Beginner’s Guide

This is a beginner’s map of the history of AI, from Turing’s question in 1936 to the generative-AI industry contest of 2025, 63 figures threading the near-ninety-year technical line from theoretical foundations, the perceptron winter, the deep-learning revival, AlphaGo and the Transformer, to ChatGPT igniting worldwide attention. The goal is not to memorize "who invented what" but to see clearly how wave after wave of technology, declared "about to arrive" and then "sunk into winter," actually rose and fell — this industry contest, still ongoing, is also the only history among all this site’s subjects whose "story is not yet finished."

Know These Few Things First

The history of AI has seen more than one "winter" — the first winter caused by the perceptron dispute, the trough of over a decade before the deep-learning revival — and each judgment that "this path leads nowhere" was thereafter proven premature, a pattern itself worth noting.
A "dispute over invention priority" is likewise common in this history — the backpropagation algorithm (Rumelhart’s team vs. Werbos’s early work), the attribution of deep-learning contributions (Hinton vs. Schmidhuber) — and this site states the historical course both sides acknowledge, without adjudicating for either who the "sole inventor" is; what you make of it is your own judgment.
Many relations in this history are of the same set of people wearing several hats at once, at once collaborators and competitors — Altman and Musk co-founded OpenAI then split openly, Amodei left OpenAI to found Anthropic, Sutskever voted to remove Altman and himself resigned a few months later — and the "nemesis" marks in the relationship graph often describe just this kind of "former comrade-in-arms."
Algorithmic bias and AI safety are not vague abstractions — Buolamwini’s "Gender Shades" research and Gebru’s "stochastic parrot" paper are both real academic work backed by concrete research methods and data, and this site tries in the relevant entries to restore the specific content of these studies rather than stop at the level of slogans.

A Few Suggestions

Look at the timeline first, to grasp "what other corners of the industry were busy with when this breakthrough appeared" — the 2012 in which AlexNet shocked academia was in fact a full decade before ChatGPT’s release, and many of the key groundworks of that decade (Transformer, GAN, AlphaGo) are often compressed into the over-simple narrative of "AI suddenly exploded."
The "master/disciple" marks in the graph mark inheritance in technical line or talent cultivation, not real face-to-face apprenticeship — between Timnit Gebru and Joy Buolamwini it is "ally," not master-disciple, so please take the specific text on the card page as authoritative.
Note the content in each card’s "version disputes" section, especially the statements around recent events such as the OpenAI board crisis and Musk’s split with the company — these disputes are quite recent, the various accounts still changing, and it is advisable to keep the mindset of "learning the publicly known course of the facts" rather than "rushing to take a side."

How to Read

Want to know "how AI went from an obscure discipline to a global headline"

Look at the timeline — from the handful at the 1956 Dartmouth conference to ChatGPT’s hundred million users within two months in 2022, over most of this near-seventy-year curve AI was an obscure research direction not favored by the mainstream, and truly winning wide public attention is a matter of only the last few years, and this contrast of "long on the cold bench, suddenly pushed into the spotlight" is itself the part of this history most worth taking in.

Want to know "in this industry race, who is contending with whom"

Look at the "nemesis" links in the graph — Altman and Musk, Altman and Sutskever, Huang and Liang Wenfeng, LeCun and Marcus — and behind these links is often real commercial competition, path disagreement, or public debate, not fictional dramatic conflict, and the specific course can be checked by tapping into the corresponding figure’s card.

Want to know "in this history, whose voice is easily overlooked"

Focus on the "who is this" and "modern echoes" parts of the cards for Fei-Fei Li, Gebru, Buolamwini, Rosenblatt, and Werbos — some whose contribution to data infrastructure was long obscured by the halo of algorithms, some whose early research was only acknowledged years later, and these "belated recognitions" are themselves a part of the history of AI worth knowing.

Go Deeper: General AI Reading

Stuart Russell & Peter Norvig, Artificial Intelligence: A Modern ApproachThe most widely adopted introductory AI textbook in universities worldwide, systematically threading the full technical line of the discipline from symbolism to deep learning.

Nick Bostrom, SuperintelligenceA systematic exploration of the existential risk superintelligent AI may bring, which deeply shaped the risk perception of several tech-industry leaders including Musk.

Cade Metz, Genius MakersA journalist’s account threading the rise of institutions such as DeepMind and OpenAI and the contest of key figures, an important reference for understanding this recent industry history.

Go Deeper: Biographies and Special Topics

Walter Isaacson, Elon MuskAn account of the full inner journey of Musk founding OpenAI, leaving the board, and thereafter setting up xAI, an important first-hand source for understanding this founders’ falling-out.

Fei-Fei Li, The Worlds I SeeFei-Fei Li’s own memoir, an account of her path from immigrant youth to creator of ImageNet and an important founding figure of the AI field.

Karen Hao, Empire of AIA journalistic work of in-depth investigation into OpenAI’s internal governance and the 2023 board crisis, offering a multi-sided view of the crisis.

A Time Coordinate

Birth & Theoretical Foundations (1936–1969)

1950 Turing publishes "Computing Machinery and Intelligence"
1956 The Dartmouth conference formally names "artificial intelligence"
1958 Rosenblatt proposes the perceptron
1959 Samuel proposes the term "machine learning"
1969 Minsky’s Perceptrons provokes the first AI winter

The Connectionist Revival & Statistical Learning (1974–1997)

1974 Werbos’s doctoral thesis proposes the principle of backpropagation
1986 Rumelhart & Hinton’s paper drives the wide application of backpropagation
1988 Pearl publishes Probabilistic Reasoning in Intelligent Systems
1995 Vapnik systematically expounds statistical learning theory
1997 Deep Blue beats Kasparov, LSTM proposed the same year

The Deep-Learning Revival & the Shock of ImageNet (2006–2016)

2006 Hinton publishes the deep-belief-network paper
2009 Fei-Fei Li releases the ImageNet dataset
2010 DeepMind founded
2012 AlexNet wins the ImageNet competition to a shock
2014 Goodfellow proposes the generative adversarial network
2016 AlphaGo beats Lee Sedol

The Race to Scale & the Industry Contest (2015–2025)

2015 OpenAI founded
2017 The Google team publishes "Attention Is All You Need"
2018 BERT released, "Gender Shades" research draws attention
2021 Anthropic founded
2022 ChatGPT released, past a hundred million users in two months
2023 The OpenAI board crisis
2025 DeepSeek released, shocking the compute narrative