Specialty Digest

Global English editionLearners’ Study Dashboard — in development
The Knowledge Library for All

History of Artificial Intelligence

nounArtificial Intelligencealso AI history, timeline of artificial intelligence

In one line

The history of artificial intelligence traces the effort to build thinking machines from ancient myths and early logic through the 1956 Dartmouth workshop, two AI winters and the deep learning boom.

The history of artificial intelligence is the story of humanity’s effort to build machines that reason, learn and act intelligently. It stretches from ancient myths of artificial beings and early work in logic to the founding of the field in 1956, repeated cycles of excitement and disappointment, and the rapid advances in machine learning that began around 2012. Understanding this history helps explain why today’s methods look the way they do and why predictions about AI have so often swung between hype and doubt.

Early ideas and precursors

Long before computers, people imagined artificial minds. Greek myths told of the bronze giant Talos, and medieval and early modern inventors built mechanical automata. The idea that reasoning could be formalized came from logic. Aristotle described rules of valid argument, Gottfried Leibniz dreamed of a calculus of reasoning, and in the 1800s George Boole developed algebraic logic. Charles Babbage designed programmable calculating machines, and Ada Lovelace wrote that such a machine could in principle manipulate symbols, though she doubted it could originate ideas. In 1936, Alan Turing defined the abstract computing machine that bears his name, and in 1943 Warren McCulloch and Walter Pitts proposed a mathematical model of neurons.

Birth of the field, 1950 to 1956

In 1950, Alan Turing published Computing Machinery and Intelligence, which proposed what became the Turing test. Early programs showed that machines could play checkers and prove theorems. In 1955, John McCarthy, Marvin Minsky, Nathaniel Rochester and Claude Shannon proposed a summer research project at Dartmouth College, using the term artificial intelligence for the first time. The workshop of 1956 brought together researchers who would lead the field, including Allen Newell and Herbert Simon, whose Logic Theorist was demonstrated there. The founders were optimistic, and some predicted that machines matching human intelligence would arrive within a generation.

Early successes and the first AI winter

The late 1950s and 1960s produced a stream of achievements. McCarthy created the LISP language in 1958, Frank Rosenblatt demonstrated the perceptron, a learning neural network, and Joseph Weizenbaum wrote the chatbot ELIZA in 1966. The robot Shakey at the Stanford Research Institute combined perception, planning and movement. Progress slowed when programs proved unable to scale beyond toy problems and computers were too weak. In 1969, Minsky and Papert exposed limits of simple perceptrons. A critical British report by James Lighthill in 1973 and cuts in United States government funding led to the first AI winter, a period of reduced support and interest in the mid-1970s.

Expert systems and the second winter

AI revived in the 1980s through expert systems that captured specialist knowledge as rules. The commercial success of programs such as XCON led companies to invest in AI software and specialized Lisp machines, and Japan launched its ambitious Fifth Generation Computer Systems project in 1982. At the same time, neural networks returned after the 1986 popularization of backpropagation. The boom ended in the late 1980s, when expert systems proved costly to maintain and the specialized hardware market collapsed, bringing a second AI winter that lasted into the early 1990s.

Statistical methods and milestones, 1990s to 2000s

Researchers turned to statistical and probabilistic methods and to machine learning, using growing amounts of data. Support vector machines, Bayesian networks and speech recognition systems based on hidden Markov models improved results. In 1997, IBM’s Deep Blue defeated world chess champion Garry Kasparov. Robots and vehicles advanced in competitions funded by the United States Defense Advanced Research Projects Agency, and in 2005 a Stanford vehicle completed a 132-mile desert course autonomously. In 2006, Geoffrey Hinton and colleagues published methods for training deep networks, helping to launch the deep learning movement.

The deep learning revolution, 2010s

Three ingredients came together: large datasets such as ImageNet, powerful graphics processors and improved training techniques. In 2012, a network called AlexNet, built by Alex Krizhevsky, Ilya Sutskever and Hinton, won the ImageNet image recognition contest by a wide margin, and deep learning rapidly spread to speech, translation and many other tasks. In 2011, IBM’s Watson won the quiz show Jeopardy. In 2014, Ian Goodfellow and colleagues introduced generative adversarial networks, and in 2016 DeepMind’s AlphaGo defeated the Go champion Lee Sedol. In 2017, researchers at Google published Attention Is All You Need, which introduced the transformer architecture.

The generative AI era, 2020s

Transformers enabled large language models. OpenAI released GPT-3 in 2020, and in November 2022 it launched ChatGPT, which reached an enormous audience within weeks and made generative AI a mainstream topic. Image generators such as DALL-E, Midjourney and Stable Diffusion appeared at the same time. Competition among technology companies and open-source communities accelerated. Governments responded with policy, including the European Union’s Artificial Intelligence Act, which entered into force in 2024, and international summits on AI safety. Debates over bias, misinformation, copyright, jobs and the possibility of artificial general intelligence moved into public life.

Lessons from the history

Two patterns repeat. First, enthusiasm often runs ahead of capability, leading to winters when promises are not met. Second, progress has come from combining ideas, data and computing power, not from a single breakthrough. Approaches once abandoned, such as neural networks, returned when conditions changed. Historians of the field also note that tasks once considered signs of intelligence, such as playing chess, are often dismissed as ordinary computing once machines can do them, a phenomenon called the AI effect.

Key figures in the history of AI

Many people shaped the field. Alan Turing laid the theoretical groundwork and posed the question of machine thinking. John McCarthy named the field and created LISP. Marvin Minsky co-founded the MIT AI laboratory and wrote on neural networks and the mind. Herbert Simon and Allen Newell built early reasoning programs and studied human problem solving. Frank Rosenblatt created the perceptron. Geoffrey Hinton, Yann LeCun and Yoshua Bengio pioneered deep learning, and the three shared the 2018 Turing Award for it. Fei-Fei Li led the creation of ImageNet, the dataset that helped spark the deep learning boom, and Demis Hassabis led DeepMind’s work on game-playing and protein structure prediction.

FAQs

Who is the father of artificial intelligence? John McCarthy, who coined the term, is often given the title, along with Marvin Minsky, Alan Turing and others who shaped the field.

What was the AI winter? An AI winter is a period of reduced funding and interest after hopes for AI outran results. There were two major ones, in the mid-1970s and the late 1980s.

When did modern AI begin? Many date it to 2012, when deep learning won the ImageNet contest, though its roots go back to the 1950s.

Key facts

  • The term artificial intelligence was coined for a 1956 summer workshop at Dartmouth College proposed by John McCarthy, Marvin Minsky, Nathaniel Rochester and Claude Shannon.1
  • Funding and enthusiasm collapsed twice, in the mid-1970s and the late 1980s, periods known as AI winters.2
  • The 2017 paper Attention Is All You Need introduced the transformer architecture used by modern large language models.3

Related concepts

Quick checkWhere and when was the term artificial intelligence coined?Show answer

At the Dartmouth summer research workshop in 1956, proposed by John McCarthy, Marvin Minsky, Nathaniel Rochester and Claude Shannon.

Sources

  1. Britannica, Artificial intelligence, https://www.britannica.com/technology/artificial-intelligence
  2. Wikipedia, History of artificial intelligence, https://en.wikipedia.org/wiki/History_of_artificial_intelligence
  3. Vaswani et al., Attention Is All You Need, https://arxiv.org/abs/1706.03762

Editorially reviewed by Specialty Digest Editorial TeamLast reviewed November 17, 2026Researched and drafted with AI assistanceReport an issue