In trying not to be vague, I veered towards writing too much. So if someone could take what I wrote and destroy half of it (or destroy all of it, and write a new and better thing) that would be lovely.
(In a slight inversion of PCT, I feel very sick of looking at my own writing.)
I share the impression that there's been a bit more talk and thought on it.
But I tried out the obvious search terms on here, and for the life of me I can't find it. (Other than a few side-mentions, which it didn't seem worth tagging.)
Free Energy Principle gets mentioned a few more times, but I don't know that it's quite the same thing.
Perceptual control theory (PCT) is a psychological theory of animal and human behavior. PCT postulates that an organism's behavior is a means of controlling its perceptions. The model is based on the principles of negative feedback [1]. It is to some extent an application of the ideas used in the engineering discipline of control theory to the modeling of the human mind and behavior.
PCT postulates that layers of control systems, which have access to a metric to optimize and some set of policies or actions, can maintain balancing-acts for difficult, high-abstraction things without developing any explicit model for how those actions relate to the metric being tracked. The brain is postulated to be one of these multi-layered PCTs.
Physical movements are a favorite case-study, withsince they're relatively easy to break down into this these sorts of layered control theory sub-problems. An important subtlety is that the theory postulating that a change incontrol systems are optimizing for the perception of where your hand should be sets off a series of sub-actionsstate, rather than for a concrete environmental state itself.
Actions and behaviors develop because they do something to minimize that "error."reduce the mismatch between internal perception, and the stimulus readings received.
It's unclear whether PCT is a valid theory. It doesn't significantly constrain the space of possible minds that could be built from it, and the advocates of the theory on the blog were unable to make a clear case for it. Experimental results for its quality as an algorithm seemed lackluster; see these critical comments about the paper version of this technical report, which claim that the correct results may have been achieved more through parameter-fitting than PCT.
Some anecdotally found it more useful for explaining bugs in human behaviorbehavior, than functionalfor modeling what would be ideal behavior.
Perceptual control theory (PCT) is a psychological theory of animal and human behavior. PCT postulates that an organism's behavior is a means of controlling its perceptions. The model is based on the principles of negative feedback [1]. It is to some extent an application of the ideas used in the engineering discipline of control theory to the modeling of the human mind and behavior.
PCT postulates that layers of control systems, which have access to a metric to optimize and some policies or actions, can maintain balancing-acts for difficult, high-abstraction things without developing any explicit model for how those actions relate to the metric being tracked. The brain is postulated to be one of these multi-layered PCTs.
Physical movements are a favorite case-study, with the theory postulating that a change in the perception of where your hand should be sets off a series of sub-actions to minimize that "error."
Perceptual control theory (PCT) is a psychological theory of animal and human behavior. PCT postulates that an organism's behavior is a means of controlling its perceptions. The model is based on the principles of negative feedback [1]. It is to some extent an application of the ideas used in the engineering discipline of control theory to the modeling of the human mind and behavior.
PCT postulates that layers of control systems, which have access to some policies or actions, can maintain balancing-acts for difficult, high-abstraction things without developing any explicit model for how those actions relate to the metric being tracked. The brain is postulated to be one of these multi-layered PCTs.
Physical movements are a favorite case-study, with the theory postulating that a change in the perception of where your hand should be sets off a series of sub-actions to minimize that "error."
It's unclear whether PCT is a valid theorytheory. It doesn't significantly constrain the space of possible minds that explains anything. Thecould be built from it, and the advocates of the theory on the blog were unable to make a clear case for it. For example,Experimental results seemed lackluster; see these critical comments about the paper version of this technical report, which claim that the correct results may have been achieved more through parameter-fitting than PCT.
Some found it more useful for explaining bugs in human behavior than functional behavior.
This seems to be a common category of complaints about Perceptual Control Theory.
This blog post called out that PCT "has no theory of information or how that information comes to be made," and this post grappled with a similar problem: struggling to find a place for implicit models, priors, and updates when working with a PCT framework. (This comment may have made a case for at least some embedded implicit priors.)
Pretty sure we had quite a bit more discussion of PCT on here in the last few years, would be great for someone to add it.
It's unclear whether PCT is a valid theory that explains anything. The advocates of the theory on the blog were unable to make a clear case for it. For example, see these critical comments about the paper version of this technical report., which claim that the correct results may have been achieved more through parameter-fitting than PCT.
Under-characterized information storage: This blog post called out that PCT "has no theory of information or how that information comes to be made," and this post grappled with a similar problem: struggling to find a place for implicit models, priors, and updates when working with a PCT framework. (This may have made a case for at least some implicit priors.)
It's unclear whether PCT is a valid theory that explains anything. The advocates of the theory on the blog were unable to make a clear case for it. For example, see these critical comments about the paper version of this technical report.
This blog post called out that PCT "has no theory of information or how that information comes to be made," and this post grappled with a similar problem: struggling to find a place for implicit models/priorsmodels, priors, and updates when working with a PCT framework.
It's unclear whether PCT is a valid theory that explains anything. The advocates of the theory on the blog were unable to make a clear case for it. For example, see these critical comments about the paper version of this technical report.
This blog post called out that PCT "has no theory of information or how that information comes to be made," and this post grappled with a similar problem: struggling to find a place for implicit models/priors when working with a PCT framework.
Perceptual control theory is T
Perceptual control theory (PCT) is a psychological theory of animal and human behavior. PCT postulates that an organism's behavior is a means of controlling its perceptions. The model of behavioris based on the principles of negative feedback [1]. It is to some extent an application of the ideas used in the engineering discipline of control theory to the modeling of the human mind and behavior.
It's unclear whether PCT is a valid theory that explains anything. The advocates of the theory on the blog were unable to make a clear case for it. For example, see thesecritical comments about the paper version of this technical report.
Friston's Free Energy doesn't have its own page. But I think a lot of the PCT-relevant conversation on LW, ends up under that term. Unlike the "mostly just intro-posts" under the PCT term proper, FFE seems to have more-recent engagement, so I think FFE has more of a presence here than PCT.
In retrospect, not including those was a mistake.
Here's a (non-exhaustive) handful of the FFE posts: