AINeural networks’ origins and the “applied statistics” label for AI
One post argues that “applied statistics” fits much, but not all, of modern AI—and that the earliest neural network predates knowledge of biological neurons.
TLDR
A post challenges the description of modern AI as “bio-inspired statistics.” It traces what it calls the first neural network—a two-layer linear network—to work published by Gauss and Legendre in 1795–1805. To support its argument, it says the first microscopic image of a nerve cell was created by Valentin in 1836, while Waldeyer coined the term “neuron” in 1891. The post accepts “applied statistics” as a label for much of modern AI, but says it does not cover everything.
Combined views
1.5K
1 Source, first seen 2h ago
Neural networks’ origins and the “applied statistics” label for AI
One post argues that “applied statistics” fits much, but not all, of modern AI—and that the earliest neural network predates knowledge of biological neurons.