@knutson_brain@sfba.social @neuralreckoning@neuromatch.social @tyrell_turing@fediscience.org @awaisaftab@mastodon.social @summerfieldlab@qoto.org @Neurograce@mastodon.online @eikofried@mastodon.online
I appreciate the collegial pushback
@tyrell_turing@fediscience.org and your excellent piece on the topic (here
https://www.frontiersin.org/articles/10.3389/fcomp.2022.810358/full) To summarize your point: whether the computer is a metaphor for the brain or not depends on how you define "computer".
Definition 1 (it's not a metaphor):
"based on the usage of the word “computer” in computer science, a computer is merely some physical machinery that can in theory compute any computable function. According to this definition the brain is literally a computer; there is no metaphor."
Definition 2 (it is a metaphor):
computers and computation as being necessarily sequential, or discrete, or restricted to passive processing of a stream of inputs using a step-by-step program ... we can say not only that brains are not computers, we can also say that computers are poor metaphors for brains, since the manner in which they operate is radically different from how brains operate.
(Van Gelder, 1998; Cisek, 1999; Brette, 2018, 2019; Cobb, 2020).
I was indeed using the second definition of computer that your article employs; the one that I see as most pervasive in brain research today (including my own work). So, as you say, it IS a metaphor 😉. But your point is appreciated and I will do better to clarify going forward.
Why did I use definition 2? As we both know, definition 2 has been tremendously successful in driving brain research toward some incredible discoveries (in eg vision and memory) and interactions between that arm of brain research and AI/ML have bene tremendously impactful.
I am thus concerned that those incredible successes will lead researchers to presume that this is a good approach to understanding everything the brain does. However, I strongly believe that deviations in thinking about the brain away from definition 2 are crucial for impactful progress in thinking about some of the brain functions related to dysfunction, such as affect/emotion.
For instance:
When thinking about emotions (like mood and anxiety), emphasizing the brain as an information processor that computes fails to capture the notion that we don't just compute/predict what's bad and good in the world, we are drawn toward good and repelled from bad with these very strong experiences we call feelings like craving, disgust, joy and fear (and those don't seem to be just tacked on at the end after goodness/badness computation; it's all wrapped up together).
Likewise, thinking about the brain as the type of thing that processes input to generate behaviors fails to capture that the input we receive depends on what we do and thus input and output operate in a continuous feedback loop (ala control systems/active sensing; approach good/hide from bad).
When some researchers deviate from definition 2, it leads them down intriguing paths. To quote Paul Cisek:" You realize that neither the term ‘decision-making’ nor the term ‘attention’ actually corresponds to a thing in the brain. Instead, there are certain very pragmatic circuits in the brain, and they do certain things like ‘approach’ or ‘avoid.’ … Some of those things are going to look a bit like attention."
These are the paths I very much believe we need to traverse for this arm of brain research to be impactful for brain disorders.
