EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC-ND 4.0

Before McCulloch and Pitts: Artificial neural networks in the 1930s

Bart de Boer

AI Magazine · 2026-12 · 47 (4)

DOI: 10.1002/aaai.70126 ↗

Kurz gefasst

Fachartikel in der Zeitschrift AI Magazine, online erschienen am 19.11.2026. Verfasst von Bart de Boer. Die Arbeit nennt 49 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

This paper presents the very earliest history of artificial neural network research. In textbooks about (deep) artificial neural networks, the 1943 paper by McCulloch and Pitts is generally taken to be the starting point of artificial neural network research. Here it is shown that their paper stems from a line of research into artificial neural models that started as early as 1928, and in which Nicolas Rashevsky was a central figure. These models were electric/electronic or mathematical in nature and already had many of the properties of modern neural networks (for example, activation functions and summation of activation) but also had some notable differences (for example, most of them worked in continuous time and did not make use of weight updates). This paper shows how this earlier work led to the ideas presented by McCulloch and Pitts in 1943 and argues that, contrary to what is generally stated in current artificial neural network literature, McCulloch and Pitts' paper did not spell the beginning of neural network research, but perhaps rather spelled the end of the 1930's wave of neural network research.

Zitieren

APA

Boer, B. D. (2026). Before McCulloch and Pitts: Artificial neural networks in the 1930s. AI Magazine, 47(4). https://doi.org/10.1002/aaai.70126

Vancouver

Boer BD. Before McCulloch and Pitts: Artificial neural networks in the 1930s. AI Magazine. 2026;47(4). doi:10.1002/aaai.70126

BibTeX
@article{Boer2026,
  author = {Boer, Bart de},
  title = {Before McCulloch and Pitts: Artificial neural networks in the 1930s},
  journal = {AI Magazine},
  year = {2026},
  volume = {47},
  number = {4},
  doi = {10.1002/aaai.70126}
}
Kostenlos starten →