History of AI: Complete Timeline From 1943 to 2026

Quick dinner-party fact. AI is older than the personal computer, the internet, and most of the tech in your pocket.

Yet many people first met it in a chat window. So AI showed up last Tuesday.

It didn’t. The history of artificial intelligence spans more than 80 years, with big promises, bigger crashes, and one very stubborn comeback. This guide walks through the whole story in plain English, whether you’re a beginner or just curious how we got here.

The history of AI began with early ideas about thinking machines in the 1940s and 1950s. The field got its name at a 1956 workshop at Dartmouth College. It then survived two “AI winters,” when funding dried up, before deep learning revived it in 2012 and ChatGPT brought it to the public in 2022.

New to the topic? Start with our guide on what AI is.

AI History Timeline at a Glance

Here’s the AI timeline from 1950 to today, and a few earlier moments, in one table.

Now let’s fill in the story behind each row.

How Artificial Intelligence Started (1943-1955)

So how did AI start? Not with a bang. It began with a question: can a machine think?

• 1943: The first mathematical model of an artificial neuron appears. It’s a simple idea, but every neural network since traces back to it.
• 1951: Marvin Minsky and Dean Edmonds build the first artificial neural network machine.
• 1952: Arthur Samuel at IBM writes a checkers program that plays and learns on its own.

In these years, scientists from math, psychology, and engineering were all circling the same puzzle. Norbert Wiener’s cybernetics, Claude Shannon’s information theory and Alan Turing’s theory of computation all hinted that a machine might copy some abilities of the brain.

Nobody had a shared name for the field yet. Terms like cybernetics and automata theory competed for attention.

Alan Turing and the Turing Test

If AI has a founding thinker, it’s Alan Turing. His contribution to artificial intelligence was mostly a question, and it was a good one.

In 1950, Turing published “Computing Machinery and Intelligence.” He asked whether machines can think and proposed the Imitation Game, now called the Turing Test. A machine passes if a person chatting with it can’t tell whether they’re talking to a person or a computer.

The test became one of the best-known ideas in AI history. Researchers still argue about what it proves, but it gave the field a target, and it gave everyone a great excuse to argue at parties.

The 1956 Dartmouth Conference

Why was the Dartmouth conference so important to AI? Because it gave the field a name and a home.

In the summer of 1956, John McCarthy, then a professor at Dartmouth College, organized a workshop on thinking machines. He proposed it with Marvin Minsky, Nathaniel Rochester of IBM, and Claude Shannon of Bell Labs. McCarthy chose the name “artificial intelligence.”

The proposal set an ambitious goal. Figure out how to make machines use language, form concepts, solve problems people usually handle, and improve themselves.

That’s why most historians treat 1956 as the birth of AI as a research field. The same year, Allen Newell, Herbert Simon, and Cliff Shaw’s Logic Theorist debuted and proved theorems in symbolic logic.

Early AI Systems: The First AI Program, Chatbot and Robot

After Dartmouth, the field moved fast. People often ask which was the first AI program. The honest answer depends on how you count.

• Logic Theorist (1956) is often called the first AI program because it proved math theorems using symbolic reasoning.
• Samuel’s checkers program (1952) is widely credited as an early self-learning game program. Samuel also introduced the term “machine learning” in a 1959 paper.
• The perceptron (1958) was Frank Rosenblatt’s early neural network model, a bridge between early AI research and today’s deep learning.
• SAINT (1961) was James Slagle’s program that solved calculus problems. It’s considered one of the first expert systems.
• ELIZA (1966) was Joseph Weizenbaum’s program at MIT. It imitated a therapist by pattern matching, and many people treat it as the first AI chatbot. It knew almost nothing, yet some users felt it understood them.
• Shakey (1966-1972) was a robot from the Stanford Research Institute, the first general-purpose mobile robot that could reason about its own actions.

So the first AI system, the first chatbot, and the first AI robot each have a different winner. That’s why history nerds enjoy this topic so much.

The history of AI chatbots picks up again later with A.L.I.C.E. in the 1990s and, much later, ChatGPT.

The First AI Winter

Big promises met hard reality. Programs that dazzled on small demos fell apart on real-world problems.

In 1969, Minsky and Seymour Papert published Perceptrons, which laid out the limits of simple neural networks. Neural network research slowed while symbolic AI thrived. Then in 1973, James Lighthill delivered a report to the British Science Research Council that criticized AI for missing its grand promises. UK funding fell.

That began the first AI winter, roughly 1974 to 1980. An AI winter is a period of reduced funding, interest, and progress after expectations outrun what the technology can do.

What caused the AI winter? Mostly three things:

• Overpromising by researchers and the press
• Limited computing power and data
• Systems that worked in the lab but not in the real world

Think of it as the morning after a party where everyone promised to start a business together.

Expert Systems and the Second AI Winter

AI bounced back in the 1980s thanks to expert systems. These programs stored the knowledge of human specialists as IF-THEN rules.

XCON, the first commercial expert system, helped configure computer orders starting in 1980. MYCIN targeted medical diagnosis. For a while, companies paid real money for these tools.

Then the costs showed up. Expert systems were expensive to build and even pricier to maintain, because every update needed a human expert. They were also brittle, which means they broke in any case their rules didn’t cover.

The bubble popped, and the second AI winter ran from roughly 1987 to 1993. When it ended, the field had shifted away from hand-written rules and toward statistical machine learning.

Here are the AI winter causes and effects at a glance:

Neural Networks, Machine Learning and Deep Learning

While the winters chilled the headlines, quieter progress kept going. This is where the story turns from classic AI history to modern AI.

• 1969: Arthur Bryson and Yu-Chi Ho describe a backpropagation learning algorithm for multilayer neural networks.
• 1986: Backpropagation, a method for training neural networks by learning from mistakes, gains wide attention. Geoffrey Hinton helped popularize it.
• 1990s: Machine learning grows into the largest branch of AI. Systems learn patterns from data instead of following hand-written rules.
• 1997: IBM’s Deep Blue beats world chess champion Garry Kasparov. It’s a huge headline, though Deep Blue was built for chess and nothing else.
• 2011: IBM Watson beats the all-time Jeopardy! champion.

Want to dig into the tech? See our guides on machine learning and deep learning.

Modern AI History: 2012 to 2022

Three ingredients finally lined up: lots of data, powerful graphics chips, and better neural network methods.

• 2009: Fei-Fei Li’s team completes ImageNet, a huge labeled image dataset.
• 2012: A neural network called AlexNet wins the ImageNet contest by a wide margin. Before it, almost none of the leading computer vision papers used neural networks. After it, nearly all did.
• 2014: Ian Goodfellow introduces generative adversarial networks (GANs).
• 2016: Google DeepMind’s AlphaGo beats Go champion Lee Sedol 4 games to 1.
• 2017: Google researchers publish “Attention Is All You Need” and introduce the Transformer. Every modern large language model builds on it.
• 2018: Geoffrey Hinton, Yann LeCun and Yoshua Bengio win the ACM Turing Award for their deep learning work.
• 2020: OpenAI releases GPT-3 with 175 billion parameters. DeepMind’s AlphaFold 2 takes a huge leap in predicting protein structures.

The AlexNet team was Alex Krizhevsky, Ilya Sutskever, and Geoffrey Hinton. Hinton later joked about credit. Ilya had the idea, Alex made it work, and Hinton got the Nobel Prize.

This is the AI history before ChatGPT, and it’s the part most people skip.

ChatGPT and Generative AI

The history of generative AI runs from GANs in 2014 to image models and large language models around 2021 and 2022. Then came the tipping point.

On November 30, 2022, OpenAI released ChatGPT as a free research preview. It reached an estimated 100 million users in about two months, one of the fastest adoptions of any app.

Suddenly, everyone had an opinion about AI. Your cousin. Your boss. Your dentist.

The race was on. In 2023, models like GPT-4, Claude, Bard, and Llama 2 opened the frontier model race. In January 2025, Chinese lab DeepSeek released R1, an openly licensed reasoning model that rivaled top Western systems and reportedly cost far less to train.

Science noticed too. In October 2024, John Hopfield and Geoffrey Hinton won the Nobel Prize in Physics for their work on artificial neural networks. Demis Hassabis, John Jumper and David Baker shared the Chemistry prize for AI-driven protein structure prediction.

Regulation followed. The EU AI Act (Regulation (EU) 2024/1689) became the first comprehensive legal framework on AI in the world.

You can read how the tech works in our Generative AI guide.

Where AI Stands in 2026

Now the newest chapter. Stanford’s 2026 AI Index Report gives a snapshot.

• Fast adoption. Generative AI reached 53% population adoption within three years, faster than the personal computer or the internet.
• Corporate money. Global corporate AI investment hit $581.7 billion in 2025. Private investment on its own was $344.7 billion.
• Growing pains. Documented AI incidents rose from 233 in 2024 to 362 in 2025.

Wikipedia’s ChatGPT entry, citing company figures, reports 900 million weekly active users in February 2026. Treat that as reported, not audited.

More in our AI Statistics roundup and our look at the future of ai.

AI Pioneers: Who Made This Happen?

AI has many parents. Here are the famous AI pioneers and what they’re known for.

What the History of AI Teaches Us

History repeats. AI history repeats a lot.

• Hype runs ahead of reality. Both AI winters followed big promises that the technology couldn’t keep.
• Data and computing power matter. Deep learning worked in 2012 because data and chips finally caught up with old ideas.
• Narrow wins beat grand dreams. Deep Blue, AlphaGo and today’s chatbots do specific jobs well. None is a general thinker.
• Comebacks happen. Neural networks were once unfashionable. Today, they power nearly everything.

So the next time someone says AI is a bubble, or the future of everything, remember that people have said both before.

FAQ

When did artificial intelligence begin? Early ideas date to the 1940s, and Alan Turing’s 1950 paper set the agenda. The field formally started at the Dartmouth workshop in 1956.

Who invented AI? No single person. Alan Turing shaped the early ideas, and John McCarthy organized the 1956 Dartmouth workshop with Marvin Minsky, Nathaniel Rochester, and Claude Shannon.

Who coined the term artificial intelligence? John McCarthy chose the name “artificial intelligence” for the 1956 Dartmouth project.

What was the first AI program? Many historians point to Logic Theorist (1956). Arthur Samuel’s checkers program (1952) is also an early self-learning program.

What was the first AI chatbot? ELIZA, created by Joseph Weizenbaum at MIT in 1966.

What was the AI winter? It was a period of reduced funding and interest after AI missed its early promises. There were two, roughly 1974 to 1980 and 1987 to 1993.

What caused the second AI winter? Expert systems proved expensive to build, costly to maintain, and brittle. The market collapsed around 1987.

What is the Turing Test? It’s a test Alan Turing proposed in 1950. A machine passes if a person can’t tell its conversation from another person’s.

When did modern AI begin? Many experts point to 2012, when AlexNet’s win sparked the deep learning boom. The public saw the boom explode with ChatGPT in November 2022.

Sources

• Alan Turing, “Computing Machinery and Intelligence,” Mind, 1950
• McCarthy, Minsky, Rochester, Shannon, A Proposal for the Dartmouth Summer Research Project on Artificial Intelligence, 1955
• GovTech Singapore, The History of AI, Part 1: https://www.tech.gov.sg/technews/the-history-of-ai-part-1/
• TechTarget, The History of Artificial Intelligence: Complete AI Timeline: https://www.techtarget.com/ai/tip/The-history-of-artificial-intelligence-Complete-AI-timeline
• TechTarget, History and Evolution of Machine Learning: A Timeline: https://www.techtarget.com/whatis/feature/History-and-evolution-of-machine-learning-A-timeline
• IEEE Spectrum, How AlexNet Transformed AI and Computer Vision Forever: https://spectrum.ieee.org/alexnet-source-code
• Vaswani et al., “Attention Is All You Need,” 2017
• Stanford HAI, The 2026 AI Index Report: https://hai.stanford.edu/ai-index/2026-ai-index-report
• Nobel Prize announcements, Physics and Chemistry, October 2024
• European Commission, AI Act (Regulation (EU) 2024/1689): https://digital-strategy.ec.europa.eu/en/policies/regulatory-framework-ai
• ChatGPT usage figures (company-reported): https://en.wikipedia.org/wiki/ChatGPT

Noman Akram

Noman Akram is the Founder and Editor-in-Chief of TWT News. He is a technology journalist with 5+ years of experience covering artificial intelligence, AI in healthcare, blockchain, cloud computing, and cybersecurity. He built TWT News to make complex emerging technologies understandable for professionals, students, and business leaders. Based in UK (United Kingdom), his reporting covers global tech developments with a focus on real world impact.