PCT
- Perceptual Control Theory
- good thread:
- the standard hierarchy levels, grouped into 3 categories
45 linked references
November 28th, 2025
- unifying PCT vs predictive processing - ARTICLE ON THIS
November 22nd, 2022
- PCT explanation of eureka moments
evolution includes WAY more than biological evolution
- PCT reorganization is an evolutionary process among the brain's neural circuits that converges on whatever solves the organism's needs
June 13th, 2022
reading Eliezer Yudkowsky on AGI Ruin › > 22. There's a relatively simple core structure that exp… › but I do also find myself feeling like I have directly en…
- left hemispherey stuff seems relevant here, as does PCT
February 22nd, 2022
- this article is from a sort of behaviorist lens overtly
- critiquing Alfie Kohn
- but seems to kinda strawman him and not engage with his actual studies
- (which Divia points out may not have replicated, although even so, to not mention them at all is unimpressive)
- and a lot of what she says sounds more PCT-esque ("sated") than stimulus-response
- the whole concept of an "intrinsically reinforced behavior"
- feels very phlogistony from PCT's perspective
- but also it makes sense!
- as long as you don't overreify who/what is doing the reinforcement
- I find myself wanting to replace "reinforcer" with "what works"
- and so you can only reinforce someone else if you are satisfying some sense of what works
- and only they are the deciders of what workingness is!
- like if you're playing chess and you make a move and your friend you're playing against says "fuck you", that might be a reinforcer! even though the same friend saying that during an argument would not
- (unless in some sense you were trying to get them worked up!)
- and you may have multiple control systems
- in this sense, non-praise feedback is letting them know that what they did works/worked, which is a reinforcement!
- and I see & like how it dovetails with what Divia's saying, in terms of "timing matters"
- like in order to have "timing matters" you need to be able to measure the reinforcement event precisely
- but it seems like the "timing matters" stuff fails to model all of the higher-order control systems that operate more slowly?
- like, entire strategies
- > It's considered the height of success to have a career based on your passion. That is, by definition. external reinforcement for an intrinsically motivating activity. Sometimes it works out great for people. Sometimes people find that what is great as a hobby absolutely sucks as a job. Always though, it sticks within the laws of learning.
- this is doing a bunch of subtle equivocation afaict
- and the equivocation is probably caused by assuming that money is always a reinforcer
- money isn't necessarily functioning as a reinforcer/reward here, depending on how the person's psychology is relating to the situation.
- it could be that they're just not thinking about money anymore and doing what they want to do, in exactly the same way they would if they had a UBI and were doing it for no pay (this requires that the job feels secure, eg a prof with tenure or whatever)
Podcast summaries
- Keywords - brain hemispheres McGilchrist's brain hemisphere model Perceptual Control Theory barn-raising Internal Family Systems blindspot blindspot feedback is hard (but possible!) it's only possible to be rational within a frame make shared or different frames explicit Change doesn't just happen Intend productivity Intend Philosophy & Paradigm how to think about goals PCT internal conflict much of human collaboration scales sublinearly but it's possible to scale superlinearly What are the components or dimensions required for collaborative culture? Geoffrey West Questions for teasing apart the concept of commitment commitment/choice is a confused way to relate to stances/mindsets Everyone is basically living in a dream mashup of their current external situation & old emotional meanings everyone is constantly overgeneralizing acceptance isn't the same as acceptance safety self-trust right hemisphere left hemisphere self-energizing
With Jonny Miller — Podcast with Jonny Miller Transcript › Outline of ideas discussed. (I'm M, Jonny is J)
- 29:11 - 43:49 - Reconnecting to desire, aliveness vs exhaustiveness Intend productivity Intend Philosophy & Paradigm how to think about goals PCT
With Jonny Miller — Podcast with Jonny Miller Transcript › Outline of ideas discussed. (I'm M, Jonny is J)
December 25th, 2021
- fascinating parallel to PCT from a hungarian physiologist
December 20th, 2021
- the PCT books and Unlocking the Emotional Brain have some interesting cross-connections, which I talk about in the convo I had with Eva de Hullu
Books or essays that answer each others' questions › meanwhile, some stuff that Book/Controlling People talks …
- and if one asks the Deutsch-inspired question of how this happens neurologically, or even something like what is necessary to achieve this, neurologically?, Unlocking the Emotional Brain and PCT both have some answers
- Controlling People in particular talks a lot more about the reorganization system, and so if you want people to not try to outsmart culture then it would naturally be necessary to keep people from consciously inspecting certain things!!!
November 24th, 2021
20:49 in evening convo, maybe like 1.2h in › what is coercion? › sort of a response to something Selentelechia asked about
- using the lens of control systems & PCT
November 12th, 2021
- Read this very post-average article on the implications of PCT
- I was kind of disappointed by it
- it's mostly taking a fairly academic stance
- which I guess makes sense in context
- but I was hoping for a bunch more of the societal implications, which are sort of in there but buried under academic frames
- the other thing I noticed about it was that it was still so so reactionary against behaviorism
- and there's value to this but I find myself wanting an articulation of PCT that barely references behaviorism, and instead just talks about how much sense the theory makes, on its own terms
- and then maybe sort of after conveying its whole frame in such a coherent way, it spends some time talking about how behaviorism is confused?
- idk, there's value to Ruling Out Everything Else
- Book/Making Sense of Behavior is pretty good at this, though it does spend a little time discussion the confusions of behaviorism
October 23rd, 2021
- Matt Brockwell > apophatic knowledge
- knowledge of what to block out
October 16th, 2021
- I mentioned PCT to someone and they named "ecological psychology"
- someone also pointed at this Gap & Gain model
- > I showed him that he was defining his success by mea-suring where he was—the goal he had achieved—against the ideal he had set out for himself. I said that the distance between his goal and his ideal is “The Gap.”
- the gap is important for motivation, from a PCT perspective
- but if you define your success by that, then you're making a fundamental confusion
- > Our ideals are a lot like the horizon in this way. They’re useful to us to figure out a direction and to plan a destination, but you can’t reach your ideal any more than you can arrive at the horizon. No matter how far you’ve traveled, the horizon will always be far off from where you currently stand.
Book/Making Sense of Behavior
- best introduction to PCT
September 25th, 2021
- quotes from Book/Making Sense of Behavior, PCT
- > Since the world we experience is the world of perception, it
usually makes little difference whether we say we’re controlling perceptions or controlling the state of the real world. It does, however, make a difference when we try to explain behavior in terms of some physical model of a behaving system. - > The two problems go together: the problem of reaching agreement with each other about reality, and the problem that all perception is fundamentally private. If our perceptions were not private, we would never disagree with each other about the world and its meanings. If they were not, somehow, similar, we wouldn't be able to tell jokes or communicate about anything. It is, of course, possible that our perceptions are both private and similar.
- gonna make a PCT meta-thread at some point, and this will go in it:
June 26th, 2021
- difference between needs & wants, in PCT terms
- one thought
- needs = intrinsic reference levels
- a few hundred of them
- many of them are nutrients, minerals, etc
- also calories, sleep, secure attachment, physical safety
- wants = all other reference levels
- but they can be quite fixed
- maybe by trauma
- maybe because identity doesn't want to change
- maybe something like, built-in inertia
- Jay Mustafi > persistent desires
- also treadmill of upgrading your sense of expectations
- eg when you make more money
- M > epictetus or that other stoic who lived like a dog
- M > creativity - people don't even know what's possible until they see it
June 15th, 2021
- starting a thread on PCT/reward is just the temporary relief of punishment - tweeted
- Reinforcement learning is backwards because it implicitly says "when reward, behavior change"
when actually, change is the *default*, and the thing being called "reward" says "stop changing, you found it"
This figure-ground inversion is key. - The whole concept here of PCT (Perceptual Control Theory) is that brains are in some state of organization that is set up to produce a set of controlled perceptions, like getting money to buy food to not be hungry, or telling jokes to make people laugh to feel connection.
- When something is perceived to be at an unwanted level (eg hunger, safety, connection, thirst, sex, meaning) the system goes into a state of reorganization, inhibiting old strategies and trying new ones.
Once the error is resolved, the music stops and the organization stays. - This is independent from whether the new strategy is actually the thing that solved the problem (which is where superstitions come from).
- There's also a question of what happens when two strategies for two different needs conflict.
Often this produces oscillatory tug-of-war dynamics, but if you can gently hold all of the desiderata at once, there's usually some sort of win-win to be found.
twitter.com/nickcammarata/status/1403558910984966150 - One important implication of this whole thing is that you can only motivate/train someone with reward *to the extent that they are otherwise deprived of the thing* and
May 27th, 2021
- thread related to PCT
April 8th, 2021
- reward is just the temporary relief of punishment - not my original thought - comes straight out of PCT
Home
- When I use the word "need" or "want" or "careabout" or "desire" in this piece of writing, I'm pointing at, fundamentally, a cybernetic control system set point. See PCT or PCT Error for more about this.
- What makes one a need, vs a want, might be essentially, as my friend Catherio put it:
- "if this goes unmet, I will make substantial changes to my life"
- Said needs & wants include things on every level from psychological safety & meaning to food & water & shelter, and probably stuff Maslow forget to mention.
March 15th, 2021
re fractal Home › something called Dragon Dreaming › oriented towards eliciting everyone's dreams and orientin… › finding underlying things that can be cared about that ar…
- PCT-style! / Method of Levels
Malcolm podcast appearances
List › November 5th, 2021 on Curious Humans with Jonny Miller › The Art of Asking Questions, Incentivising Negative Effor…
February 24th, 2021
- things we need better language for - the ones below are all broadly related to PCT
- intersubjectivity
- Trinley Goldenberg > another thing we need language for is this distinction between
- everything you're aware of is your own perception / in your head
- there's still a territory though
Intend and Improvise
- talk about PCT for a bit - what is a goal/intention/etc?
- gap between desired state and perceived state
- can be on different levels - give basic physical examples
- can involve unclear end-state, as long as recognizeable, etc
- lower-level ones in service of higher-level ones; sometimes get dropped completely
- also conflict
- exercise: reflect on some of the goals/intentions you've held during the hack-week
- ideally finding at least one of each of these:
- something you finished
- something you dropped
- something you're still intending
- for the one you dropped, did you drop the intention it was in service of too? can you go up levels until you find an intention that didn't drop?
February 18th, 2021
call with Qiaochu, co-reading PJ Eby's Self-Help Myth › > What if you could really do everything effortlessly? Ge…
- in a command hierarchy, higher-ups tell lower-downs "what to do"
- in a control hierarchy, the higher levels tell lower levels "what to achieve" and the lower levels can reorganize however they want to achieve that
call with Qiaochu, co-reading PJ Eby's Self-Help Myth › Q > something feels bad to be made out of thermostats! › Q > uhghhhhh... I'm imagining my body made out of thermos… › M > yeah, blows my mind that all these models seem good
August 23rd, 2020
- also a PCT book by William Powers
thingness isn't conceptuality
- fill this in with case study re PCT levels & turning off thingness
David Deutsch ↔ McGilchrist's brain hemisphere model
- M: PCT & disrupting objectifier layer - thingness isn't conceptuality
- C: stan brakhage films
- C: I do a similar thing with my kids, we find random objects and try to come up with stories
- although I guess that's not on the level of objects, but more like on the level of meanings