Models of the Mind: How physics, engineering, and mathematics have shaped our understanding of the brain

2021-06-16 · Source: Unsupervised Thinking · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Data Science & Analytics · Depth: Intermediate, extended

Summary

Grace Lindsay's book, "Models of the Mind: How Physics, Engineering, and Mathematics Have Shaped Our Understanding of the Brain," explores the historical and current influence of mathematical methods on neuroscience, from single-cell models to behavioral equations. The book, discussed in a cross-posted episode of the Brain Inspired podcast, aims to be an accessible history of computational neuroscience, highlighting interdisciplinary interactions. Lindsay, a postdoc at the Gadsby Computational Neuroscience Unit at University College London, emphasizes that mathematical models force explicit hypotheses and allow for rigorous testing, despite requiring simplification of biological systems. The discussion also touches on the birth of AI, the Minsky-Papert criticisms of perceptrons, the rediscovery of old mathematical concepts like graph theory in neuroscience, and the ongoing debate about rate versus temporal coding. Lindsay's personal research involves using artificial neural networks to understand vision and cognition, viewing them as a controlled environment to test neuroscience tools.

Key takeaway

For AI Scientists developing models of brain function, recognize that mathematical rigor demands explicit assumptions, which can reveal gaps in understanding. Your models should aim to clarify thinking and test hypotheses, not just describe phenomena. Be mindful that simplifying complex biological systems for mathematical representation can lead to overlooking crucial details, so balance abstraction with empirical grounding. Consider how your work contributes to a broader, interdisciplinary understanding, rather than seeking a single "grand unified theory."

Key insights

Mathematical models are crucial for explicit hypothesis testing and understanding the brain's complex, evolved systems.

Principles

• Mathematical models force explicit hypothesis formulation.
• Simplification is necessary but can lead to overlooked details.
• Interdisciplinary interaction drives scientific progress.

Method

Computational neuroscience uses mathematical models to link cognitive function to brain activity, serving as testable hypotheses for understanding brain function and guiding experimental design.

In practice

• Use artificial neural networks as model organisms for neuroscience research.
• Quantify neural activity to understand functional changes.
• Be wary of averaging data across individuals due to diverse underlying mechanisms.

Topics

• Computational Neuroscience
• Mathematical Brain Models
• Artificial Neural Networks
• Neuroscience History
• Brain Information Processing

Best for: AI Scientist, AI Researcher, Research Scientist, General Interest

Related on AIssential

• Your Brain Doesn't Command Your Body. It Predicts It. [Max Bennett] — The brain synthesizes diverse information into a coherent model, enabling perception by inference and mental simulation.
• OpenAI just SOLVED MATH.... — AI's breakthrough stemmed from its ability to connect disparate mathematical fields, revealing novel solutions.
• Clever Hans Has a Brain Scanner — Brain-encoding models' high prediction scores may reflect shortcut learning, not true brain understanding, due to noise-free synthetic data.
• They Copied a Fly's Brain Into a Computer. It Started Walking. Nobody Taught It How — Biological brain wiring can inherently produce complex behaviors without explicit training or optimization.
• Artificial Intelligence for Mathematical Reasoning: An Integrated Survey of Language Models, Neuro-symbolic Systems, and Verified Discovery — AI for mathematical reasoning is a rapidly evolving frontier, integrating diverse techniques for verified discovery.
• A Three-Layer Framework for AI in Scientific Discovery — True scientific discovery in AI hinges on qualitative reasoning for model formation, transcending mere search or execution.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Unsupervised Thinking.