The TWIML AI Podcast (formerly This Week in Machine Learning & Artificial Intelligence) Spiking Neural Networks: A Primer with Terrence Sejnowski - #317
Nov 14, 2019
Terrence Sejnowski, a pioneer in computational neuroscience and head of the Computational Neurobiology Lab at the Salk Institute, joins to unravel the complexities of spiking neural networks. He discusses how these networks mimic biological brain functions, boosting energy efficiency in machine learning. The conversation also delves into the challenges of training these networks, the synergy between neuroscience and AI, and their transformative potential in robotics. Sejnowski shares insights on the future of neuromorphic hardware and its implications for technology.
AI Snips
Chapters
Transcript
Episode notes
Brain vs. Deep Learning Efficiency
- Biological brains operate remarkably efficiently on just 20 watts.
- Deep learning systems, in contrast, consume vast amounts of energy, a growing concern.
Spiking as a Signaling Method
- Spikes, or action potentials, are the brain's primary signaling method, lasting about a millisecond.
- These signals travel relatively slowly compared to computer chips but encode information effectively.
Complexity of Spikes and Synapses
- Neural spikes, while digital in their all-or-none nature, exhibit analog timing and varied sizes and shapes.
- Synapses introduce further complexity with dynamic systems influencing signal strength and modulation.