Schmidhuber: Gauss and Legendre invented 2-layer neural nets back in 1795-1805
SchmidhuberAI · x · 2026-09-20
- Jürgen Schmidhuber pushes back on framing modern AI as mere "applied statistics" inspired by biology, citing his own historical research.
- He argues the first "modern" two-layer linear neural nets were created by Gauss (1795, unpublished) and Legendre (1805) — over two centuries ago and decades before the first microscopic image of a nerve cell (Valentin, 1836) or the coining of "neuron" (Waldeyer, 1891).
- Since biological neurons were unknown at the time, the original nets weren't bio-inspired; and Gauss-era compute was trillions of times more expensive than in 2025.
More from AGI Musings
- Jack Clark: 'stochastic parrot' was a cognitive virus that blinded elites for years — teortaxesTex · 2026-09-20
- Frontier model outputs now need another model to read them, insider says — DavidSKrueger · 2026-09-20
- A single web search makes LLM inputs OOD, argues developer — menhguin · 2026-09-20
- Two types of daily AI users: becoming more slop or more human — StewartalsopIII · 2026-09-20
- Spotify reframes recommender systems for the agentic web in weekly IR papers roundup — _reachsumit · 2026-09-20
- AI moral status debate: does conceding machine suffering imply utilitarian math? — teortaxesTex · 2026-09-20