The history of artificial intelligence

From the first mathematical neuron model in 1943 to autonomous AI agents in 2026 — a complete timeline of more than 80 years of artificial intelligence.

1940–1955 — The Foundations

1943The first artificial neuron

Warren McCulloch and Walter Pitts publish the first mathematical model of an artificial neuron — the foundation for all neural networks.

1950The Turing Test

Alan Turing introduces in 'Computing Machinery and Intelligence' the question of whether machines can think, and the test that bears his name.

1951SNARC — the first neural network in hardware

Marvin Minsky and Dean Edmonds build SNARC, the first artificial neural network as a physical device, constructed from electrical components.

1956–1974 — The Birth of AI

1956The Dartmouth Conference

John McCarthy, Marvin Minsky, Claude Shannon, and others introduce the term 'artificial intelligence' at the Dartmouth Conference — the official birth of AI as a field.

1957The Perceptron

Frank Rosenblatt develops the Perceptron: the first learning artificial neural network that recognizes patterns based on examples.

1965ELIZA — the first chatbot

Joseph Weizenbaum develops ELIZA at MIT: a chatbot that simulates a psychotherapist and is so convincing that many users believe they are talking to a human.

1974–1993 — AI Winters

1974–1993The AI winters

Two periods of cutbacks and disappointment: the first AI winter (1974–1980) and the second (1987–1993), following the failure of expert systems and inflated expectations.

1993–2010 — Machine Learning

1997Deep Blue defeats Kasparov

IBM's chess computer Deep Blue defeats world champion Garry Kasparov 3.5–2.5 — the first time a computer defeats the best human chess player in the world.

2006The deep learning breakthrough

Geoffrey Hinton publishes a method to effectively train deep neural networks via pre-training, reopening research into neural networks.

2009ImageNet

Fei-Fei Li launches ImageNet: a dataset of 14 million labeled images that becomes the benchmark for image recognition and drives the deep learning revolution.

2010–2017 — Deep Learning Revolution

2011Watson wins Jeopardy! — Siri launched

IBM Watson defeats the best Jeopardy! champions; Apple launches Siri as the first mainstream voice assistant on the iPhone 4S.

2012AlexNet — the deep learning revolution

AlexNet by Krizhevsky, Sutskever, and Hinton wins the ImageNet competition by an unprecedented margin and ushers in the deep learning revolution.

2016AlphaGo defeats Lee Sedol

DeepMind's AlphaGo defeats world Go champion Lee Sedol 4–1, twenty years earlier than experts expected.

2017Attention Is All You Need — the Transformer

Google Brain publishes the Transformer architecture, which replaces all recurrent networks and forms the foundation for GPT, BERT, Claude, Gemini, and all modern LLMs.

2018–2022 — Large Language Models

2018BERT

Google launches BERT, a bidirectional Transformer language model that becomes the standard for search engines and language understanding tasks.

2020GPT-3

OpenAI launches GPT-3 with 175 billion parameters: the first language model that convincingly generates human-like text, code, and translations, making the world realize what LLMs can do.

2021AlphaFold 2 and GitHub Copilot

DeepMind solves the 50-year-old protein folding problem with AlphaFold 2; GitHub Copilot brings AI code completion to software developers.

2022Stable Diffusion and ChatGPT

Stable Diffusion democratizes image generation as an open source tool; ChatGPT reaches 100 million users in two months in November — the fastest growth of a consumer application ever.

2023 — The Year of the AI Explosion

March 2023GPT-4, Claude v1, and Gemini

OpenAI launches the multimodal GPT-4; Anthropic releases Claude v1; Google launches Bard (later Gemini). The AI race at the frontier level has begun.

2023Llama, Mistral, and open source AI

Meta launches Llama and later Llama 2 as open source LLMs; Mistral AI from Paris releases Mistral 7B that outperforms closed models — open source AI comes of age.

2024 — Multimodality and Agents

2024GPT-4o, Claude 3, and Gemini 1.5

OpenAI launches GPT-4o with real-time text/image/audio; Anthropic releases Claude 3 Haiku/Sonnet/Opus; Google's Gemini 1.5 Pro processes 1 million tokens of context.

2024AI agents go mainstream

AI agents — LLMs that independently use tools, make plans, and execute multiple steps — go mainstream. Claude Computer Use shows AI operating a computer.

2025 — Reasoning Models and Open Source Breakthrough

January 2025DeepSeek-R1 — the open source shock

Chinese DeepSeek launches R1: an open source reasoning model that matches GPT-4o at a fraction of the cost, shattering the myth that frontier AI is exclusively American.

February 2025Claude 3.7 Sonnet and OpenAI o3

Anthropic introduces Claude 3.7 Sonnet with 'extended thinking'; OpenAI's o3 achieves human level on mathematical and scientific benchmarks.

2025Gemini 2.5, Llama 4, Grok 3, and MCP

Google launches Gemini 2.5 Pro; Meta releases Llama 4; xAI introduces Grok 3. Anthropic's Model Context Protocol (MCP) becomes the industry standard for AI tool integrations.

2025EU AI Act takes effect

The EU AI Act enters into force in phases: the first binding international regulation for AI, with prohibitions on high-risk applications and transparency requirements for LLMs.

2025AI coding tools become standard

AI coding tools such as Cursor, Windsurf, and GitHub Copilot become standard in software development; it is estimated that more than half of all professional code is now co-generated by AI.

2026 — Autonomous AI Agents

2026GPT-5

OpenAI launches GPT-5: the first model that consistently reaches human level across a broad spectrum of cognitive tasks, with strong reasoning and multimodal capabilities.

2026Claude Opus 4.8 and Sonnet 4.6

Anthropic launches Claude Opus 4.8 and Sonnet 4.6 — the most capable models in the Claude family, with particular strengths in reasoning, safety, and long-context tasks.

2026Autonomous AI agents everywhere

Autonomous AI agents independently execute multi-step tasks — writing code, building websites, conducting research — and become standard integrated into business processes worldwide.


Ster Software

The most complete knowledge platform on artificial intelligence.

Kraaienjagersweg 24
7341 PT Beemte Broekland, Netherlands


© 2026 Ster Software BV · Chamber of Commerce 75474913

Content generated by Claude (Anthropic) · model: claude-sonnet-4-6