1950 — The question

Alan Turing publishes “Computing Machinery and Intelligence”, asking whether machines can think and proposing the imitation game, later called the Turing test.

1956 — The name

A summer workshop at Dartmouth College, organized by John McCarthy, Marvin Minsky, Nathaniel Rochester and Claude Shannon, gives the field its name: artificial intelligence.

1960s — Optimism

Early programs prove theorems, play checkers and hold simple conversations (such as ELIZA, 1966). Researchers predict human-level AI within a generation.

1970s and late 1980s — The AI winters

Progress proves far harder than expected. Funding is cut in the mid-1970s and again in the late 1980s after the commercial boom in “expert systems” fades. These periods become known as AI winters.

1997 — Deep Blue

IBM’s Deep Blue defeats world chess champion Garry Kasparov in a six-game match.

2012 — Deep learning takes off

A neural network called AlexNet wins the ImageNet image recognition challenge by a large margin, trained on GPUs. Deep learning quickly becomes the dominant approach.

2016 — AlphaGo

DeepMind’s AlphaGo defeats Go champion Lee Sedol, a milestone many expected to be a decade away.

2017 — Transformers

Google researchers publish Attention Is All You Need, introducing the transformer architecture behind today’s language models.

2022 — Generative AI goes mainstream

Image generators reach the public, and on November 30, 2022, OpenAI releases ChatGPT, which becomes one of the fastest-adopted consumer applications in history.

Since then

Rapid progress in multimodal models, reasoning models, open-weight models and AI agents — alongside new laws such as the EU AI Act and intense debate about safety, jobs and copyright.

Lessons from 75 years of AI

Approaches come and go. Symbolic AI dominated early decades; neural networks were dismissed, then revived to dominate today. Ideas that seem dead can return when data and computing catch up.

Predictions are unreliable. Early pioneers expected human-level AI within a generation; later, many experts underestimated how fast deep learning would progress.

Hardware matters. GPUs, originally built for games, made deep learning practical. Progress often follows advances in computing.

Data matters. Large datasets like ImageNet and web-scale text enabled breakthroughs as much as new algorithms.

Hype cycles have consequences. Overpromising contributed to funding collapses known as AI winters. Today’s investment boom raises the question of whether expectations are again running ahead of reality.

Key people along the way

  • Alan Turing (1950) asked whether machines can think and proposed the imitation game, now called the Turing test.
  • John McCarthy, Marvin Minsky, Claude Shannon and Nathaniel Rochester organized the 1956 Dartmouth workshop that named the field.
  • Frank Rosenblatt built the perceptron in the late 1950s, an early neural network.
  • Edward Feigenbaum championed expert systems, the commercial AI of the 1980s.
  • Geoffrey Hinton, Yann LeCun and Yoshua Bengio kept neural networks alive through lean years and shared the 2018 Turing Award.
  • Fei-Fei Li led ImageNet, the dataset behind the 2012 breakthrough.
  • Demis Hassabis and the DeepMind team built AlphaGo and AlphaFold.
  • Ashish Vaswani and co-authors introduced the Transformer in 2017.

Behind each name are many collaborators, students and engineers — AI progress has always been a collective effort.

Further reading

For deeper accounts, see our list of classic books to understand AI and our ten milestones of the deep learning revolution, which cover the period since 2012 in more detail.

What to watch next

  • Whether current investment levels are sustained or followed by a correction.
  • New paradigms beyond scaling, such as reasoning and world models.
  • AI’s role in scientific discovery.
  • Historians’ reassessments of AI’s recent breakthroughs.

Key takeaways

  • Turing’s 1950 paper and the 1956 Dartmouth workshop launched AI as a field.
  • Optimism in the 1960s gave way to AI winters in the 1970s and late 1980s.
  • Deep learning took off in 2012; Transformers arrived in 2017.
  • ChatGPT brought generative AI to the mainstream in 2022.
  • Hardware, data and realistic expectations shape AI progress.

Frequently asked questions

Who invented artificial intelligence?

No single person. Alan Turing posed key questions in 1950, and John McCarthy coined the term for the 1956 Dartmouth workshop, with many contributors since.

What were the AI winters?

Periods of reduced funding and interest, after early promises failed to materialize, notably in the 1970s and late 1980s.

Why did AI suddenly improve after 2012?

Large datasets, GPU computing and improved neural network techniques came together, enabling deep learning breakthroughs.