 | René Descartes proposed that bodies of animals are nothing more than complex machines (but that mental phenomena are of a different "substance"). |
|---|
| 1654 | Blaise Pascal described how to find expected values in probability, in 1662 Antoine Arnauld published a formula to find the maximum expected value, and in 1663, Gerolamo Cardano's solution to the same problems is published 116 years after it was written. The theory of probability is further developed by Jacob Bernoulli and Pierre Laplace in the 18th century. Probability theory would become central to AI and machine learning from the 1990s onward. |
|---|
| 1672 | Gottfried Wilhelm Leibniz improved the earlier machines, making the Stepped Reckoner to do multiplication and division. |
|---|
| 1676 | Leibniz derived the chain rule.[8] The rule is used by AI to train neural networks, for example the backpropagation algorithm uses the chain rule. |
|---|
| 1679 | Leibniz developed a universal calculus of reasoning (alphabet of human thought) by which arguments could be decided mechanically. It assigned a specific number to each and every object in the world, as a prelude to an algebraic solution to all possible problems. |
|---|
| 1726 | Jonathan Swift published Gulliver's Travels, which includes this description of the Engine, a machine on the island of Laputa: "a Project for improving speculative Knowledge by practical and mechanical Operations" by using this "Contrivance", "the most ignorant Person at a reasonable Charge, and with a little bodily Labour, may write Books in Philosophy, Poetry, Politicks, Law, Mathematicks, and Theology, with the least Assistance from Genius or study."[9] The machine is a parody of Ars Magna, one of the inspirations of Gottfried Wilhelm Leibniz' mechanism. |
|---|
| 1738 | Daniel Bernoulli introduces the concept of "utility", a generalization of probability, the basis of economics and decision theory, and the mathematical foundation for the way AI represents the "goals" of intelligent agents. |
|---|
| 1739 | David Hume described induction, the logical method of learning generalities from examples. |
|---|
| 1750 | Julien Offray de La Mettrie published L'Homme Machine, which argued that human thought is strictly mechanical. |
|---|
| 1763 | Thomas Bayes's work An Essay Towards Solving a Problem in the Doctrine of Chances, published two years after his death, laid the foundations of Bayes' theorem and used in modern AI in Bayesian networks. |
|---|
| 1769 | Wolfgang von Kempelen built and toured with his chess-playing automaton, The Turk, which Kempelen claimed could defeat human players. The Turk was later shown to be a hoax, involving a human chess player. |
|---|
| 1795-1805 | The simplest kind of artificial neural network is the linear network. It has been known for over two centuries as the method of least squares or linear regression. It was used as a means of finding a good rough linear fit to a set of points by Adrien-Marie Legendre (1805)[10] and Carl Friedrich Gauss (1795)[11] for the prediction of planetary movement.[12] |
|---|
| 1800 | Joseph Marie Jacquard created a programmable loom, based on earlier inventions by Basile Bouchon (1725), Jean-Baptiste Falcon (1728) and Jacques Vaucanson (1740).[13] Replaceable punched cards controlled sequences of operations in the process of manufacturing textiles. This may have been the first industrial software for commercial enterprises. |
|---|
| 1818 | Mary Shelley published the story of Frankenstein; or the Modern Prometheus, a fictional consideration of the ethics of creating sentient beings. |
|---|
| 1822–1859 | Charles Babbage & Ada Lovelace worked on programmable mechanical calculating machines. |
|---|
| 1837 | The mathematician Bernard Bolzano made the first modern attempt to formalize semantics.[14] |
|---|
| 1854 | George Boole set out to "investigate the fundamental laws of those operations of the mind by which reasoning is performed, to give expression to them in the symbolic language of a calculus", inventing Boolean algebra. |
|---|
| 1863 | Samuel Butler suggested that Darwinian evolution also applies to machines, and speculates that they will one day become conscious and eventually supplant humanity.[15] | |
|---|
| Date | Development |
|---|
| 1910-1913 | Bertrand Russell and Alfred North Whitehead published Principia Mathematica, which showed that all of elementary mathematics could be reduced to mechanical reasoning in formal logic. |
|---|
| 1912-1914 | Leonardo Torres Quevedo built an automaton for chess endgames, El Ajedrecista. He was called "the 20th century's first AI pioneer". In his Essays on Automatics (1914), Torres published speculation about thinking and automata and introduced the idea of floating-point arithmetic.[16] |
|---|
| 1923 | Karel Čapek's play R.U.R. (Rossum's Universal Robots) opened in London. This is the first use of the word "robot" in English. |
|---|
| 1920-1925 | Wilhelm Lenz and Ernst Ising created and analyzed the Ising model (1925)[17] which can be viewed as the first artificial recurrent neural network (RNN) consisting of neuron-like threshold elements. In 1972, Shun'ichi Amari made this architecture adaptive.[18] |
|---|
| 1920s and 1930s | Ludwig Wittgenstein's Tractatus Logico-Philosophicus (1921) inspires Rudolf Carnap and the logical positivists of the Vienna Circle to use formal logic as the foundation of philosophy. However, Wittgenstein's later work in the 1940s demonstrates that context free symbolic logic is incoherent without human interpretation. |
|---|
| 1931 | Kurt Gödel encoded mathematical statements and proofs as integers, and showed that there are true theorems that are unprovable by any consistent theorem-proving machine. Thus "he identified fundamental limits of algorithmic theorem proving, computing, and any type of computation-based AI," laying foundations of theoretical computer science and AI theory. |
|---|
| 1935 | Alonzo Church extended Gödel's proof and showed that the decision problem of computer science does not have a general solution.[19] He developed the Lambda Calculus, which will eventually be fundamental to the theory of computer languages. |
|---|
| 1936 | Konrad Zuse filed his patent application for a program-controlled computer.[20] |
|---|
| 1937 | Alan Turing published "On Computable Numbers", which laid the foundations of the modern theory of computation by introducing the Turing machine, a physical interpretation of "computability". He used it to confirm Gödel by proving that the halting problem is undecidable. |
|---|
| 1940 | Edward Condon displayed Nimatron, a digital machine that played Nim perfectly. |
|---|
| 1941 | Konrad Zuse built the first working program-controlled general-purpose computer.[21] |
|---|
| 1943 | Warren Sturgis McCulloch and Walter Pitts publish "A Logical Calculus of the Ideas Immanent in Nervous Activity" (1943), the first mathematical description of an artificial neural networks. |
|---|
| Arturo Rosenblueth, Norbert Wiener and Julian Bigelow coin the term "cybernetics". Wiener's popular book by that name published in 1948. |
| 1945 | Game theory which would prove invaluable in the progress of AI was introduced with the 1944 paper "Theory of Games and Economic Behavior" by mathematician John von Neumann and economist Oskar Morgenstern. |
|---|
| Vannevar Bush published "As We May Think" (The Atlantic Monthly, July 1945) a prescient vision of the future in which computers assist humans in many activities. |
| 1948 | Alan Turing produces "Intelligent Machinery" report, regarded as the first manifesto of Artificial Intelligence. It introduces many concepts including the logic-based approach to problem solving, that intellectual activity consists mainly of various kinds of search, and a discussion of machine learning in which he anticipates the Connectionism AI approach.[22] |
|---|
| John von Neumann (quoted by E.T. Jaynes) in response to a comment at a lecture that it was impossible for a machine (at least ones created by humans) to think: "You insist that there is something a machine cannot do. If you will tell me precisely what it is that a machine cannot do, then I can always make a machine which will do just that!". Von Neumann was presumably alluding to the Church–Turing thesis which states that any effective procedure can be simulated by a (generalized) computer. |
| 1949 | Donald Hebb develops Hebbian Learning, a possible algorithm for learning in neural networks. | |
|---|
| Date | Development |
|---|
| 1950 | Alan Turing published "Computing Machinery and Intelligence", which proposes the Turing test as a measure of machine intelligence and answered all of the most common objections to the proposal "machines can think". |
|---|
| Claude Shannon published a detailed analysis of chess playing as search. |
| Isaac Asimov published his Three Laws of Robotics. |
| 1951 | The first working AI programs were written in 1951 to run on the Ferranti Mark 1 machine of the University of Manchester: a checkers-playing program written by Christopher Strachey and a chess-playing program written by Dietrich Prinz. |
|---|
| 1952–1962 | Arthur Samuel (IBM) wrote the first game-playing program, for checkers (draughts), to achieve sufficient skill to challenge a respectable amateur. His first checkers-playing program was written in 1952, and in 1955 he created a version that learned to play.[23] |
|---|
| 1956 | The Dartmouth College summer AI conference is organized by John McCarthy, Marvin Minsky, Nathan Rochester of IBM and Claude Shannon. McCarthy coins the term artificial intelligence for the conference.[24] |
|---|
| The first demonstration of the Logic Theorist (LT) written by Allen Newell, J.C. Shaw and Herbert A. Simon (Carnegie Institute of Technology, now Carnegie Mellon University or CMU). This is often called the first AI program, though Samuel's checkers program also has a strong claim. This program has been described as the first deliberately engineered to perform automated reasoning, and would eventually prove 38 of the first 52 theorems in Russell and Whitehead's Principia Mathematica, and find new and more elegant proofs for some.[25] Simon said that they had "solved the venerable mind/body problem, explaining how a system composed of matter can have the properties of mind".[26] |
| 1958 | John McCarthy (Massachusetts Institute of Technology or MIT) invented the Lisp programming language. |
|---|
| Herbert Gelernter and Nathan Rochester (IBM) described a theorem prover in geometry. It exploited a semantic model of the domain in the form of diagrams of "typical" cases. |
| Teddington Conference on the Mechanization of Thought Processes was held in the UK and among the papers presented were John McCarthy's "Programs with Common Sense" (which proposed the Advice Taker application as a primary research goal) Oliver Selfridge's "Pandemonium", and Marvin Minsky's "Some Methods of Heuristic Programming and Artificial Intelligence". |
| 1959 | The General Problem Solver (GPS) was created by Newell, Shaw and Simon while at CMU. |
|---|
| John McCarthy and Marvin Minsky founded the MIT AI Lab. |
| Late 1950s, early 1960s | Margaret Masterman and colleagues at University of Cambridge design semantic nets for machine translation. | |
|---|
| Date | Development |
|---|
| 1960s | Ray Solomonoff lays the foundations of a mathematical theory of AI, introducing universal Bayesian methods for inductive inference and prediction. |
|---|
| 1960 | "Man-Computer Symbiosis" by J.C.R. Licklider. |
|---|
| 1961 | James Slagle (PhD dissertation, MIT) wrote (in Lisp) the first symbolic integration program, SAINT, which solved calculus problems at the college freshman level. |
|---|
| In Minds, Machines and Gödel, John Lucas[27] denied the possibility of machine intelligence on logical or philosophical grounds. He referred to Kurt Gödel's result of 1931: sufficiently powerful formal systems are either inconsistent or allow for formulating true theorems unprovable by any theorem-proving AI deriving all provable theorems from the axioms. Since humans are able to "see" the truth of such theorems, machines were deemed inferior. |
| Unimation's industrial robot Unimate worked on a General Motors automobile assembly line. |
| 1963 | Thomas Evans' program, ANALOGY, written as part of his PhD work at MIT, demonstrated that computers can solve the same analogy problems as are given on IQ tests. |
|---|
| Edward Feigenbaum and Julian Feldman published Computers and Thought, the first collection of articles about artificial intelligence.[28] [29] [30] [31] |
| Leonard Uhr and Charles Vossler published "A Pattern Recognition Program That Generates, Evaluates, and Adjusts Its Own Operators", which described one of the first machine learning programs that could adaptively acquire and modify features and thereby overcome the limitations of simple perceptrons of Rosenblatt. |
| 1964 | Danny Bobrow's dissertation at MIT (technical report #1 from MIT's AI group, Project MAC), shows that computers can understand natural language well enough to solve algebra word problems correctly. |
|---|
| Bertram Raphael's MIT dissertation on the SIR program demonstrates the power of a logical representation of knowledge for question-answering systems. |
| 1965 | Alexey Grigorevich Ivakhnenko and Valentin Lapa developed the first deep learning algorithm for multilayer perceptrons in Soviet Union.[32] [33] |
|---|
| Lotfi Zadeh at U.C. Berkeley publishes his first paper introducing fuzzy logic, "Fuzzy Sets" (Information and Control 8: 338–353). |
| J. Alan Robinson invented a mechanical proof procedure, the Resolution Method, which allowed programs to work efficiently with formal logic as a representation language. |
| Joseph Weizenbaum (MIT) built ELIZA, an interactive program that carries on a dialogue in English language on any topic. It was a popular toy at AI centers on the ARPANET when a version that "simulated" the dialogue of a psychotherapist was programmed. |
| Edward Feigenbaum initiated Dendral, a ten-year effort to develop software to deduce the molecular structure of organic compounds using scientific instrument data. It was the first expert system. |
| 1966 | Ross Quillian (PhD dissertation, Carnegie Inst. of Technology, now CMU) demonstrated semantic nets. |
|---|
| Machine Intelligence[34] workshop at Edinburgh – the first of an influential annual series organized by Donald Michie and others. |
| Negative report on machine translation kills much work in natural language processing (NLP) for many years. |
| Dendral program (Edward Feigenbaum, Joshua Lederberg, Bruce Buchanan, Georgia Sutherland at Stanford University) demonstrated to interpret mass spectra on organic chemical compounds. First successful knowledge-based program for scientific reasoning. |
| 1967 | Shun'ichi Amari was the first to use stochastic gradient descent for deep learning in multilayer perceptrons.[35] In computer experiments conducted by his student Saito, a five layer MLP with two modifiable layers learned useful internal representations to classify non-linearily separable pattern classes. |
|---|
| 1968 | Joel Moses (PhD work at MIT) demonstrated the power of symbolic reasoning for integration problems in the Macsyma program. First successful knowledge-based program in mathematics. |
|---|
| Richard Greenblatt (programmer) at MIT built a knowledge-based chess-playing program, MacHack, that was good enough to achieve a class-C rating in tournament play. |
| Wallace and Boulton's program, Snob (Comp.J. 11(2) 1968), for unsupervised classification (clustering) uses the Bayesian Minimum Message Length criterion, a mathematical realisation of Occam's razor. |
| 1969 | Stanford Research Institute (SRI): Shakey the Robot, demonstrated combining animal locomotion, perception and problem solving. |
|---|
| Roger Schank (Stanford) defined conceptual dependency model for natural-language understanding. Later developed (in PhD dissertations at Yale University) for use in story understanding by Robert Wilensky and Wendy Lehnert, and for use in understanding memory by Janet Kolodner. |
| Yorick Wilks (Stanford) developed the semantic coherence view of language called Preference Semantics, embodied in the first semantics-driven machine translation program, and the basis of many PhD dissertations since such as Bran Boguraev and David Carter at Cambridge. |
| First International Joint Conference on Artificial Intelligence (IJCAI) held at Stanford. |
| Marvin Minsky and Seymour Papert publish Perceptrons, demonstrating previously unrecognized limits of this feed-forward two-layered structure. This book is considered by some to mark the beginning of the AI winter of the 1970s, a failure of confidence and funding for AI. However, by the time the book came out, methods for training multilayer perceptrons by deep learning were already known (Alexey Grigorevich Ivakhnenko and Valentin Lapa, 1965; Shun'ichi Amari, 1967). Significant progress in the field continued (see below). |
| McCarthy and Hayes started the discussion about the frame problem with their essay, "Some Philosophical Problems from the Standpoint of Artificial Intelligence". |
| | |
|---|
| Date | Development |
|---|
| Early 1970s | Jane Robinson and Don Walker established an influential Natural Language Processing group at SRI.[36] |
|---|
| 1970 | Seppo Linnainmaa publishes the reverse mode of automatic differentiation. This method became later known as backpropagation, and is heavily used to train artificial neural networks.[37] |
|---|
| Jaime Carbonell (Sr.) developed SCHOLAR, an interactive program for computer assisted instruction based on semantic nets as the representation of knowledge. |
| Bill Woods described Augmented Transition Networks (ATN's) as a representation for natural language understanding. |
| Patrick Winston's PhD program, ARCH, at MIT learned concepts from examples in the world of children's blocks. |
| 1971 | Terry Winograd's PhD thesis (MIT) demonstrated the ability of computers to understand English sentences in a restricted world of children's blocks, in a coupling of his language understanding program, SHRDLU, with a robot arm that carried out instructions typed in English. |
|---|
| Work on the Boyer-Moore theorem prover started in Edinburgh.[38] |
| 1972 | Prolog programming language developed by Alain Colmerauer. |
|---|
| Earl Sacerdoti developed one of the first hierarchical planning programs, ABSTRIPS. |
| 1973 | The Assembly Robotics Group at University of Edinburgh builds Freddy Robot, capable of using visual perception to locate and assemble models. (See Edinburgh Freddy Assembly Robot: a versatile computer-controlled assembly system.) |
|---|
| The Lighthill report gives a largely negative verdict on AI research in Great Britain and forms the basis for the decision by the British government to discontinue support for AI research in all but two universities. |
| 1974 | Ted Shortliffe's PhD dissertation on the MYCIN program (Stanford) demonstrated a very practical rule-based approach to medical diagnoses, even in the presence of uncertainty. While it borrowed from DENDRAL, its own contributions strongly influenced the future of expert system development, especially commercial systems. |
|---|
| 1975 | Earl Sacerdoti developed techniques of partial-order planning in his NOAH system, replacing the previous paradigm of search among state space descriptions. NOAH was applied at SRI International to interactively diagnose and repair electromechanical systems. |
|---|
| Austin Tate developed the Nonlin hierarchical planning system able to search a space of partial plans characterised as alternative approaches to the underlying goal structure of the plan. |
| Marvin Minsky published his widely read and influential article on Frames as a representation of knowledge, in which many ideas about schemas and semantic links are brought together. |
| The Meta-Dendral learning program produced new results in chemistry (some rules of mass spectrometry) the first scientific discoveries by a computer to be published in a refereed journal. |
| Mid-1970s | Barbara Grosz (SRI) established limits to traditional AI approaches to discourse modeling. Subsequent work by Grosz, Bonnie Webber and Candace Sidner developed the notion of "centering", used in establishing focus of discourse and anaphoric references in Natural language processing. |
|---|
| David Marr and MIT colleagues describe the "primal sketch" and its role in visual perception. |
| 1976 | Douglas Lenat's AM program (Stanford PhD dissertation) demonstrated the discovery model (loosely guided search for interesting conjectures). |
|---|
| Randall Davis demonstrated the power of meta-level reasoning in his PhD dissertation at Stanford. |
| 1978 | Tom Mitchell, at Stanford, invented the concept of Version spaces for describing the search space of a concept formation program. |
|---|
| Herbert A. Simon wins the Nobel Prize in Economics for his theory of bounded rationality, one of the cornerstones of AI known as "satisficing". |
| The MOLGEN program, written at Stanford by Mark Stefik and Peter Friedland, demonstrated that an object-oriented programming representation of knowledge can be used to plan gene-cloning experiments. |
| 1979 | Bill VanMelle's PhD dissertation at Stanford demonstrated the generality of MYCIN's representation of knowledge and style of reasoning in his EMYCIN program, the model for many commercial expert system "shells". |
|---|
| Jack Myers and Harry Pople at University of Pittsburgh developed INTERNIST, a knowledge-based medical diagnosis program based on Dr. Myers' clinical knowledge. |
| Cordell Green, David Barstow, Elaine Kant and others at Stanford demonstrated the CHI system for automatic programming. |
| The Stanford Cart, built by Hans Moravec, becomes the first computer-controlled, autonomous vehicle when it successfully traverses a chair-filled room and circumnavigates the Stanford AI Lab. |
| BKG, a backgammon program written by Hans Berliner at CMU, defeats the reigning world champion (in part via luck). |
| Drew McDermott and Jon Doyle at MIT, and John McCarthy at Stanford begin publishing work on non-monotonic logics and formal aspects of truth maintenance. |
| Late 1970s | Stanford's SUMEX-AIM resource, headed by Ed Feigenbaum and Joshua Lederberg, demonstrates the power of the ARPAnet for scientific collaboration. | |
|---|
| Date | Development |
|---|
| 1980s | Lisp machines developed and marketed. First expert system shells and commercial applications. |
|---|
| 1980 | First National Conference of the American Association for Artificial Intelligence (AAAI) held at Stanford. |
|---|
| 1981 | Danny Hillis designs the connection machine, which utilizes parallel computing to bring new power to AI, and to computation in general. (Later founds Thinking Machines Corporation) |
|---|
| 1982 | The Fifth Generation Computer Systems project (FGCS), an initiative by Japan's Ministry of International Trade and Industry, begun in 1982, to create a "fifth generation computer" (see history of computing hardware) which was supposed to perform much calculation utilizing massive parallelism. |
|---|
| 1983 | John Laird and Paul Rosenbloom, working with Allen Newell, complete CMU dissertations on Soar (program). |
|---|
| James F. Allen invents the Interval Calculus, the first widely used formalization of temporal events. |
| Mid-1980s | Neural Networks become widely used with the Backpropagation algorithm, also known as the reverse mode of automatic differentiation published by Seppo Linnainmaa in 1970 and applied to neural networks by Paul Werbos. |
|---|
| 1985 | The autonomous drawing program, AARON, created by Harold Cohen, is demonstrated at the AAAI National Conference (based on more than a decade of work, and with subsequent work showing major developments). |
|---|
| 1986 | The team of Ernst Dickmanns at Bundeswehr University of Munich builds the first robot cars, driving up to 55 mph on empty streets. |
|---|
| Barbara Grosz and Candace Sidner create the first computation model of discourse, establishing the field of research.[39] |
| 1987 | Marvin Minsky published The Society of Mind, a theoretical description of the mind as a collection of cooperating agents. He had been lecturing on the idea for years before the book came out (c.f. Doyle 1983).[40] |
|---|
| Around the same time, Rodney Brooks introduced the subsumption architecture and behavior-based robotics as a more minimalist modular model of natural intelligence; Nouvelle AI. |
| Commercial launch of generation 2.0 of Alacrity by Alacritous Inc./Allstar Advice Inc. Toronto, the first commercial strategic and managerial advisory system. The system was based upon a forward-chaining, self-developed expert system with 3,000 rules about the evolution of markets and competitive strategies and co-authored by Alistair Davidson and Mary Chung, founders of the firm with the underlying engine developed by Paul Tarvydas. The Alacrity system also included a small financial expert system that interpreted financial statements and models.[41] |
| 1989 | The development of metal–oxide–semiconductor (MOS) Very Large Scale Integration (VLSI), in the form of complementary MOS (CMOS) technology, enabled the development of practical artificial neural network (ANN) technology in the 1980s. A landmark publication in the field was the 1989 book Analog VLSI Implementation of Neural Systems by Carver A. Mead and Mohammed Ismail.[42] |
|---|
| Dean Pomerleau at CMU creates ALVINN (An Autonomous Land Vehicle in a Neural Network), which was used in the Navlab program. | |
| Date | Development |
|---|
| 1990s | Major advances in all areas of AI, with significant demonstrations in machine learning, intelligent tutoring, case-based reasoning, multi-agent planning, scheduling, uncertain reasoning, data mining, natural language understanding and translation, vision, virtual reality, games, and other topics. |
|---|
| Early 1990s | TD-Gammon, a backgammon program written by Gerry Tesauro, demonstrates that reinforcement (learning) is powerful enough to create a championship-level game-playing program by competing favorably with world-class players. |
|---|
| 1991 | DART scheduling application deployed in the first Gulf War paid back DARPA's investment of 30 years in AI research.[43] |
|---|
| 1992 | Carol Stoker and NASA Ames robotics team explore marine life in Antarctica with an undersea robot Telepresence ROV operated from the ice near McMurdo Bay, Antarctica and remotely via satellite link from Moffett Field, California.[44] |
|---|
| 1993 | Ian Horswill extended behavior-based robotics by creating Polly, the first robot to navigate using vision and operate at animal-like speeds (1 meter/second). |
|---|
| Rodney Brooks, Lynn Andrea Stein and Cynthia Breazeal started the widely publicized MIT Cog project with numerous collaborators, in an attempt to build a humanoid robot child in just five years. |
| ISX corporation wins "DARPA contractor of the year"[45] for the Dynamic Analysis and Replanning Tool (DART) which reportedly repaid the US government's entire investment in AI research since the 1950s.[46] |
| 1994 | Lotfi Zadeh at U.C. Berkeley creates "soft computing"[47] and builds a world network of research with a fusion of neural science and neural net systems, fuzzy set theory and fuzzy systems, evolutionary algorithms, genetic programming, and chaos theory and chaotic systems ("Fuzzy Logic, Neural Networks, and Soft Computing", Communications of the ACM, March 1994, Vol. 37 No. 3, pages 77–84). |
|---|
| With passengers on board, the twin robot cars VaMP and VITA-2 of Ernst Dickmanns and Daimler-Benz drive more than one thousand kilometers on a Paris three-lane highway in standard heavy traffic at speeds up to 130 km/h. They demonstrate autonomous driving in free lanes, convoy driving, and lane changes left and right with autonomous passing of other cars. |
| English draughts (checkers) world champion Tinsley resigned a match against computer program Chinook. Chinook defeated 2nd highest rated player, Lafferty. Chinook won the USA National Tournament by the widest margin ever. |
| Cindy Mason at NASA organizes the First AAAI Workshop on AI and the Environment.[48] |
| 1995 | Cindy Mason at NASA organizes the First International IJCAI Workshop on AI and the Environment.[49] |
|---|
| "No Hands Across America": A semi-autonomous car drove coast-to-coast across the United States with computer-controlled steering for 2797miles of the 2849miles. Throttle and brakes were controlled by a human driver.[50] [51] |
| One of Ernst Dickmanns' robot cars (with robot-controlled throttle and brakes) drove more than 1000 miles from Munich to Copenhagen and back, in traffic, at up to 120 mph, occasionally executing maneuvers to pass other cars (only in a few critical situations a safety driver took over). Active vision was used to deal with rapidly changing street scenes. |
| 1996 | Steve Grand, roboticist and computer scientist, develops and releases Creatures, a popular simulation of artificial life-forms with simulated biochemistry, neurology with learning algorithms and inheritable digital DNA. |
|---|
| 1997 | The Deep Blue chess machine (IBM) defeats the (then) world chess champion, Garry Kasparov. |
|---|
| First official RoboCup football (soccer) match featuring table-top matches with 40 teams of interacting robots and over 5000 spectators. |
| Computer Othello program Logistello defeated the world champion Takeshi Murakami with a score of 6–0. |
| Long short-term memory (LSTM) was published in Neural Computation by Sepp Hochreiter and Juergen Schmidhuber.[52] |
| 1998 | Tiger Electronics' Furby is released, and becomes the first successful attempt at producing a type of A.I to reach a domestic environment. |
|---|
| Tim Berners-Lee published his Semantic Web Road map paper.[53] |
| Ulises Cortés and Miquel Sànchez-Marrè organize the first Environment and AI Workshop in Europe ECAI, "Binding Environmental Sciences and Artificial Intelligence".[54] [55] |
| Leslie P. Kaelbling, Michael Littman, and Anthony Cassandra introduce POMDPs and a scalable method for solving them to the AI community, jumpstarting widespread use in robotics and automated planning and scheduling[56] |
| 1999 | Sony introduces an improved domestic robot similar to a Furby, the AIBO becomes one of the first artificially intelligent "pets" that is also autonomous. |
|---|
| Late 1990s | Web crawlers and other AI-based information extraction programs become essential in widespread use of the World Wide Web. |
|---|
| Demonstration of an Intelligent room and Emotional Agents at MIT's AI Lab. |
| Initiation of work on the Oxygen architecture, which connects mobile and stationary computers in an adaptive network. | |
| Date | Development |
|---|
| 2000 | Interactive robopets ("smart toys") become commercially available, realizing the vision of the 18th century novelty toy makers. |
|---|
| Cynthia Breazeal at MIT publishes her dissertation on Sociable machines, describing Kismet (robot), with a face that expresses emotions. |
| The Nomad robot explores remote regions of Antarctica looking for meteorite samples. |
| 2002 | iRobot's Roomba autonomously vacuums the floor while navigating and avoiding obstacles. |
|---|
| 2004 | OWL Web Ontology Language W3C Recommendation (10 February 2004). |
|---|
| DARPA introduces the DARPA Grand Challenge requiring competitors to produce autonomous vehicles for prize money. |
| NASA's robotic exploration rovers Spirit and Opportunity autonomously navigate the surface of Mars. |
| 2005 | Honda's ASIMO robot, an artificially intelligent humanoid robot, is able to walk as fast as a human, delivering trays to customers in restaurant settings. |
|---|
| Recommendation technology based on tracking web activity or media usage brings AI to marketing. See TiVo Suggestions. |
| Blue Brain is born, a project to simulate the brain at molecular detail.[57] |
| 2006 | The Dartmouth Artificial Intelligence Conference: The Next 50 Years (AI@50) AI@50 (14–16 July 2006) |
|---|
| 2007 | Philosophical Transactions of the Royal Society, B – Biology, one of the world's oldest scientific journals, puts out a special issue on using AI to understand biological intelligence, titled Models of Natural Action Selection[58] |
|---|
| Checkers is solved by a team of researchers at the University of Alberta. |
| DARPA launches the Urban Challenge for autonomous cars to obey traffic rules and operate in an urban environment. |
| 2008 | Cynthia Mason at Stanford presents her idea on Artificial Compassionate Intelligence, in her paper on "Giving Robots Compassion".[59] |
|---|
| 2009 | An LSTM trained by connectionist temporal classification[60] was the first recurrent neural network to win pattern recognition contests, winning three competitions in connected handwriting recognition.[61] |
|---|
| 2009 | Google builds autonomous car.[62] | |
|---|
| Date | Development |
|---|
| 2010 | Microsoft launched Kinect for Xbox 360, the first gaming device to track human body movement, using just a 3D camera and infra-red detection, enabling users to play their Xbox 360 wirelessly. The award-winning machine learning for human motion capture technology for this device was developed by the Computer Vision group at Microsoft Research, Cambridge.[63] [64] |
|---|
| 2011 | Mary Lou Maher and Doug Fisher organize the First AAAI Workshop on AI and Sustainability.[65] |
|---|
| IBM's Watson computer defeated television game show Jeopardy! champions Rutter and Jennings. |
| 2011–2014 | Apple's Siri (2011), Google's Google Now (2012) and Microsoft's Cortana (2014) are smartphone apps that use natural language to answer questions, make recommendations and perform actions. |
|---|
| 2012 | AlexNet, a deep learning model developed by Alex Krizhevsky, wins the ImageNet Large Scale Visual Recognition Challenge with half as many errors as the second-place winner. This is a turning point in the history of AI; over the next few years dozens of other approaches to image recognition were abandoned in favor of deep learning. Krizhevsky is among the first to use GPU chips to train a deep learning network. |
|---|
| 2013 | Robot HRP-2 built by SCHAFT Inc of Japan, a subsidiary of Google, defeats 15 teams to win DARPA’s Robotics Challenge Trials. HRP-2 scored 27 out of 32 points in eight tasks needed in disaster response. Tasks are drive a vehicle, walk over debris, climb a ladder, remove debris, walk through doors, cut through a wall, close valves and connect a hose.[66] |
|---|
| NEIL, the Never Ending Image Learner, is released at Carnegie Mellon University to constantly compare and analyze relationships between different images.[67] |
| 2015 | Two techniques were developed concurrently to train very deep networks: highway network,[68] and the residual neural network (ResNet).[69] They allowed over 1000-layers-deep networks to be trained. |
|---|
| In January 2015, Stephen Hawking, Elon Musk, and dozens of artificial intelligence experts signed an open letter on artificial intelligence calling for research on the societal impacts of AI.[70] [71] |
| In July 2015, an open letter to ban development and use of autonomous weapons was signed by Hawking, Musk, Wozniak and 3,000 researchers in AI and robotics.[72] |
| Google DeepMind's AlphaGo (version: Fan)[73] defeated three-time European Go champion 2 dan professional Fan Hui by 5 games to 0.[74] |
| 2016 | Google DeepMind's AlphaGo (version: Lee) defeated Lee Sedol 4–1. Lee Sedol is a 9 dan professional Korean Go champion who won 27 major tournaments from 2002 to 2016.[75] |
|---|
| 2017 | Asilomar Conference on Beneficial AI was held, to discuss AI ethics and how to bring about beneficial AI while avoiding the existential risk from artificial general intelligence. |
|---|
| Deepstack[76] is the first published algorithm to beat human players in imperfect information games, as shown with statistical significance on heads-up no-limit poker. Soon after, the poker AI Libratus by different research group individually defeated each of its four-human opponents—among the best players in the world—at an exceptionally high aggregated winrate, over a statistically significant sample.[77] In contrast to Chess and Go, Poker is an imperfect information game.[78] |
| In May 2017, Google DeepMind's AlphaGo (version: Master) beat Ke Jie, who at the time continuously held the world No. 1 ranking for two years,[79] [80] winning each game in a three-game match during the Future of Go Summit.[81] [82] |
| A propositional logic boolean satisfiability problem (SAT) solver proves a long-standing mathematical conjecture on Pythagorean triples over the set of integers. The initial proof, 200TB long, was checked by two independent certified automatic proof checkers.[83] |
| An OpenAI bot using machine learning played at The International 2017 Dota 2 tournament in August 2017. It won during a 1v1 demonstration game against professional Dota 2 player Dendi.[84] |
| Google Lens image analysis and comparison tool released in October 2017, associates millions of landscapes, artworks, products and species to their text description. |
| Google DeepMind revealed that AlphaGo Zero—an improved version of AlphaGo—displayed significant performance gains while using far fewer tensor processing units (as compared to AlphaGo Lee; it used same amount of TPU's as AlphaGo Master). Unlike previous versions, which learned the game by observing millions of human moves, AlphaGo Zero learned by playing only against itself. The system then defeated AlphaGo Lee 100 games to zero, and defeated AlphaGo Master 89 to 11. Although unsupervised learning is a step forward, much has yet to be learned about general intelligence.[85] AlphaZero masters chess in four hours, defeating the best chess engine, StockFish 8. AlphaZero won 28 out of 100 games, and the remaining 72 games ended in a draw. |
| Transformer architecture was invented, which led to new kinds of large language models such as BERT by Google, followed by the generative pre-trained transformer type of model introduced by OpenAI. |
| 2018 | Alibaba language processing AI outscores top humans at a Stanford University reading and comprehension test, scoring 82.44 against 82.304 on a set of 100,000 questions.[86] |
|---|
| The European Lab for Learning and Intelligent Systems (aka Ellis) proposed as a pan-European competitor to American AI efforts, with the aim of staving off a brain drain of talent, along the lines of CERN after World War II.[87] |
| Announcement of Google Duplex, a service to allow an AI assistant to book appointments over the phone. The Los Angeles Times judges the AI's voice to be a "nearly flawless" imitation of human-sounding speech.[88] |
| 2019 | DeepMind's AlphaStar reaches Grandmaster level at StarCraft II, outperforming 99.8 percent of human players.[89] | |
|---|
| Date | Development |
|---|
| 2020 | In February 2020, Microsoft introduces its Turing Natural Language Generation (T-NLG), which is the "largest language model ever published at 17 billion parameters".[90] |
|---|
| In November 2020, AlphaFold 2 by DeepMind, a model that performs predictions of protein structure, wins the CASP competition.[91] |
| OpenAI introduces GPT-3, a state-of-the-art autoregressive language model that uses deep learning to produce a variety of computer codes, poetry and other language tasks exceptionally similar, and almost indistinguishable from those written by humans. Its capacity was ten times greater than that of the T-NLG. It was introduced in May 2020,[92] and was in beta testing in June 2020. |
| 2022 | ChatGPT, an AI chatbot developed by OpenAI, debuts in November 2022. It is initially built on top of the large language model. While it gains considerable praise for the breadth of its knowledge base, deductive abilities, and the human-like fluidity of its natural language responses,[93] [94] it also garners criticism for, among other things, its tendency to "hallucinate",[95] [96] a phenomenon in which an AI responds with factually incorrect answers with high confidence. The release triggers widespread public discussion on artificial intelligence and its potential impact on society.[97] [98] |
|---|
| A November 2022 class action lawsuit against Microsoft, GitHub and OpenAI alleges that GitHub Copilot, an AI-powered code editing tool trained on public GitHub repositories, violates the copyrights of the repositories' authors, noting that the tool is able to generate source code which matches its training data verbatim, without providing attribution.[99] |
| 2023 | By January 2023, ChatGPT has more than 100 million users, making it the fastest-growing consumer application to date.[100] |
|---|
| On January 16, 2023, three artists, Sarah Andersen, Kelly McKernan, and Karla Ortiz, file a class-action copyright infringement lawsuit against Stability AI, Midjourney, and DeviantArt, claiming that these companies have infringed the rights of millions of artists by training AI tools on five billion images scraped from the web without the consent of the original artists.[101] |
| On January 17, 2023, Stability AI is sued in London by Getty Images for using its images in their training data without purchasing a license.[102] [103] |
| Getty files another suit against Stability AI in a US district court in Delaware on February 6, 2023. In the suit, Getty again alleges copyright infringement for the use of its images in the training of Stable Diffusion, and further argues that the model infringes Getty's trademark by generating images with Getty's watermark.[104] |
| OpenAI's model is released in March 2023 and is regarded as an impressive improvement over, with the caveat that GPT-4 retains many of the same problems of the earlier iteration.[105] Unlike previous iterations, GPT-4 is multimodal, allowing image input as well as text. GPT-4 is integrated into ChatGPT as a subscriber service. OpenAI claims that in their own testing the model received a score of 1410 on the SAT (94th percentile),[106] 163 on the LSAT (88th percentile), and 298 on the Uniform Bar Exam (90th percentile).[107] |
| On March 7, 2023, Nature Biomedical Engineering writes that "it is no longer possible to accurately distinguish" human-written text from text created by large language models, and that "It is all but certain that general-purpose large language models will rapidly proliferate... It is a rather safe bet that they will change many industries over time."[108] |
| In response to ChatGPT, Google releases in a limited capacity its chatbot Google Bard, based on the LaMDA and PaLM large language models, in March 2023.[109] [110] |
| On March 29, 2023, a petition of over 1,000 signatures is signed by Elon Musk, Steve Wozniak and other tech leaders, calling for a 6-month halt to what the petition refers to as "an out-of-control race" producing AI systems that its creators can not "understand, predict, or reliably control".[111] [112] |
| In May 2023, Google makes an announcement regarding Bard's transition from LaMDA to PaLM2, a significantly more advanced language model.[113] |
| In the last week of May 2023, a Statement on AI Risk is signed by Geoffrey Hinton, Sam Altman, Bill Gates, and many other prominent AI researchers and tech leaders with the following succinct message: "Mitigating the risk of extinction from AI should be a global priority alongside other societal-scale risks such as pandemics and nuclear war."[114] [115] |
| On July 9, 2023, Sarah Silverman files a class action lawsuit against Meta and OpenAI for copyright infringement for training their large language models on millions of authors' copyrighted works without permission.[116] |
| In August, 2023, the New York Times, CNN, Reuters, the Chicago Tribune, Australian Broadcasting Corporation (ABC) and other news companies block OpenAI's GPTBot web crawler from accessing their content, while the New York Times also updates its terms of service to disallow the use of its content in large language models.[117] |
| On September 13, 2023, in a serious response to growing anxiety about the dangers of AI, the US Senate holds the inaugural bipartisan "AI Insight Forum", bringing together senators, CEOs, civil rights leaders and other industry reps, to further familiarize senators with the nature of AI and its risks, and to discuss needed safeguards and legislation.[118] The event is organized by Senate Majority Leader Chuck Schumer (D-NY),[119] and is chaired by U.S. Senator Martin Heinrich (D-N.M.), Founder and co-chair of the Senate AI Caucus. Reflecting the importance of the meeting, the forum is attended by over 60 senators,[120] as well as Elon Musk (Tesla CEO), Mark Zuckerberg (Meta CEO), Sam Altman (OpenAI CEO), Sundar Pichai (Alphabet CEO), Bill Gates (Microsoft co-founder), Satya Nadella (Microsoft CEO), Jensen Huang (Nvidia CEO), Arvind Krishna (IBM CEO), Alex Karp (Palantir CEO), Charles Rivkin (chairman and CEO of the MPA), Meredith Stiehm (president of the Writers Guild of America West), Liz Shuler (AFL-CIO President), and Maya Wiley (CEO of the Leadership Conference on Civil and Human Rights), among others. |
| On October 30, 2023, US President Biden signed the Executive Order on the Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence.[121] |
| In November 2023, the first global AI Safety Summit was held in Bletchley Park in the UK to discuss the near and far term risks of AI and the possibility of mandatory and voluntary regulatory frameworks.[122] 28 countries including the United States, China, and the European Union issued a declaration at the start of the summit, calling for international co-operation to manage the challenges and risks of artificial intelligence.[123] [124] |
| Google releases Gemini 1.0 Ultra. |
| 2024 | On February 15, 2024, Google releases Gemini 1.5 in limited beta, capable of context length up to 1 million tokens. |
|---|
| Also, on February 15, 2024, OpenAI publicly announces Sora, a text-to-video model for generating videos up to a minute long. |
| Google DeepMind unveils DNA prediction software AlphaFold which helps to identify cancer and genetic diseases. |
| On 22 February, StabilityAI announces Stable Diffusion 3, using a similar architecture to Sora. |
| On June 10, Apple announced "Apple Intelligence" which incorporates ChatGPT into new IPhones and Siri. | |
