Transcript
Speaker: Hello and welcome to Imagine an Apple, the podcast about our inner mental worlds.
Speaker: My name is Francis and today our guest is Mike Johnson.
Speaker: Hi everyone.
Speaker: Also with me is Vin, my co-host.
Speaker: Hello Vin.
Speaker: Hey Francis, hi Mike.
Speaker: So Mike is formerly worked at the Qualia Institute but now is currently doing other stuff but
Speaker: Today we will be interviewing Mike about his internal experiences, specifically about the stuff that he specializes in, Tana, qualia, and active influence among some of the things we'll be talking about.
Speaker: So what we're interested in talking about today is inner experiences and how you might experience some of these phenomenal, your phenomenology, whether it's more visual or auditory or olfactory.
Speaker: And I was wondering if you have
Speaker: special kinds of phenomenology that you think are uncommon in general populations?
Speaker: Thanks.
Speaker: That's a really good question.
Speaker: So I would open with, you know, no one exactly knows how common or uncommon their phenomenology is.
Speaker: but we can sort of analyze base rates and different sort of physiologies.
Speaker: And yeah, I was actually talking about this with a friend today.
Speaker: And I was saying, well, you know, I think that I have...
Speaker: a somewhat unusual phenomenology.
Speaker: And they were saying, okay, cool, Mike, why do you think that?
Speaker: What makes you say that?
Speaker: And I actually wrote a decently long article about something very similar.
Speaker: It's called...
Speaker: autism as a disorder of dimensionality and it basically makes the case that uh people vary in a in a very interesting way so part of um physiology is understanding dimensions of natural variation um and uh kind of the the interesting dimensions that give rise to to interesting personality types or in this sense uh interesting phenomenologies and um
Speaker: my thesis in that piece was that people vary in terms of literally how many neurons they have in a unit volume of brain mass.
Speaker: And this is based on some research, some autopsies of children with autism, basically that
Speaker: And this is actually a pretty crazy result.
Speaker: They did autopsies and found that autistic children had on average 67% more neurons in the same volume of prefrontal cortex, which is a lot.
Speaker: It's not just 1% or 10%, it's 67% more.
Speaker: So that got me thinking about, okay, well, what does that mean?
Speaker: And can we sort of derive interesting things from this sort of one factor model?
Speaker: And my thesis was that, yes, we can.
Speaker: And I guess I'd sort of frame that as if you take a, so everyone's kind of familiar with how LLMs work, at least a little bit these days.
Speaker: And yeah,
Speaker: and sort of the challenges of if you sort of have a certain number of parameters in your LLM, then you have certain alignment challenges.
Speaker: Certain things work to align the LLM, and certain things maybe don't work as well.
Speaker: And in this case, basically, you'd be jacking the parameter count by 67%.
Speaker: And so you get into a different kind of
Speaker: different kinds of failure mode.
Speaker: And I promise this is answering your question.
Speaker: Yeah.
Speaker: So the observation that I made in the piece was that the natural sort of circuitry that evolution gives us
Speaker: sort of has pre-built optimization into it, that this is a very finely tuned evolved circuit that produces a specific result.
Speaker: But basically, if you make these networks thicker, if you take an optimized circuit,
Speaker: I don't know if you're familiar with in computing, there are ASICs, application-specific integrated circuits, basically a custom chip for a custom task.
Speaker: And it does one thing, but it does with this one thing really well.
Speaker: So like a Bitcoin miner.
Speaker: And then you have FPGAs, Field Programmable Gate Arrays, that basically it can be anything.
Speaker: This is a chip that you can reprogram at the hardware level.
Speaker: You can make it look like any other chip, or you can sort of solve any problem with hardware, but you actually have to do the programming.
Speaker: And so if this sort of autists have more thicker networks, we can say, more neurons and more connections between neurons, then their circuits are more like an FPGA than an ASIC.
Speaker: more like a general unoptimized network than sort of a sleek, highly tuned, thin network.
Speaker: Right.
Speaker: So yeah.
Speaker: So in this, so I'll pause a little bit.
Speaker: Does that all make sense?
Speaker: So I'm a little bit lost at the moment.
Speaker: I'm not entirely sure what an ASIC is or the other thing that you mentioned.
Speaker: And a PAG?
Speaker: and not entirely sure how emotion, how this relates to emotions and qualia.
Speaker: Can you
Speaker: expand a little bit more what you mean by dimensionality.
Speaker: What does that mean for autistic brains to have different dimensionality?
Speaker: Yeah, yeah, you bet.
Speaker: So an ASIC is a specialized chip.
Speaker: It's very optimized.
Speaker: And an FPGA is a much more generalized chip.
Speaker: It's not optimized, but you can program it yourself.
Speaker: And the sense is that
Speaker: If a brain has sort of extra neurons in it, it's more like an FPGA than an ASIC.
Speaker: It has a lot of potential in it, but it's not particularly optimized.
Speaker: Right.
Speaker: But what is it optimized to do?
Speaker: Yeah, yeah.
Speaker: So we generally, like a lot of circuits are pretty optimized in the brain.
Speaker: Like we have circuits to detect faces.
Speaker: We have circuits to detect noises and scary noises.
Speaker: We have circuits to figure out which foods are tasty and which are not.
Speaker: And the thesis is that people on the spectrum, on the autism spectrum, might
Speaker: have to do a little bit of post-production tuning on their circuits.
Speaker: They're not so optimized because it's just they have more neurons than what evolution was planning for.
Speaker: Right, okay.
Speaker: So does this translate to a different inner experience, so to speak?
Speaker: Do they have qualia that might be different from someone who's not autistic?
Speaker: Yeah, you know, that's, I want to say yes.
Speaker: And I think that part of that is
Speaker: every ASIC is going to be this.
Speaker: Every highly evolved circuit is going to be very similar.
Speaker: But every FPGA is going to be wired up a little bit differently.
Speaker: And so you're going to get a lot more diversity in the solutions that the brain finds.
Speaker: Oh, so I guess this would like I don't particularly make autism part of my identity or not part of my identity.
Speaker: It's just a it's just the word.
Speaker: But I do think that
Speaker: like this, I sometimes interact with the community and there's this focus on neurodivergence.
Speaker: And I think that's interesting because it's like everyone with autism is going to be different than people without autism, but also different from other people with autism.
Speaker: It's like...
Speaker: If you have a very small circuit, there's one way to make it work.
Speaker: If you have a very big circuit, maybe there are 100 different ways to make it work.
Speaker: And maybe you found solution 57, whereas someone else found solution 62.
Speaker: And I think that may show up in how sensory experiences present to people.
Speaker: Right.
Speaker: So here you're talking about the phenomenology of direct sense.
Speaker: So how people, when they're actually seeing something, experience seeing it.
Speaker: Is that what you mean would vary?
Speaker: Yeah.
Speaker: Yeah.
Speaker: I think that you actually get different results in a phenomenology.
Speaker: Okay.
Speaker: And in terms of the kind of more imaginative internal phenomenology, so like inner imagination and inner voices, that kind of thing, do you think use of those would vary or is that like a separate question?
Speaker: I think it's all linked.
Speaker: So yeah, I would also apply this to that frame as well.
Speaker: So imagination has a practice component as well.
Speaker: And for many people, they engage in some kind of attentional training.
Speaker: Meditation being a pretty good example of this, but other techniques, Alexander Tick, various other things also come into play.
Speaker: I'm aware that you are a meditator, am I correct on this?
Speaker: Yeah, that's right.
Speaker: Do you think that your meditation practices lead to different inner experiences, whether it regards to your imagination or how you experience qualia differently?
Speaker: Yeah, I would say so.
Speaker: If I had to say some words on that, I do think that just practice looking at things makes them more clear.
Speaker: And I would also say that, just like you mentioned, Alexander Technique, and I guess, especially lately, after my latest piece on vessel computation, I've been sort of practicing just
Speaker: there's this phrase, opening the hand of thought.
Speaker: And I think that practicing that is a very powerful move.
Speaker: And often there are clenches and clamps in our phenomenology and just practicing, OK, would it be so bad?
Speaker: Would anything bad happen if I just released and just tried to be an open experience?
Speaker: And I think that's very, very powerful.
Speaker: Yeah.
Speaker: Opening the hand of thought.
Speaker: I really liked that phrase.
Speaker: Um, you mentioned vasocomputation, which is something that I've seen you written about.
Speaker: Can you tell us more about what vasocomputation is and how that relates to Tana, which you mentioned quite a lot in your blog post?
Speaker: Uh, sure.
Speaker: Yeah.
Speaker: Uh, well, I I'm
Speaker: a bit afraid that I'm kind of throwing listeners into the deep end here.
Speaker: That's OK.
Speaker: We'll unpack it once a bit.
Speaker: Nice.
Speaker: Nice.
Speaker: Good.
Speaker: So I wrote this piece called Principles of Vaso Computation.
Speaker: And
Speaker: It's basically a unification of three or four domains.
Speaker: And the first domain is Buddhist phenomenology.
Speaker: And especially this term that the Buddhists call tanha.
Speaker: And this is often translated as grasping or thirst or clenching or desire.
Speaker: and uh according to the buddhists it's um responsible for you know maybe 90 of of all sort of moment by moment and so so it's kind of it's a big deal it's important just to i think just to make it clear what you mean by tan can you when you experience it yourself can you describe how that comes to you do you get like is it like a verbal thought or do you
Speaker: see things?
Speaker: What's the actual experience of having Tanha?
Speaker: Yeah, so there's, I would call it a contraction in awareness.
Speaker: And there's both the general contraction and a contraction around a specific part of awareness.
Speaker: And you can kind of gauge how flat your phenomenology is, in a way.
Speaker: And when Tanha is going,
Speaker: your phenomenology feels more lumpy, I guess I would say, for lack of a better word.
Speaker: Cool.
Speaker: Okay.
Speaker: So how does Tanha relate to this thing called free energy and active inference?
Speaker: And can you help unpack those terms and what they mean?
Speaker: Sure, sure.
Speaker: So principles of vessel computation connects three domains.
Speaker: And one is Buddhist phenomenology.
Speaker: And the second domain is the active inference framework, or the more full mouthful there would be the free energy principle active inference framework.
Speaker: And so this was developed by Fristen and others.
Speaker: And it's basically a framework how the brain is a prediction machine, and we sort of hallucinate our reality.
Speaker: Rao is trying to predict our sensations.
Speaker: And we live in that story, we can say.
Speaker: And active inference is another layer on top of this that says, not only do we make these predictions all the time, not only do we try to predict our sensations, but sometimes we predict
Speaker: false sensations and we hold these predictions until we act in the world to make them true.
Speaker: So this is a very subtle, important thing, I think, that, for example, if I'm thirsty, maybe I'll predict, oh, I'm not thirsty.
Speaker: I have the taste of water in my mouth and down my throat.
Speaker: Of course, that's not true.
Speaker: That's a false prediction.
Speaker: But I'll hold it in my mind until I go, I pour myself a glass of water and take a drink.
Speaker: Then I can release the prediction.
Speaker: Or if I turn the burner on on the stove, I'll hold the prediction that I need to turn off the burner until I turn off the burner.
Speaker: So I sort of keep track of the state of things by active inference.
Speaker: So yeah, this is a really cool framework.
Speaker: Sorry.
Speaker: I was going to say, so the Burnham one was super interesting.
Speaker: So there you said, once you've turned it on,
Speaker: you would have like in your present mind like a prediction that you need to turn off rather than predicting that it will turn off?
Speaker: Yeah, there's some kind of accounting system in the brain where, okay, you know, I know
Speaker: Now I'm turning this on and I need to sort of hold this state in my brain.
Speaker: And I know that I can stop holding the state once I turn it off.
Speaker: So we can sort of frame that as what I'm holding is the prediction that the burner is off.
Speaker: Gotcha.
Speaker: Right.
Speaker: Yeah.
Speaker: I have a question about this because you say that making a prediction
Speaker: example, that you are not thirsty is said to be a false prediction according to this framework.
Speaker: Whereas I typically interpret my desire to grab a glass of water to be a type of wanting, it's a type of desire, rather than thinking of it as a false prediction which I hold true until it becomes true by way of my acting in the world.
Speaker: And is there a difference?
Speaker: Is there a difference between falsely predicting and simply
Speaker: wanting or desiring, or are these like functionally the same thing?
Speaker: Same phenomena, same mechanism for achieving?
Speaker: Yeah, that's a good question.
Speaker: And I would say that with it, it of course depends on the mechanism.
Speaker: And with some mechanisms, it's implemented in such a way that they would be the same.
Speaker: And some mechanisms it would not be.
Speaker: So I guess this would be an argument for sort of dropping down and
Speaker: paying attention to implementation, which is the third domain, of course.
Speaker: Right.
Speaker: Cool.
Speaker: So can you tell us more about Buddhist conceptions of Tana and how that might have influenced your development of your framework in
Speaker: baso-computational theories?
Speaker: Yeah, sure.
Speaker: So I want to give a shout out to both Romeo Stevens and Nick Camerado.
Speaker: They've really done a lot of foundational work on thinking about Tanha.
Speaker: Novel translations of Tanha are like sorting through, okay, what's a good translation?
Speaker: and what's not.
Speaker: And first-person exploration that, you know, can we see this?
Speaker: Can we put a time frame on?
Speaker: You know, does it happen within 25 to 50 milliseconds or 100 milliseconds or whatnot?
Speaker: So they've done really, really great work.
Speaker: I think so my focus is generally sort of connecting these domains and saying, okay, this Buddhist tanha is sort of using this active inference system in an unskillful way, trying to control sensations in ways that don't make sense in some sort of like
Speaker: There are, in programming, we can talk about type errors, trying to, you know, add two letters together or multiply two letters together or something like that.
Speaker: It's just like, that's not the sort of thing you do with letters.
Speaker: And the tanha as unskillful active inference frames as the brain engages in a lot of that stuff when it tries to manipulate sensations, that it's often unskillful in the ways that it tries to...
Speaker: make predictions about sensations.
Speaker: And that in a moment by moment sense, this happens a lot.
Speaker: And this really adds up to like a lot of sort of submerged unpleasantness in your moment to moment experience.
Speaker: And that, I mean, the Buddhists say that, and if you fix this, if the Vipassana frame is, you can see this happening.
Speaker: And once you see it, you can't really unsee it.
Speaker: and you'll start to not do wrong things.
Speaker: So when you say see it, what do you mean by that?
Speaker: Do you mean like a literal seeing?
Speaker: I would say that when you observe what the mind is doing, I wouldn't say it necessarily flashes into your visual field, but I do think that everyone's going to be a little bit different in how it presents.
Speaker: Mm-hmm.
Speaker: But I would say that I think the takeaway would be sort of a bit of a shock, like, oh, wow, I'm doing that all the time?
Speaker: This feels suboptimal.
Speaker: And I think that progress on the path, on sort of getting better at not doing this, is sort of developing a
Speaker: a taste for what it feels like and a distaste for like okay no thanks like I don't need to do that to myself.
Speaker: In your blog post you use the word stress just in a passing and part of it is that potentially one of the ways someone would experience this so if you're grasping a lot partly that might add to your stress?
Speaker: Yeah, so I sort of break down the Tanha as unskillful active inference into three buckets.
Speaker: The first bucket is just if you try to control your sensations or control your environment more than you have the energy to.
Speaker: And we all know people that try to very, very tightly control their environment.
Speaker: And we can all kind of feel, OK, that runs into problems.
Speaker: And you just get stressed trying to control everything, and it doesn't work.
Speaker: I think what the Buddhists are pointing to, though, is a little bit different.
Speaker: It's that we try to apply active inference or predictions to sensations in ways that don't make sense, could never come true, or just are sort of bound to cause us a lot of suffering.
Speaker: Yeah, so a couple simple examples would be
Speaker: Like if a sensation is good, it's a nice, pleasant, tasty sensation.
Speaker: For example, if you're eating some cake and you really like it and you make the prediction, I will have this taste in my mouth forever.
Speaker: Of course, you're going to run into trouble because, OK, you're going to eat all the cake, trying to eat that taste in your mouth, and then you're going to be out of cake.
Speaker: And then you're just going to be left with tension.
Speaker: you can't make that prediction true and you don't have sort of any easy way to release that tension.
Speaker: The word forever in there is very powerful because yeah I wouldn't have thought I feel like I want the taste forever but maybe there is an aspect to that of the desire of the taste.
Speaker: Yeah.
Speaker: So, um, Romeo Stevens has this, uh, this nice frame that we tried to make our sensations stable, controllable and satisfying.
Speaker: And, uh, of course, uh, the Buddhists would say, you know, we can never ensure any sensation stable or controllable or satisfying.
Speaker: Um, so we're sort of doomed to discomfort when we assume that.
Speaker: Right.
Speaker: Going back, you mentioned before that there are three unskillful active inferences.
Speaker: And the first one was that if you try to control an environment more than you have energy to, that leads to stress.
Speaker: What are the other two?
Speaker: causes stress or unskillful yeah yeah so the the first is just trying to control too much um the second was trying to control in ways that don't make sense for example uh you know i'll have this taste forever um or you know if you if you drop a drink on the floor you make the prediction that didn't happen
Speaker: Well, it's not going to make it true.
Speaker: You can hold it, but it's not the appropriate prediction for the event.
Speaker: And then the third bucket or category of ways that we can be unskillful in using active inference.
Speaker: And I think this is really a cool, important thing.
Speaker: It's context desynchronization.
Speaker: So if you're doing a very hard math problem and you're sort of holding, using tension as medium term memory and sort of holding a problem in your mind by sort of let's pinch here and let's hold here and so on.
Speaker: And then someone knocks on the door and asks to get a cup of flour, a cup of sugar, whatever is the Xanadu scenario.
Speaker: Or someone says, oh, we have
Speaker: an emergency or there's a fire alarm or something, the context gets switched very rapidly.
Speaker: The tension that you're holding in your mind to remember the things in context one no longer has sense in context two.
Speaker: It's just stress.
Speaker: It's just tension.
Speaker: It's just suffering.
Speaker: It only has meaning within the right context.
Speaker: And it's like the, you know, if you're
Speaker: working on a computer program and the memory allocation changes, then your pointers are not going to point to the right things.
Speaker: And I think this actually happens a lot.
Speaker: And that a lot of our stress is sort of these fragments of tension that made sense in one context, but now we're in a new context where they don't make sense.
Speaker: Okay.
Speaker: So what is the, that's what unskillful active and looks like.
Speaker: What does skillful management's tension look like?
Speaker: Right.
Speaker: I'm not using it a lot.
Speaker: I mean, yeah, just kind of, you know, a little bit, a touch here, a touch there.
Speaker: We still need to navigate the world.
Speaker: And we may need tension to do that.
Speaker: We need, may need predictions to do that.
Speaker: And we may need active inference to do that.
Speaker: but uh don't like make as few as possible and don't hold them to skillfully choose your predictions and then like update them more often or replace them yeah yeah just uh hold them gently right so how does this now like lead to um predictions of and what is like the optimal level prediction or act inferences to do in order to
Speaker: have a healthy relationship, internal states with it?
Speaker: Right.
Speaker: That's a good question.
Speaker: I think that, as we were saying, being sparing with the predictions that understand that a prediction is a commitment to tension and a little bit of suffering until it comes true.
Speaker: But you also need them to navigate the world.
Speaker: I think that, so we were all at Just Camp together, where we met.
Speaker: And Vin and I were talking about the self.
Speaker: the self as kind of the sum total of tension in the mind.
Speaker: And I think this sort of nicely dovetails with some of the ways the Buddhists talk about the self and tension as sort of not the optimal way to navigate certain situations.
Speaker: So I guess I would just say that everything is contextual.
Speaker: There may not be a, this is always good, th is is always bad.
Speaker: But there will always be a dimension of, is this skillful and is this not skillful?
Speaker: Can you talk about what good and bad mean in this class?
Speaker: I know that's like... Right, right.
Speaker: Yeah, that's a hard problem.
Speaker: I mean, so in my more formal philosophy research, I study pain and pleasure.
Speaker: Right.
Speaker: And I'd be I'd be hesitant to say, yes, you know, pleasure is always good.
Speaker: Pain is bad because, you know, sometimes these sensations are trying to do things for us.
Speaker: And I think that I mean, good and bad could be framed in terms of the free energy principle.
Speaker: And sort of it could be framed in terms of adaptiveness to the environment and so on.
Speaker: So I'm not sure I have a sort of clean, like, this is what good is, this is what bad is.
Speaker: But just an appreciation.
Speaker: There's a lot of theories here.
Speaker: Yeah, there's a lot of context.
Speaker: Yeah, that's a pretty tall order to deliver.
Speaker: Anyway, just moving on from that, I suppose I really want to talk about qualia and
Speaker: I mean, you worked in an institute that had QALIA in its name.
Speaker: And I was wondering what you thought, what do you take QALIA to?
Speaker: And can you tell us more about that?
Speaker: Yeah, yeah.
Speaker: So I always introduce qualia as the components of subjective experience.
Speaker: And it's a little bit, you know, the easy explanation would be, you know, there's red, there's heavy, there's hot, et cetera.
Speaker: But what the actual natural kinds of qualia are is a very interesting unsolved question.
Speaker: Thanks, Mike.
Speaker: Yeah, you bet.
Speaker: Thanks, Vin.
Speaker: And thanks, Francis.
Speaker: Thank you.
Speaker: It was super interesting.
Speaker: Yeah, connecting to some really important things about how we experience our life every day.
Speaker: But also a theory that connects them up, which is great.
Speaker: you
