Brain's Neural Network as a service
summary
Proposes treating a brain’s 86 billion neurons as a service by download-store-upload pipeline: sample snapshots over a window (minutes to ~30), record which neurons fire (each sends ~1,000 impulses/sec and makes tens of thousands of connections), store ensembles, run AI processing to extract experience-level ensembles, and then stimulate a target brain to replay them. Cites optogenetic experiments (virus that lights active neurons, optical fiber forcing a rat to circle) and neural-sensor advances (2016 biosensors) plus large models (NVIDIA Megatron ~530B). Flags accuracy limits, ethical risks, and likelihood of only achieving “closest to” reproduction.
AI-written summary · about
chapters (12)
- 0:00Introduction
- 0:25How does the Neural Network work
- 3:22Neural Network 3D Visualization
- 5:02Bungee Jumping
- 7:02Planning
- 9:33Proposal
- 14:26Adam Cohen Experiment
- 16:08Service Offering of People
- 19:56Service Offering of Experience
- 23:54Neural Network Development in Humans
- 26:40Challenges
- 31:44Questions
related talks
24mFrom Neural Networks to Digital Brains: The Next Leap in AI
30mCracking the code: neuronal networks in the brain
41mBrain-inspired computational primitives for reliable neuromorphic intelligence
45mMachine Learning, Deep Networks and Interpretability in Systems, Cognitive and...
01h 14mTissue vs silicone musings on the future of deep learning hardware and software
29mDeep Learning, Neuroscience and the future of AI
21mOrganoid Computing – Can Nature Save AI?
50mAI deciphers the brain code: from real synapse to a neuron's digital twin