Schmidhuber Reiterates: Roots of Current AI Boom Trace Back to 1991 Munich
SchmidhuberAI · x · 2026-08-06
In a recent blog update, Jürgen Schmidhuber reiterates that the technological foundations of today's trillion-dollar AI boom trace back to his team's 1991 research in Munich. He lists core concepts including early linear Transformers, Pre-Training, Neural Net Distillation, Deep Residual Learning, and GANs, asserting these form the basis of modern LLMs like ChatGPT. He also claims the heavily promoted JEPA architecture is essentially identical to his 1992 Predictability Maximization system.
More from Research
- AgentGUI Preprint: Humans Understand Agent Traces 38% Faster — Michael_D_Moor · 2026-08-07
- SIGNPOST-Bench: Evaluating Text-Vision Conflict Resolution in MLLMs — Sirun Li · 2026-08-07
- PolyLayout: Multi-Room 3D Layout Estimation with Joint Optimization Improves Accuracy — anselm · 2026-08-07
- Open-Sourcing a Multi-Agent Adversarial AI Pipeline for Medical Data Analysis — JWojoMojo · 2026-08-07
- Meta's New Architecture Validates Scaling Laws in Ads, Lifting IG Conversions by 6% — Meta_Engineers · 2026-08-07
- Qwen-3D: Enhancing Spatial Reasoning via Multi-View Geometric Cues — udmrzn · 2026-08-07