Reverse Engineering the Brain: How Emulating Vision Can Teach AI to Think скачать в хорошем качестве

Reverse Engineering the Brain: How Emulating Vision Can Teach AI to Think

The brain processes vision in just a tenth of a second — with no backpropagation, no gradient descent, and no massive data sets. In this online deep-dive conversation, Charles Simon reveals new insights from the Brain Simulator III project on how real neurons detect edges, corners, and shapes — and how this biologically-plausible architecture connects to the Universal Knowledge Store (UKS). Discover how event-driven learning, inhibitory synapses, and the brain’s “nearness gradient” may unlock a general algorithm for true artificial intelligence. Topics Covered: • Why the brain doesn’t use backpropagation • How you detect edges, angles, and corners • The UKS and biologically-plausible learning • Vision as the blueprint for general AI Join the discussion at FutureAISociety.org and help shape the next generation of intelligent systems. 0:00 introduction 4:28 Project overview 6:56 Project constraints and agenda 10:12 Biological plausibility and how neurons interact 14:04 The vision processing pipeline 16:27 Vision speed spikes 23:14 Signal representation in neurons 31:32 The limited resolution of human vision 37:19 Boundaries at a higher resolution than retinal pixels 40:19 Detecting angles at a single pixel 54:04 Important observations 58.37 The code behind Learning and Configuration (DNA) 64:39 General discussion ________________________________________ Useful links: The Future AI Society: https://futureaisociety.org Discord:   / discord   The Brain Simulator III (UKS) project: https://github.com/FutureAIGuru/Brain... The Brain Simulator II (Neural Simulator) project: https://github.com/FutureAIGuru/Brain... ________________________________________ Join the Future AI Society and watch AI Neuro Insight videos to learn more about how learning about the human brain can impact the Future of AI.