AI timeline
About this app
Explore seventy years of AI history on an interactive 3D timeline, from Turing and the AI winters to ChatGPT, with selectable milestones and event details.
Waves of hype and disappointment
The history of artificial intelligence has moved in waves. Early theories and the first learning models were followed by inflated expectations, cut funding and long periods of stagnation known as AI winters. Rule-based expert systems and individual successes such as chess computers kept the field visible. With larger datasets, graphics processors and better training methods, deep learning then took hold. The Transformer architecture became the basis of today’s language models.
What you see
A 3D timeline with three colour-coded eras: the pioneer years, AI winters and expert systems, and the learning revolution. Along it sit milestones including the Turing test (1950), the Dartmouth conference (1956), the perceptron (1958), ELIZA (1966), backpropagation (1986), Deep Blue (1997), AlexNet (2012), AlphaGo (2016), the Transformer (2017), GPT-3 (2020) and ChatGPT (2022). Each milestone comes with a short explanation.
Things to try
Click a node or use “Next” and “Back” to walk through the milestones in order. Notice how far apart the events are in the middle era and how closely they follow one another in the last one. A useful comparison: what separates the 1958 perceptron from the networks that recognise images from 2012 onwards?
Frequently asked questions
- What was the Turing test?
- In 1950 Alan Turing proposed an imitation game: a person converses in writing with a human and a machine and has to work out which is which. If they cannot do so reliably, the machine can be credited with intelligent behaviour by that criterion. The test remains a reference point but is widely debated.
- What is an AI winter?
- An AI winter is a period in which expectations of AI were disappointed and funding and interest dropped sharply. A first such slump is usually dated to the mid-1970s. These periods typically followed times of big promises.
- Why was the 2017 Transformer so important?
- The Transformer, introduced in the paper “Attention Is All You Need”, processes text with self-attention instead of recurrent networks. This makes it easy to train in parallel on large amounts of data. It became the basis of models such as BERT and GPT.
- What can I do in this app?
- You move through a 3D timeline, select milestones by clicking or with “Next” and “Back”, and read a short explanation for each. The three eras are colour-coded.
Subject: Machine Learning
More from Machine Learning
- AdaBoost
- Anatomy of a prompt
- Data privacy traffic lights
- Decision Tree
- Ensemble: Bagging vs. Boosting
- Finding faces: masks versus learning from examples
- k-Means
- k-Nearest Neighbors (kNN)
A VisuApp by heyprof: interactive, free in your browser, no sign-up. What is a VisuApp? · All apps