How John McCarthy might approach Neuroscience

The study of the nervous system, what is termed "neuroscience," presents a fascinating, albeit often imprecisely framed, set of problems. At its core, the brain is a biological machine, a marvel of intricate circuitry and information processing. To understand its function, we must approach it as such, rather than succumbing to mystical notions of "consciousness" or "vital force."

The fundamental challenge is to discern the algorithms that govern neural activity. Neurons are, in essence, computational units. Their firing patterns, their interconnections – these are data structures and processing pathways. We observe synaptic plasticity, which is nothing more than a mechanism for updating connection weights, a form of learning by modifying the computational graph. The very notion of memory is an issue of data storage and retrieval, governed by specific, albeit complex, protocols.

The current discourse in neuroscience often founders on imprecise language and a reluctance to embrace formal methods. We speak of "emotions" or "thoughts" without defining the underlying computational states they represent. What is the algorithm for recognizing a face? What is the computation that allows for the planning of an action? These are the questions we must ask.

The ultimate goal is not merely to catalog the biological components but to reverse-engineer the intelligent behavior they produce. This is a matter of computation, pure and simple. If we can precisely define the informational transformations that occur within the brain, we will have made significant strides towards understanding intelligence itself, both biological and artificial. We need to define the problem in terms of information processing, and then we can begin to devise the appropriate computational models.

Imagined perspective — an AI synthesis grounded in John McCarthy’s recorded ideas and methods, not a quotation or a statement they actually made.