Timeline
How the idea of superintelligence took shape.
Eleven dated moments, from the first paper to ask whether machines can think to the year AI reached a mass audience. Later entries will be added once their importance is clear.
Turing asks whether machines can think
Alan Turing publishes “Computing Machinery and Intelligence” and proposes the imitation game, later called the Turing test.
The field gets its name
A summer workshop at Dartmouth College brings researchers together under a new label: artificial intelligence.
The intelligence explosion
I. J. Good describes an ultraintelligent machine that could design better machines, and the runaway improvement that would follow.
The singularity
Vernor Vinge's essay “The Coming Technological Singularity” argues that superhuman intelligence would end the period in which people can forecast the future.
A machine beats the chess champion
IBM's Deep Blue defeats world champion Garry Kasparov in a six-game match. The system is superhuman at chess and nothing else.
Deep learning breaks through
AlexNet, a deep neural network trained on graphics processors, wins the ImageNet image recognition contest by a wide margin.
The vocabulary is fixed
Nick Bostrom publishes “Superintelligence: Paths, Dangers, Strategies”, which defines the term and its three forms.
Go falls
DeepMind's AlphaGo defeats Lee Sedol, one of the strongest Go players in the world, four games to one.
The Transformer
The paper “Attention Is All You Need” introduces the architecture behind today's large language models.
A scientific problem yields
AlphaFold 2 predicts protein structures at the CASP14 assessment with accuracy close to experimental methods.
AI reaches a mass audience
ChatGPT is released on 30 November and puts a large language model in front of the general public.