Artificial neural networks: Discovery
The story
How the idea was found and came to be accepted
In 1943 Warren McCulloch and Walter Pitts published a paper arguing that, because nerve cells fire all or nothing, networks of them can be described with propositional logic [1, 2]. In 1949 Donald Hebb proposed that a connection grows stronger when the two neurons it joins are repeatedly active together [7]. In 1958 Frank Rosenblatt described the perceptron in Psychological Review, and the US Office of Naval Research unveiled a machine that learned after 50 trials to tell cards marked on the left from those marked on the right [3, 4]. In 1969 Minsky and Papert published 'Perceptrons', and the limits of a single layer, shown by the XOR problem, are blamed for the AI winter of the 1970s [4, 5]. Hopfield published a network model of associative memory in 1982, and in 1986 Rumelhart, Hinton and Williams showed that backpropagation could train networks with a hidden layer [7, 8]. Hopfield and Hinton shared the 2024 Nobel Prize in Physics [6].
as of 2026-11-12
People
The people behind it, their years and what each contributed
| Name | Years | Contribution |
|---|---|---|
| Warren McCulloch | 1898 to 1969 | With Walter Pitts, gave neural networks their logical form in 1943 [1, 2] |
| Frank Rosenblatt | 1928 to 1971 | Invented the perceptron, a network that learns by adjusting its connections [3, 4] |
| Geoffrey Hinton | born 1947 | Used Hopfield's network as the basis of the Boltzmann machine; shared the 2024 Nobel Prize in Physics [6] |
as of 2026-11-12
Milestones
The key moments, year by year
| Year | Event |
|---|---|
| 1943 | McCulloch and Pitts publish 'A logical calculus of the ideas immanent in nervous activity' [2] |
| 1949 | Hebb proposes that when two neurons are repeatedly active together, the connection between them grows stronger [7] |
| 1958 | The US Office of Naval Research unveils Rosenblatt's perceptron, a machine that taught itself from examples [4] |
| 1969 | Minsky and Papert publish 'Perceptrons', showing the limits of single-layer networks [4, 7] |
| 1982 | Hopfield publishes a network model of associative memory [7] |
| 1986 | Rumelhart, Hinton and Williams show that backpropagation can train networks with a hidden layer [7, 8] |
| 2024 | Hopfield and Hinton receive the Nobel Prize in Physics [6] |
as of 2026-11-12
Sources
- A logical calculus of the ideas immanent in nervous activity The Bulletin of Mathematical Biophysics (W. S. McCulloch and W. Pitts), Springer, 1943, vol. 5, December 1943
- McCulloch, Warren Sturgis | Encyclopedia.com Complete Dictionary of Scientific Biography (entry by Kenneth Aizawa), on Encyclopedia.com
- The Perceptron: A Probabilistic Model for Information Storage and Organization in the Brain Psychological Review (F. Rosenblatt), 1958
- Professor’s perceptron paved the way for AI - 60 years too soon Cornell Chronicle, Cornell University (Melanie Lefkowitz), 2019
- Cornell Virtual Workshop > Scientific Machine Learning (SciML) > Multi-Layer Perceptron > The Need for Depth - The XOR Problem: A Historical Turning Point Cornell University Center for Advanced Computing
- Press release: The Nobel Prize in Physics 2024 The Royal Swedish Academy of Sciences, NobelPrize.org, 2024
- Scientific Background to the Nobel Prize in Physics 2024 Royal Swedish Academy of Sciences, 2024
- Learning representations by back-propagating errors Nature (D. E. Rumelhart, G. E. Hinton and R. J. Williams), via the SAO/NASA Astrophysics Data System, 1986