with Spencer Greenberg
the podcast about ideas that matter

Episode 088: How many minds do you have? (with Kaj Sotala)

Enjoying the episode? Want to listen later? Subscribe on any of these apps or stores to be notified when we release new episodes:

January 20, 2022

What are the advantages of viewing the mind through the multi-agent model as opposed to (say) the rational / optimizing agent model? What is the "global workspace" theory of consciousness? What's going on during concentration meditation according to the global workspace theory? If our brains are composed of multiple sub-agents, then what does it mean when I say, "I believe such-and-such"? Are beliefs context-dependent (i.e., you believe P in one context and not-P in a different context)? What effects do the various therapeutic modalities and meditation practices have on our beliefs? What are the advantages of transformational therapy over other approaches?

Kaj Sotala is interested in finding ways that would allow everyone to reach their fullest potential with regard to agency, inner harmony, and well-being. He believes that the upper limits on these are somewhere very high indeed. Kaj has worked as an emotion coach, software developer, and researcher focusing on the long-term consequences of advanced artificial intelligence. One of his last research projects was the Multiagent Models of Mind article series, combining perspectives from a variety of fields ranging from psychotherapy to AI and neuroscience, for better understanding how the mind works. His website is, and he also posts articles on You can also follow him on Twitter and Facebook or email him at

JOSH: Hello, and welcome to Clearer Thinking with Spencer Greenberg, the podcast about ideas that matter. I'm Josh Castle, the producer of the podcast. And I'm so glad you joined us today. In this episode, Spencer speaks with Kaj Sotala, about theories of consciousness, how the mind deals with multiple processes, and transforming self-perception.

SPENCER: Kaj, welcome. Thanks for coming on.

KAJ: Thank you. Delighted to be here.

SPENCER: You and I have had some fun debates on social media and unless you're on-site, I was really eager to get you on, because I really enjoyed reading some of the articles you've written about how the mind works. And I think you have a lot of cool insights. I also think we might have some interesting disagreements as well, which I'd like to dig into at some point. But you want to start by just telling us a little bit about this model of the mind as a bunch of sub-agents.

KAJ: Yeah, so the way I would describe it would be to sort of comparing it to it, what I could call a classical way of thinking about the mind, which is something that might say, in economics, or AI, or just generally people's folk impressions of how the mind sometimes works, which is basically that people have as goals that they are trying to pursue, and people have some current construct beliefs about what they should be doing. And that they are just pursuing the goals, according to the best ways to believe they should be pursuing the goals. So for instance, like, maybe you want to become rich, and then you believe that buying lottery tickets is the best way to get rich, which might or might not be an accurate belief, but at least you might be acting according to your beliefs if you're buying lottery tickets.

SPENCER: Is it fair to call this the rational agent model of the mind? Or is that unfair?

KAJ: Yeah, that sounds fair. But the pairs, this alternative way of thinking about it, which pop up in common speech, which is something like one way it is sometimes expressed is that you might be facing a decision. And then you might be saying, well, there's one part of me that wants a piece, one thing, but then another part of my mind wants this alternative thing. And it's almost as if you're talking of different parts of your mind, having different preferences, possibly even different beliefs, and places to kind of a model that you also find in places like some forms of psychotherapy, science to it in the theory of meditation, some neuroscience models, and I think it is an alternative way, where you basically think of your mind as being composed of many parts with conflicting beliefs and preferences is often in many situations, it feels like a more fruitful way of thinking about how the mind actually works.

SPENCER: So I discussed this multi-agent view with [unclear] on the podcast a bit, but I think that you're going to have somewhat of a different take on it. And so I want to get your opinion on this. But how is this view of multi-agents different than just saying, “We're this sort of approximately rational agent, you know, far from perfect, but an approximately rational agent, and we have multiple goals so sometimes our goals come into conflict.”

KAJ: Yeah, so the distinction is, I wouldn't consider it a clear call to the difference you can, in some situations, it's more useful to think of it as one unified agent with shifting goals and depending on the situation. So one way of looking at it is that we have these different situations, where before you go into a situation, you think you know, what your goals are, and what you want to be doing. But then when you actually go into the situation, you find yourself acting completely differently. Like maybe you and your romantic partner always have a taste stupid argument about some particular thing. And then, after having had tough many times who decided, okay, next time his thing comes up, I'm not going to argue about it, I'm going to be smarter about it. And then the situation comes up anyway. And then you find yourself getting drawn into exactly the same pattern that you are always having had to before. And if we were just having different goals, depending on the situation, that's kind of hard to fit into the framework.

SPENCER: Yeah. And just a simpler level. It seems pretty clear that we don't have stable goals. The model where it says, well, we just have a whole bunch have goals as soon as they come into conflict doesn't really work unless you add this extra ability for us to change our goals on the fly, right? Like in the example you gave where, you know, it seems like you come in with a certain goal, and then your goal suddenly changes. And that seems to be a problem for this view of like, we're sort of just an optimizing agent, maybe optimizing agent is a better name for that more traditional view than a rational agent, because it's not that we're rational, it's that we have these goals, and then we're trying to optimize towards them. And that seems to contradict this idea of sometimes our goals seem to suddenly change when we put it in a situation. And then they change back where we're out of the situation again, right? So there's something really unstable about them. So we've got this view of humans as sort of multiple agents. Do you want to flesh out this model a little bit more? Do you want to say some things about like, what's the nature of these agents? Or how do they interact with each other, anything like that?

KAJ: So maybe just starting out by some fairly ordinary examples, right now, when you're having a conversation with me, whatever you feel like you are trying to do in this conversation?

SPENCER: I feel like I'm trying to, one, produce a conversation that will give people value, two, give you a positive experience and make you feel comfortable and three, nature has something to say, which is there is a loop like okay, picture something to say,

KAJ: Yeah. And if I'm thinking about what is my mind trying to do, like, I have a part of my mind that's looking at my notes for this conversation and trying to pick out sensible things to be saying, and more generally, trying to produce the kind of an explanation of my model that sounds good. And because we're talking in English, that's the kind of a situation where another part of my mind is paying attention to how I'm pronouncing different words and trying to make sure I'm pronouncing them right.

SPENCER: It creates this extra cognitive burden that you had to deal with that a lot of my guests don't have to.

KAJ: Yeah. And luckily, I don't feel too tired or hungry right now. But if I did feel that way, I might have like other parts of my mind that were trying to make me stop having this discussion to make me go eat or sleep instead. But fortunately, I'm not having that problem.

SPENCER: Can I just do a meta-comment on this? I love this idea of talking about this experience of podcasting, while podcasting. This is like the circling version of podcasting.

KAJ: True.

SPENCER: It's like someone could create a whole podcast that's in this theme of like, what are we experiencing right now? I think maybe Ella actually tried to do this. So I think we're both experiencing this sort of multi-agent phenomenon right now, right, we both have these kinds of different drives or motives occurring simultaneously.

KAJ: Yeah and sort of the simplest possible way of explaining the idea would be if you are just thinking of, sort of like specialized programs, in the same way, that you might have computer programs on your computer, or a very simple example would be if you had a robot, let's say cleaning robot has basically had two different programs when pure robot happens to have a lot of electricity in its battery, can it just go around cleaning places, and then, when it notices that it's low on battery, can it switch to a completely different program just turns off the cleaning program and goes back to recharge? And obviously, humans are a lot more complicated than this. But you can kind of think of yourself acting in a similar way, like we are here, having this conversation, and tap activates a certain set of priorities that we might have. And when we stop recording this and go spend time with friends, we'll have other things that we care about, want to think about. And if you happen to get angry at someone then you would think and act in a different way than if you were in love with them. Or if you were jealous of them, and so on. So one way of thinking about it might just be that depending on our external and internal environment, we get different priorities kicking in and tear something into the brain, hearts recognizes a particular situation and thinks in a situation it should be running this subprogram or this set of priorities. And obviously, the example of the priorities we have going on currently, such as did in humans, we usually have like, multiple sets of priorities in any given situation, and multiple sets of sorts of subprograms or parts of the mind are running and trying to make sure we achieve whatever our brain thinks we should achieve in that situation.

SPENCER: Yeah, the way that I think about it, and I'm curious to hear whether you agree with this is that we have all these different motives and they get activated at different times. So maybe we have it I don't know, 50 possible motives or something like this, or 100 or 1000, I don't, I don't know the actual number. But then at a given point, maybe two motors will actually be active, right, like, I'll be hungry, and I kinda want to go eat. But also, I'm like trying to pay attention to something, and I don't want to stop paying attention to it, right. So there's the like, keep paying attention to the thing, you're engaged in motive. And then there's the, I want to eat motive, and then there's kind of a conflict, and there's some kind of bid each one is making and each one can sort of be intrinsically stronger or weaker. Like some motives just feel like they inherently have more strength. But then in addition to that, based on the context, like some of them will, like bid harder, and some will bid weaker, right? Like, if you're a little bit hungry, you're gonna not bid for like going and eating as much as if you're really, really hungry. Like, if you haven't eaten in days, the hunger motive might just be all empowering. And like, if there's any chance to get food, it might just override all the other motives. So I'm curious, do you agree with that framing?

KAJ: Yeah, that sounds pretty accurate. Yeah. And I might also say that, at any given time, we have some set of motives that are active at that current moment. And at the same time, we have all of these motives that could be active at that moment. And our mind is constantly sorting of monitoring for things that have been important to notice before. I think at this point, I could have mentioned a few sorts of fields, which have been talking about this kind of thing, just to make clear to our listeners that we're not just speculating.

SPENCER: Yeah, sounds great.

KAJ: So as some people know, in AI, terawatts, this like classic book in the 80s, by Marvin Minsky, Society of Mind, which was basically just speculating on the topic, but like, I guess, the main three directions or fields where I first noticed this thing, and it got me interested, more interested in the topic. And I realized that several fields that seemed to be talking about the same thing from different directions was basically psychotherapy, neuroscience, and some theories of meditation. So in cognitive psychology, and cognitive neuroscience, there's what's called a global workspace theory. That's basically one of our best-supported theories of consciousness, which basically says that consciousness is a place that the different parts of the brain can use to communicate with each other. Like the brain has all of these subsystems, doing stuff, and trying to send information into what's called a global workspace. And there's this constant competition of who gets to, what things tear then whatever gets put into a global workspace is sort of information that gets broadcast to all the systems. So for instance, if I am talking to you right now, and I hear what you're saying, and probably if I'm not being distracted, by the content of your words, is considered important enough, it gets into my consciousness and into my global workspace and then all the subsystems in my brain that might be listening to the content of that workspace, get to hear what you're saying. And there might be a systems test or paying attention to particular things. Like if suddenly, you were to sound angry, then, some subsystem in my brain tap was picking up signs of potential threat, might notice it and react to it some way, such as making me more anxious or whatever,

SPENCER: All these different kinds of modules are listening, or watching this global workspace and reacting to it essentially, right?

KAJL Yeah, exactly.

SPENCER: This unpacks the word consciousness for a second because I just find that people are extremely confused about this word because people mean it in a lot of different ways. Usually, when I think of consciousness, I think of what it's like to be a being or what it's like to have experienced. So a chair is not conscious, but human is because there's something that it's like to be the human right. There's an internal state in the human. And so is that what you're talking about, like the experience we have of being aware of anything?

KAJ: Yeah, roughly. If we are to be the exact definition of consciousness used in global workspace theory is conscious access. That means that we have conscious access to information if we can report that information to other people. So if I hear what you're saying, I can report to you the content of your word.

SPENCER: Right. So you can't report your kind of heart rate. Normally, maybe if you maybe could learn to pay really close attention, but normally you can't. But you can report the thought you had or the fact that you're feeling hungry or something like that.

KAJ: Yeah.

SPENCER: I like the definition because it makes it really clear cut, kind of sidesteps a lot of philosophical issues. Okay, so can we think of it then sort of like a spotlight in the mind and then these different sub-agents in our mind can kind of stick something in the spotlight or compete to put something in the spotlight, but only one of them can have something in the spotlight at a time. So they're in a competition, they're trying to put them in the spotlight, there's only one thing in the spotlight and then another agent can kind of swap it for the thing they want to be in the spotlight and so on.

KAJ: Yeah, that's pretty good. And I also liked the spotlight metaphor intended to help highlight tack, even though we usually feel like the only thing that we are conscious about exists. A spotlight sort of help highlight the fact that the brain also has a lot of stuff going on that we are not conscious of. And there's a lot of information that's getting processed to doesn't make it all the way to conscious awareness. But the stuff that does get put into consciousness, it is sort of like, strengthened and gets everyone special attention and sort of has all or at least most of the modules paying extra attention to it on top of whatever else they might be focusing on.

SPENCER: Right. Because presumably, the different parts of the brain are communicating anyway, in sort of the darkness, right, without us being aware of it. And this is just another form of communication that maybe has special properties. Is that right?

KAJ: Yeah.

SPENCER: One of the coolest studies on this that sort of illustrates that different stuff is happening without our awareness is these cases where someone will have brain damage, and they won't be able to see. But then if you to toss them a ball, they'll still be able to catch it, which is super strange like that somehow, some part of their brain actually can still see, because it's not that their eyes are damaged. In this case, it's that there's damage in the brain that's preventing information from flowing to certain parts of awareness, even though on some level, they can still see. I mean, I haven't dug into the details of the studies to kind of confirm them. But this is at least what the studies claimed to show.

KAJ: Yep. And another like fun, clash of studies is terrorists that are saying terrorist's cases where people are learning to do something, for instance, maybe are showing him two countries of pictures, and they are supposed to recognize different kinds of issues that might have some medical problem or might not. And you're trying to teach them to diagnose the pictures of the issue. And then what might be noticeable is that at the point when they are starting to get the hang of it, you can sort of pick up subconscious signs of him starting to have a sense of which pictures are the correct ones and to which ones are not before they actually have conscious awareness.

SPENCER: Oh, it's right. So this is sort of the idea that when we start learning to do a new skill, we often do it very consciously where we're like thinking about what we're doing, like okay, I've got to first move my left hand and put in this position, then my right hand. Jujitsu would be a good example. You're like learning a new jiu-jitsu move in Brazilian jujitsu. And it's like, your first hand, I'm like, really thinking about, I gotta move my leg and then my arm. And then after a little while, you just kind of start, you're like subconscious starts executing. It's kind of like that. You can eventually learn to do it in a way where the subconscious starts doing it before your conscious mind even realizes that.

KAJ: Yep.

SPENCER: Cool. So is it an aspect of this workspace theory that only one thing can be in the spotlight at a time?

KAJ: So as far as I know, there's some debate about it. But the sort of standard interpretation of it is that only one thing can be in the spotlight at any time. But there's also some debate about what exactly is one thing, but roughly –

SPENCER: Yeah, cuz it does seem that the conscious mind is very serial-like as opposed to parallel, right? Like it does kind of one thing after another. And it's very hard to do two things at once. And when I tried to pay attention to myself multitasking, it seems like I'm actually just switching back and forth really fast. And that like I can fake parallel tasking by this rapid switching. But it's hard to pay close enough attention, like such fine-grained attention to really be totally confident and you're not doing two things at once, in your conscious mind. At least I'm maybe I'm not at that level of meditation, you're paying attention to be able to notice for sure,

KAJ: Yeah, that's sort of the standard interpretation. As far as I know that your mind is actually switching between tasks very quickly, or switching between pieces of conscious content very quickly, rather than actually having multiple pieces of content. And that's actually also the claim made in some of those theories of meditation, that is suggesting some of those theories or framing things like concentration practice in terms of, okay, you have all of these sub minds in your brain, which are trying to direct attention into different things. And they might have disagreements about what you want to concentrate on. But if you sort of try and turn into noticing that actually concentrating on one thing feels really good. And according to these frameworks, the way you make progress in something like concentration meditation is to just teach your brain internal noticing how good it feels to concentrate on one thing, and then in tough situations, at least all of your sub minds will agree to be focusing the attention on your breath, or whatever you are concentrating on.

SPENCER: So what's happening in concentration meditation, according to global workspace theory? Are you kind of giving full access to the spotlight to just one sub-agent? Like take a case where you're trying to focus fully on your breath, we're really common for meditation, right? And the thing that everyone experiences who's not credibly in meditating is that their brain constantly wants to not focus on the breath, right? Like their mind, you know, will dart between the breath and then they will notice some thought, and then it will notice that you've got a pain in your leg, and then they'll notice that there's noise over there. And then I'll notice, notice an itch, and then maybe we'll go back to the breath, right? And then you're kind of bouncing around, you're trying to keep it on the breath, but it's bouncing around. But like, as you get better and better, and you start paying more and more attention to the breath for longer and longer. What is actually happening?

KAJ: Yeah, so one of the assumptions in these theories is that your brain is constantly learning about how exactly case different sub-agents should be given more power to control take the content of your consciousness in any given situation. And one way of thinking about what's happening in training concentration meditation will be. So suppose you manage for a little bit to actually follow your breath. And you might notice that it feels kinda good to be following your breath for a moment. And then you get distracted thinking about whatever you are thinking and 10. At some point, you remember that oh, yeah, I was supposed to be meditating. And then if you are sort of approaching this from a view where you think that oh, great, it's great that I noticed that I got distracted. And the theory would say that there was some particular sub-agent that was doing the monitoring and happened to come online in order to notice that you had actually been distracted. And maybe that sub-agent gets a tiny bit of reward, making it more likely for it to be activated again. So if that sub-agent just notices destructions, and focuses your attention back on the breath, keeps being rewarded, when it gets more likely to activate again, and then your periods of destruction might get shorter and shorter. And as they do so, you might, again, if you're paying attention to what's actually happening in your mind, you might notice that following your breath is actually feeling nicer, hadn't get distracted all the time. And then, if all goes well, he sub-agents tat or directing your attention on the breath, keep getting rewarded and then, you're more likely to spend longer times actually focusing on the breath, or whatever you're focusing on and feeling good.

SPENCER: Would that suggest that if you feel frustrated when you notice you're no longer meditating properly, like, you know, you notice that you're distracted, and you think, oh, man, I suck at this or something like that, that actually might have the opposite effect and actually make it harder and harder for you to focus on the breath.

KAJ: Yeah, that's actually emphasized in some of the meditation frameworks, I'm drawing on how it's like, actually pretty important to try to be sort of compassionate with yourself and not get too frustrated over whenever you get distracted because you might easily be actually training yourself not to notice destruction if you just feel annoyance at yourself whenever you notice it.

SPENCER: Interesting. I always assumed that those instructions were there, which are really common, just because it would be really unfun to meditate, and you get discouraged. But that's super interesting that it's actually maybe much more important even than that, that it's actually about the way you're rewarding your sub-agents. But let's say you actually achieve this state, at least for like, you know, a glimmer of time, you know, 10 seconds, where you really are just focusing your breath, you really are undistracted. What's your model of that is that just one sub-agent you've kind of given the full spotlight to?

KAJ: Um, so at this point, it is like, getting sort of the model I'm having at in this particular case is getting a little fussy about what exactly is one sub-agent, what do we count as several sub-agents. The theory I was reading was like, putting it in terms of many different sub-agents all agreeing to focus the attention just on one thing, but whether that's actually like one or several sub-agents on a neural level, I don't know. Could be either.

SPENCER: It seems like a pretty difficult question. Because you imagine you have a bunch of Python or C++ code, right? And within that code, there's a whole bunch of algorithms, right? And you might ask the question, well, how many algorithms are there? That actually seems like a nice its, not a very well-defined question, because you might have one function that calls three other functions, and each of them calls other functions. And it's like, what do you count as a separate algorithm? And so it's not that clear to me that there even is a clear-cut answer to where one algorithm ends and one begins. What are your thoughts on that?

KAJ: Yeah, I agree. And I think that's similar to what's the case here, with sub-agents into the brain, like there or some situations where we can roughly say that, okay, it makes sense to talk about his particular thing as one sub agent. But like one thing, where this also comes up is that sometimes you might have a source of different memories, pulling you in different directions. For instance, maybe you have had positive experiences meditating. And the memories of those experiences are encouraging you to motivate. But you've also had some bad memories, which are also getting activated. And then, it's kind of like, is each memory trace in your brain, its own sub-agent, its cluster of related memories, its own sub-agent, its general memory system, the sub-agent and the different memories, just patterns within the subsystem, it gets fuzzy.

SPENCER: I also want to ask, we're working memory fits into this model, I'm not sure whether it is part of this model or not. But it seems like we have this little memory buffer that we can store a certain amount of stuff in. You know, like, if you're thinking about doing a math problem, like, it would be really bad if when you're adding two numbers, you couldn't think about what they're gonna then get added to afterward, like, you just lose your thread. And so you have to build keep a few things in mind. But also working memory seems very limited. Like, for many people, it seems to build a store maybe like seven numbers, or that kind of thing, or maybe six or maybe eight, but you know, not 20. So is that part of this model at all? Or is that just a separate idea?

KAJ: Um, so the interpretation for that workspace itself, or your consciousness itself, as we mentioned, it can have one piece of content at a time. But at the same time, there are these buffers that you mentioned, that can sort of taking a thing from consciousness, hold it active outside consciousness for a while, and then fade it back in and help to keep refreshing pretty regularly. But for a while, take and maintain several pieces of information.

SPENCER: This is a kind of side tangent, but it's kind of interesting to me, it seems that in addition to working memory, where we can keep like a few things in our mind, we have this replay loop where whatever happened, like the last, I don't know, five seconds, whether it's a sound or something you saw, it seems like we can just replay it. And we can use this to try to pick out information that we missed the first time. Like, I don't know, if you've had this experience where someone says something, and you're kind of distracted, so you didn't really pay attention, but then you kind of replay what they said, and you're like, oh, okay, now I got it. It's a kind of interesting kind of memory.

KAJ: Yeah, I've had that same experience. I'm not sure what exact neuroscience on that is. But I recall that being thought that, besides consciousness, we also have a like unconscious buffer of just maintaining whatever information happened to come up within the last few seconds. And it has that exact same role that you mentioned that if our consciousness happens to be busy processing something else in a brief period of time for which we have actually stored the information before it gets lost entirely.

SPENCER: Yeah, it makes sense with regard to this spotlight theory. Because imagine the spotlight is currently being trolled by one sub-agent. But then something interesting happens. It's very important to another sub-agent. And then other sub-agents like, “Go replay the tape. Look, there's this thing that just happened,” right? Like you missed it because you were distracted by doing this other thing. But like you could see it being useful as a way of sub-agents like pulling back to attention, something that was missed by the sub-agent, those currently in control at the moment something happened.

KAJ: Yeah. And also the third field I wanted to mention the sub-agent idea is that in psychotherapy, you often have these models of having some kind of a part of your mind or some kind of a sub-agent that you originally learned in childhood, or when you are younger, which was useful at the time, but doesn't necessarily produce very good reactions right now. So for example, let's say, suppose when you were a child, you wanted to play with some other children, but perhaps other children didn't like you, for whatever reason. So if you tried approaching them, they would it be mean to you to tell you to go away and As a result, you developed some social anxiety. You learned not to approach other children and learn to play alone instead. And in that situation, that was actually a reasonable strategy, because by not approaching them, you avoided the pain of rejection, and of children bullying you and you get to play alone, which was my third time getting picked on. But then when you become an adult, that particular theory isn't necessarily relevant anymore. You can just choose to be friends with people who like you. But regardless, you might still have these early sub-agents that learned to notice situations when you're approaching other people and learned that this is their interest, it should be avoided. And whenever it notices you approaching other people, it will trigger that same sense of anxiety. And then an adult, you might need to do some work with that to sort of overcome that old programming.

SPENCER: That seems to relate to things like PTSD, where someone will have an extremely traumatic event happen like they're in a war, and one of their friends gets shot right in front of them. And then later, when there's, let's say, a loud noise, it sounds like it could be then shot, but it's not a gunshot. You know, they have a sudden extreme fear reaction. It's like they've learned this behavior, like, “Oh, I'm in an extreme threat.” And it gets associated with this pattern, like a loud noise. And you know, in their home environment where there are no gunshots, this is actually dysfunctional, but it was functional in the kind of wartime environment that they learned the pattern.

KAJ: Yeah, exactly. That's sort of an extreme example. And well, some of these old sub-agents, might also be similarly extreme if your childhood was really bad. But at the same time, there might be all kinds of behaviors that you don't necessarily even know the kind of origin, like terror, might be some things that you just generally vaguely feel like, you always end up procrastinating on or just feel vaguely unpleasant, and you aren't sure why exactly that is. And it's not the kind of sort of capital letter Trauma that you are having. It's just something that your brain noticed that, oh, maybe this should be avoided. And it keeps doing it even if you aren't quite sure why exactly

SPENCER: Those who remind me of the idea of transference from like kind of the psychoanalytic tradition. Where, as I understand it, transference is when, let's say you're seeing a psychoanalytic therapist, and you start acting as though the therapist is a person in your life, like, for example, I think, typically, it'd be like a person from your childhood, right? Like, maybe you throw a tantrum, the way you would, as a child, maybe not as extreme, but like some sort of tantrum in reaction to what the therapist does. And you're kind of treating them as though they were a parent figure. I don't know how strong the evidence is for transference. But it seems like we sometimes do have these kinds of saved-up patterns, and then being with a person can kind of put us in the same role that we were in previously, for example, as a child, and then that starts bringing out more childlike behaviors. I think a lot of people notice this, like when they go home to visit their families, and they suddenly start acting more childlike.

KAJ: Yeah, and to take an example of going home, and suddenly finding yourself acting in completely different ways. Actually, one of the examples I was thinking of maybe using, like, it's such a strong shift, that thriller makes it obvious how, actually, when you are in this particular situation, just an alternative set of priorities completely kicked in, and it's kind of weird to notice it.


SPENCER: Okay, so we've talked about this multi-agent model from different perspectives from meditation from there from therapy. One thing that I wanted to ask you about is, it seems like it's not just different behaviors we have in different situations based on kind of what sub-agents are activated, but our beliefs can change, too. Do you want to comment on that?

KAJ: Yeah. So if we go back to the example of you got bullied as a kid, and had social anxiety, to know as an adult, you might feel like on some intellectual level, you do realize that not every other person is out to get you. And so it seems like there's the part of you that originally developed in childhood, which is making a prediction of pieces, a social situation, and I will be bullied here and compares to other parts of you, which has experience of situations that didn't go so badly. And where people actually like you have, you might think something like people actually does like me. So you could kind of think of this as two different parts of you who are believing opposite things. And an obvious question here is like, why does that happen? Like, how come we have different parts of our mind that have opposite beliefs? And here's actually some research on it. But a big part of the issue is that it's really hard for the brain to know when are you facing essentially the same situation in a slightly different form? Or when are you facing an entirely different situation?

SPENCER: So these might actually be two different beliefs. If you think of them as being situation-dependent? Is that the idea?

KAJ: Yep, exactly. Like maybe you grew up in a family where all of your family members liked you. But then these children at school dislike you. And in a tough situation, it makes perfect sense that when you're around your family, you act under the assumption that people like you. And then when you're at school, you act under the assumption that people don't like you. And there's some research on like, when exactly does a brain treat a situation as something it has encountered before? And when does it treat this as something novel? It gets a little complicated and hard to interpret. But basically, it's just like, pretty commonsensical to more difference the context, the more likely this for the brain to lay down a new memory trace rather than update an existing one and then If the situation is somehow ambiguous like your brain isn't sure of is this more like a kind of a situation where people used to like me? Or is this more of a kind of a situation where people used to dislike me? Can you might have conflicting feelings, as both of those parts are getting somewhat activated, and both of which are trying to get you to behave in a particular way.

SPENCER: So this reminds me of things about conditional probability. It's like saying, like, probability, someone likes me, given that I'm in environment x versus probably we've given them environment y, right? And sort of like, it's that conditioning of what environment you're in, that might affect the answer. But what kind of bugs me about this explanation is, it seems like when we believe these things, it often feels universalized? Like, it doesn't feel like oh, people like me, because I'm in my home environment. It just feels like people like me. Why would that be?

KAJ: Yeah. So I think what's going on here is that in this situation, basically, the function of your belief is to get you to act in a particular way. Like on a subconscious level, there's probably some more detailed role model. So what happens is, you get into that situation, and then you end up having a belief of people like me, or people dislike me. And the important thing is that when you believe that people like me, will you act accordingly. And vice versa. And it's not actually necessary for your behavior to keep track of the fact consciously had a belief is only valid for this particular environment seems to maintain is that it's sort of your subconscious, which is triggering which belief is active at any given moment.

SPENCER: Yeah. So if the subconscious is already tracking what environment you're in, your conscious doesn't need to know that the information is conditional. The conscious just needs to think. This is true right now that that conditional information is going to be adjusted automatically in the background. If this is true, it would help explain why people's egos seem to be unstable while like, on one day someone might think, you know, I'm really good. And like, I can't really get things done. And then like, you know, something goes mildly wrong. And they're like, oh, man, you know, I suck and I can't get anything done. And it's like, well, you just got a tiny bit of new evidence, why is it that your view towards yourself seemed to shift so dramatically? And I mean, this seems to be especially true. And among people with really high levels of narcissism that their egos couldn't, like, oscillate wildly between, like, I'm the best person in the world and I'm worthless, but it doesn't feel like conditional just feels like oh, it's through in that moment.

KAJ: And there's also some, like some theories on the topic of personality disorders. One explanation that I've seen proposed for some of the behaviors here get in borderline personality disorder when you might have this sort of splitting thing. Have you thought of other people as either being all bad, or all good, and sort of nothing in between? Obviously, that's sort of the extreme case, you can also have like milder cases of borderline, which are not so serious, I want to add, but one proposed explanation for why that might happen is that if you grow up in an abusive family, and your parents are sort of, in some situations, they might be really horrible and abusive towards you. And then in other situations, they might be like, basically, okay. And if you were to sort of always keep in mind how terrible and abusive your parents could be, can you be basically freaking out all the time, because it would just be so scary to remember that they can do it at any time. But if you can sort of compartmentalizing the good memories, and the bad memories of what has happened with your parents can, okay, when they are abusing you going to do anything anyway. And you can just like freak out and post memories get assigned to its the evil parent side. But then when your parents or not being horrible at you, you can just think of them as great and good and perfect. And in those situations, you can sort of act normally. And you don't need to remember all the terrible things that they might be doing to you in other situations.

SPENCER: It's interesting because this suggests that there's a sort of information hidden from the conscious mind. And by hiding from the conscious mind, we get some kind of advantage there, right?

KAJ: Yeah, exactly. And also like sort of one implication of this model is that we sort of has the conscious beliefs and subconscious predictions. So you might have a subconscious prediction of if I do this thing then something good or bad will follow. And then you have a conscious belief, which is trying to get you to do the thing that your brain thinks is good or bad. So for example, Robin Ticic's textbook on therapy, called Unlocking the Emotional Brain, which had to take an example that, there was this guy called Richard, who found himself being really on self-confident when he was at work. He was like, pretty accomplished and competent. But whenever they had technical meetings, where he could express an opinion, he found himself having thoughts like, who am I to know what's right, or this could be wrong, and so on. And those thoughts were preventing him from saying his opinion. And then, he went to work out with a therapist and had the therapist tell him too much in [unclear]. Imagine you actually were confident and expressed your opinion, pay attention to what do you feel would happen, and retrieve the data. And then he started having this feeling that if he were to act confident, then his co-workers would all really dislike him and start hating him, and eventually turned out thought Richard's father, had to be in a sort of really arrogant, domineering person, which had made everyone hate Richard's father. And Richard had this belief that if he were to express any confidence at all, then, everyone would hate him, exactly the way people had hated his father. And in that case, there was this subconscious prediction, saying that if you're confident, people will hate you. And that manifested itself as a conscious belief of him being somehow incapable of actually knowing what's right, or not having the right to express his opinion or something like that.

SPENCER: I just want to zoom in on that for a second. Because I think what you're saying is that there was a subconscious belief that was different than the conscious belief, right? Like the subconscious belief had to do with like, Oh, if people perceive you as arrogant, then nobody's gonna like you. But that's not what he said to himself when he was telling himself that he shouldn't speak right.

KAJ: Yeah, exactly. The content of his conscious mind was different from the acceleration of why he was having those thoughts.

SPENCER: What and why would that be? I mean, you know, naively, when you think that the role there was actually being used, and his explanation of the role would be the same.

KAJ: Yeah, you would think that and there are a few different angles on this. Like one perspective would be thought he made a thing for the brain is to actually get the right behavior. And as long as you get the right behavior, it doesn't matter what you consciously think. So in that case, he was avoiding an expression of confidence and what was then his brain was trying to do. But then we also get into stuff about hidden motives. So one way of framing, this is that these sub-agents or trying to get us to act and behave in ways that will be positive for us somehow. And in particular, one important thing that determines whether things will be positive for us is how people around us react to us. And authentic people around us also have opinions about how we should be thinking, or what makes a good person or something like that.

SPENCER: Right. So like, if you are nice to your friends, but your friends discover that the only reason you're being nice to them is that you want to get something from them later. Like they're gonna be really unhappy with that behavior, right? So they have an opinion, not just in how you behave towards them, but how you conceptualize how you behave towards them. Right?

KAJ: Exactly. One example I've been thinking a lot recently is, there's this like attachment theory, which talks about, among other things, how very early interactions between a baby and the baby's caregivers end up influencing terror, later emotional landscape and terror, relationship behaviors and that kind of thing. And one of the possible forms of attachment is what's called preoccupied attachment. One version of how preoccupied attachment happens is if your caregiver is sort of inconsistent about reacting to your needs, but they might react particularly strongly to whenever you are expressing some sort of distress, fear or other negative emotions. And at the same time, they might be feeling somewhat uncomfortable with you expressing independence, like your stereotypical helicopter parent who's very worried about her child's safety, and doesn't really let her child explore. And the theory claims that one of the things that might happen in this case is that you're very early emotional programming. On a subconscious level, you'll notice that, okay, if I am dependent, and if I experienced a lot of negative emotions, then that will get my attention. And that will get my needs met. And like a sort of one extreme test, what tackle a sort of dependence schema, which basically has you believing that you are incompetent and incapable of taking care of yourself, and people with this schema, they are not faking it in any way, at least not in any conscious way. It's totally their experience that they are genuinely incompetent, and need someone else to take care of them. And I've had instances of this schema myself at some parts of my life. But at the same time, it's sort of an emotional strategy, which is adapted to the circumstances you were in when you were a child. And, obviously, if your emotional strategy is that in order to get your caregivers to help you and to give you attention, you need to be needy, and dependent on time. Then, for obvious reasons why your mind would be evolved to keep tapping this motivation hidden, because if it was just a conscious, intentional strategy, to remain incompetent, and dependent and just believe that you were dependent for the sake of getting attention. Well, not only would that be bad, from a from the sort of social perspective of other people, not reacting well to that pairs, also, the fact that everyone's brain has like lots of conflicting motivations. And okay, you might have some sub-agents that learned that being incompetent and dependent on others is useful, but most likely, those people will at the same time also have other motivations. And they will still also have the same basic needs of wanting to feel competent, that everyone does. And there's sort of a conflict that the sub-agents that actually want to feel competent and want to learn to do things should also be kept in the dark about his motive to not learn skills because if the sorts of competency things of ancients would realize it, they would try to stop this strategy from being the case.

SPENCER: I find that really interesting. It seems like the generalization of that is that there can be strategies to get what you want or need, that if you knew about them. It would actually undermine the strategy, or at least to some extent undermine it. So to execute the strategy, you sort of have to be at least partially in the dark about it.

KAJ: Yeah.

SPENCER: Another example might be supposed that you find that in order to get the attention you need, you need to act really distressed and anxious, right? And, you know, if you act really anxious, then your caregiver will swoop in and help you. But if you don't act anxious, they won't, right. So you start learning to act anxious like you're getting rewarded for this behavior. But if you were just trying to fake this strategy, you're like, okay, I need to act anxious right now, to get this reward. Well, first of all, your attempt to fake being anxious is probably less realistic than actually being anxious. So like, it won't work as well. And second, if your caregiver wherever to find out that you were like faking, being anxious, that would completely undermine the strategy, because then they would view it as sort of manipulation instead of oh, you're needed, right? So you're like, the strategy is much more effective if you actually get anxious than if you try to fake being anxious.

KAJ: Yeah, exactly.

SPENCER: But then, from your point of view, you just are anxious, right? You get anxious and then they swoop in? It's not like, are you trying to be anxious?

KAJ: Yeah. And one thing I also notice is that it's sort of like really difficult to know how to talk about these kinds of things. Because as we have been talking, on one level, it's not faking at all. People have this, like strong subconscious programming, that they need to know how to change it, nor the fact that they even have it in the first place. But at the same time, the key environment had people are in shaping their subconscious calculations, and it feels like it gets really tricky because if you view people as having these subconscious modes, they will just try to, on one level, optimize for things in a sort of manipulative way, you might start thinking of people as being way more manipulative and conscious of time to actually are. But on the other hand, if you don't at all pay attention to various subconscious incentives that people have, can you might actually also like be incentivizing that kind of behavior had people would actually be happier with overall if terrible is much less of it?

SPENCER: Hmm, that makes sense. So I want to put all this together what we've been talking about and discuss, what can you do with it, right.

KAJ: So I've used to stuff both for big things and relatively small things. An example of a big thing was when I first applied a transformative technique, I didn't really know any of this theory or think to scratch with terms back then. So up to this point, I'd had some pretty strong feelings of sort of a generalized shame, feelings of being a terrible person. And I'm still not entirely certain of all of the reasons for that. But at least one contributing factor to it was that I had memories of various experiences that were basically normal childhood behavior, like acting out, getting in fights with my parents, and so on. But in part, because I was an only child, and didn't have that many friends of my own age until I was older. I didn't realize it was normal for children to act in that way. And I thought there was something uniquely bad about me. So I had this underlying sense that I'm bad, and I shouldn't be ashamed. I'd had it just for a very long time. And I wouldn't always feel that way. But theorists got feeling that way, was always there in my mind. So I had this underlying sense of insecurity, a sense of unease from the fact that anything might cause those feelings to come back. And I did also have a memory of the nice things and being generally a good person. And I also kept doing things to prove to myself that I was a good person. But positive examples were associated with a different context. So there was kind of split off from the original memories and then didn't affect that underlying sense of shame. And I basically needed a constant stream of external validation in order to keep my attention on current anti-positive memories. Because the moment the validation was cut off, or if something happened to suggest that I might be bad, the temptation for memories might trigger again, and I'd feel horrible. And trust this process. I wasn't really consciously aware of what was happening on the kind of level to the time now describing. My actual experience was more of being generally afraid of my own mind, and feeling bad and discouraged a lot. I would get flashes of those memories that were making me feel bad, but because I would always push them away as soon as taken him up, I didn't recognize the role in the process and didn't get the chance to reconsolidate. And this went on till 2017 when I read a book called Transforming Yourself by Steve Andrea, who has a theory of how to work on these kinds of issues, and reading it helped me pay attention to what was actually happening in my mind. So a scrunch of memories came up, I could see how they were making me feel bad. And I could also see that while I hadn't behaved as well as by my toddler, like, it was actually just normal childhood behavior. And it didn't reflect who I was as an adult. And then working through those memories basically fixed it, in those persistent and pervasive feelings of shame, trust that disappeared. Today they are so thoroughly gone today, if I hadn't written down my experience at the time, I honestly wouldn't even remember ever having had him anymore. Another thing is authority down from the time was cut, it used to be that if I felt physically tired, then I would also frequently feel emotionally bad because the shame would come up. And while I can very vaguely recall that because I have it written down, it simply isn't the case for me anymore. So unlike before, it's now perfectly normal for me to feel good emotionally, while feeling tired physically. And that's just completely different from how it used to be. A more recent example is that I had, just like creating books for entertainment. And I would have noticed his thing that, I might start reading some book just for fun. And at some point, I would start having this feeling that it sort of became an obligation for me that I had this feeling that I should finish reading this book, and then, I should record it somewhere that I have actually read it and turn it off and started feeling like okay, now this is no longer fun, that I'm no longer actually just reading this book for entertainment. I'm reading it because I have some sense of just needing to read all the books, and has actually had the effect that I wasn't reading very much. Because often, reading just turned into a sense of obligation. So tonight I had that feeling of reading books feels like work. And I sort of just held it in my mind and asked myself, “Why does it feel like work?” And what it came down to was basically attached by having this feeling that I needed to get sort of tangible accomplishments. If I was doing something it was like there was a field to actually make a record that I had done something useful with my time. And there was the feeling that I should also be doing useful things with my time. And I should be able to somehow show it to other people that I have been doing something useful, even if had useful thing was just like creating books. And when I asked myself, “Why start?” I had so many taste memories of care was an instance where I felt like someone whose opinion I cared about had felt like I wasn't doing anything, and felt like I was basically just wasting my time and not doing anything. While in fact, my own experience was that I was actually like, doing quite a lot of things and feeling stressed out and really needed a vacation. I needed the rest somewhat. But then I had nothing that I could use to prove to this other person. I actually have been doing quite a lot. And after that experience, here, develop tests need to actually like keep a record of the things that I'm doing so that I can if need to be document to other people what I have been doing, and after I managed to dig up that original motivation I did notice that, okay, this is a sensible thing to be concerned about. In this specific situation in which I was in where I really needed that risk and really needed to justify paths. I had actually been doing things, but in my particular situation, I don't actually need to justify it to anyone that if I want to take a break or rest or whatever, can after I worked with that issue, basically the need to have those objective accomplishments completely disappeared. And after that, I've just felt fine reading books or doing whatever without that kind of a need that would turn into obligations that I need to finish, for the sake of having accomplishments.


SPENCER: Do you want to go back to that example of the person who would struggle to express opinions at work and talk about how did they get through that struggle? And then we can bring that into a broader discussion of, how can you use these ideas in your own life?

KAJ: Yeah, so this Richard, who we were talking about. So as we mentioned, his therapist guided his attention in such a way that he became aware of the fact that he had this underlying belief of if I expressed confidence then I will be hated. And what happened was that he had become aware of having this belief and his therapist instructed him to just sort of feel that belief and just generally be aware of it for a while. And he did. And like, often he reflected on his feeling, and it felt through to him. If he were to express confidence, then people would hate him. And as he was thinking about this, there was one day when he was in one of those meetings, and a topic came up and he had this thought of something he could say, but then he had his doubts coming up again, and he remained quiet. And just then his coworker happened to say, basically exactly the thing Richard had been thinking about saying, and his coworker also expressed it in a pretty confident and self-assured way. And then Richard noticed that all the people in this room seems to be reacting very positively, to taste, coworkers commented, and he realized that there was a big sort of contradiction. He had been believing if I say this thing confidently, everyone will hate me. And then he sees his co-worker saying exactly the same thing in a confident way and getting a very positive reaction. And a sort of principle of how can you actually change these kinds of beliefs is that there's this principle called memory reconsolidation in neuroscience and the theory is basically, what if your brain predicts that a particular thing will happen? And then it notices that the prediction is somehow incorrect, like exactly here, predicting that, saying one thing we'll get one reaction and then it gets another reaction then if it gets contradicting information when the previous prediction has been active, that had may change the underlying prediction and sort of cause it to update, and more generally, if your brain sort of notices that it is, at the same time, believing one thing, and also believing another thing, at the birth of this, things feel really true to you. But at the same time you realize that they're logically incompatible, you can't believe in both of these things at the same time, then your brain will actually change its beliefs and integrate these two predictions with each other. So in Richard's case, he had the memory of how people had reacted to his father. But he came to realize that people had been reacting to his father wasn't actually about his father's confidence, but about his father's arrogance, and his therapist guided him to do an exercise that would sort of strengthening this effect. So his therapist basically just told Richard to recall his experience of how his co-workers had reacted to this other co-worker, and hold in memory key experience, while also mentally contrasting it to the sorts of certainty of what would happen if he were to say the same thing. And then when he just contracted first to experiences, gradually, his sort of self doubts just completely dissolved, as his subconscious made the update and stopped feeling the need to maintain those, as his mind no longer found it necessary to avoid being confident.

SPENCER: So is the idea that you take this belief on the one hand, like, oh, if I speak competently, people aren't gonna like me, or they're gonna think I'm, you know, asshole. And then you take this experience, on the other hand, and you kind of bring them into mind at the same moment, and kind of let them sit there, you're not trying to kind of squash the original belief, you're just letting it sit at the same time with this memory, wherein this case, the other person said something competently and got rewarded for it. And then this sort of integration somehow just happens underneath?

KAJ: Yeah, basically. And one unintuitive thing about that is that when you're doing it, it's important not to try to sort of actively disprove whatever belief or memory you're trying to change, and not try to argue against it. Because what the process depends on is that you need to experience both of them as feeling through, and then a part of your brain realizes that wait, I'm strongly believing in doing things that cannot be true, that cannot actually both be true.

SPENCER: Since releasing, you say that, if we take this strategy, it looks a lot like something used in many different therapeutic modalities. But in the different therapeutic modalities, it's used in a sort of different ways with different details. So in cognitive therapy, for example, it's very common to have people take some kind of belief that seems to be causing problems for them. And then try to dispute it, basically try to notice ways that you might be engaging in general overgeneralization or black and white thinking, once you kind of find the kind of error in that form of thinking, then maybe find evidence against the belief being true. And a lot of people do seem to find this really helpful. But you're saying specifically not to try to dispute the belief. So I'm wondering, that seems to be the intention there with the cognitive therapy technique. And I'm wondering if you just believe that kind of therapy technique is not very effective, or you think that there's something else going on there?

KAJ: Yeah. So the book that I'm drawing this specific model from makes his distinction between what it calls transformative and counteractive therapy. And in transformative therapy, it's basically taking I described. An example of counteractive therapy would be what you mentioned, like noticing the flaws in your thinking and sort of trying to argue against them. And in that case, you're sort of setting up a new subroutine or new sub-agent in your mind, whose job might be to notice the irrational behavior and try to counteract it. And the argument in the book, which I tend to virtually agree with, is that if you can do it, it's often better to sort of go with transformative change. Because if you can do the transformative change then that would just like completely remove the source of the problem. And can you basically need to deal with it again, whereas in counteractive change, you are basically creating a new conflict in your mind? So whether that counteractive strategy works depends on whether the conditions happen to be right for you to actually always succeed in making the counterargument inside your mind. But of course, I'm also not saying the authors of that book are also not saying that you shouldn't ever use counteractive approaches.

SPENCER: I'm not sure the proponents of character approaches would agree with that, though, because I think that they would argue that over time. It actually starts to shift that belief or shift your sense that that original theory was true, you know, as you kind of counteract it, maybe the first time you counteract it's sort of, you know, you're just kind of pushing against it, but you know, maybe the 20th time now you just don't have the sense that that thing is true.

KAJ: Right. I think it probably depends a lot on to exact circumstances and details or issues. Or I could imagine that if you are doing something counteractive then you might basically end up accidentally crossing over to the transformative version. Like you have this maybe black and white thought. And then you get to a habitual thought of why that isn't right. And maybe it spontaneously happens that actually, they happen to be active at the same time into kind of a way that does transformation.

SPENCER: Very well, thinking of counter-evidence might easily bring to mind that other memories that counteractive, right. And that's starting to look very transformational.

KAJ: Yeah, although it doesn't necessarily always work that way. One specific example that I'm thinking of is, is when you use CBT, to treat anxiety or phobias. So let's say you've been stung by a bee, and now you have a phobia for bees. And something that may happen is that CBT does help you with it. If you do something like exposure therapy, but then later, you have a relapse and phobia comes up again. And as far as I understand it, the explanation for why that happens is similar to the thing we were discussing before, where your brain might believe different things in different contexts. So when you originally got stung by a bee, maybe in your backyard, your brain started believing that bees are always dangerous. And then you did exposure therapy in a clinical context. But when what ended up happening was that your brain created the belief that bees in a clinical context are not dangerous. And turn, you end up having an ambiguous belief about how these bees are in general, at the exact details of the context you're in, and determining whether or not you do have anxiety. Another thing that I have heard from people more anecdotally is that many folks tend to react negatively to CBT. Because it tends to feel pretty invalidating. For instance, beliefs that cause negative emotions might be labeled as dysfunctional beliefs, which kind of implies that there's nothing good about those beliefs. And if we assume that the brain created those beliefs, for some reason. Then any approach that suggests simply rejecting them without considering what their benefit might be, might be expected to be less effective than optimal. Personally, I often have experiences where, for instance, I'm using the example of social anxiety, I might be having an experience where I'm feeling social anxiety, and then there are some other parts of me, which feels that actually, there's no reason to feel this way. And I have all of these good reasons for why I should be confident and not stressed out. And then I might be sort of really hard thinking about the reasons why I should be more confident, or others might be thinking really hard about tasks. But the overall effect tends to be more of like, sort of pushing down the social anxiety and sort of like trying to overpower it and sort of use the content of why I should not be anxious to just like, push out the anxiety out of my consciousness, rather than actually integrating it, too. And I feel like that pretty easily happens if I'm not careful.

SPENCER: So I want to contrast three different ways of dealing with a situation like this. So let's, let's use a new situation, just to clarify further. Imagine that you find when you make a mistake, you often say something like, oh, I always screw everything up, right? And it makes you feel really bad. And you kind of belief on some level, you always screw everything up, right? So there's the transformational strategy, which tells me if I'm getting this correct. But basically, what you would do is maybe you would find some scenario from your past where you actually did things just right, you didn't screw anything up. And then maybe you would simultaneously hold in mind this idea that feels true to you that you always screw everything up, along with this memory of doing everything just right and not screw anything up and kind of just let them both sit there until some kind of integration seems to happen. Is that right? For the transformational.

KAJ: So like two basic things you described sounds right, by mostly thinking about in this specific case of feeling like you screwed up. If you are bringing to mind past instances of not having screwed up that probably doesn't change your assessment of whether this particular thing was a screw up.

SPENCER: I'm talking about someone who just sort of has a general attitude that they screw everything up.

KAJ: So the thought they are having is that “I always screw everything up” and then they are bringing in past experiences of times when they did not screw things up.

SPENCER: Right and they're kind of letting them both sit in their mind simultaneously. Those are things both feel true to them, right?

KAJ: Yeah, for the specific cases of feeling like they always screw up that feels like the kind of a case where the conscious content isn't necessarily the actual thing that's going on. Likely the reason why they are having the feeling had they always screw everything up is that it feels for some particular reason it feels dangerous to do something that we're doing. And the feeling of always screwing everything up. It's probably something more specific about what's going to happen if they screw up and why that is dangerous.

SPENCER: Right? That makes all sense. So you might want to ask them something like, what would be bad about not believing you're gonna screw everything up? Or what is leaving you always screw everything up protecting you from or something like that? How would you phrase it to help them kind of pull that out?

KAJ: Yeah. Sounds like good questions.

SPENCER: So then let's say that let's say you kind of raise that what's good about believing you, you screw everything up or woods protecting you from? And then they said, “Well if I didn't believe that, I would probably take on big risks that I couldn't afford” or something like that, you know, I don't know what a real example would be. But then so then it's that belief that you'd actually want to do the memory integration thing with is that right? Like that underlying kind of motivation?

KAJ: Yeah, exactly.

SPENCER: Got it. Okay, cool. That makes a lot of sense. So we've got the transformation approach. And I want to contrast that with two other approaches. The second is this counteractive approach like you have the cognitive therapy – cognitive therapy being kind of a subset of cognitive-behavioral therapy, where you might look for evidence against the belief or challenge the belief or look for what kind of biases might be in your thinking, right. But then there's a sort of the third approach, which you might see in ACT, which is Acceptance and Commitment Therapy, or in a kind of mindfulness approach, where you're just kind of observing these thoughts from the outside, and you're saying, “Well, I know that I had the thought that I would screw everything up. But that's just a thought that's just something running through my mind. I don't have to believe it. I don't have to assume it's true. I don't have to feel it from the inside. I can just watch that thought and let it go.” Any comments on that kind of third approach?

KAJ: Yeah. So the, like mindfulness approach is, it's good if it's enough. And it can also be like very valuable in helping you notice that your beliefs or actual beliefs, and are not facts about the world human sort of limitation. I would see what it is that, for instance, in my own experience, it might be that I can be mindful about a particular belief and watch it from the outside, and then find myself basically acting in accordance with it anyway.

SPENCER: Interesting that somehow it's still affecting you. Yeah, I suppose also, you can make the argument that like, you haven't done anything, necessarily, it's gonna make it not occur again, five minutes later, right?

KAJ: Yeah.

SPENCER: The hope of the other two techniques, at least is that over time that diminishes that sort of belief, or that sort of motivation.

KAJ: I think one of two things that may often sort of make the mindfulness thing actually change your behavior is that if you can sort of getting to distance from your beliefs, and actually watch some of your own thinking, then, doing that may actually help you notice some of your some of the tariffs, subconscious assumptions you have going on, like some of the motivation, sort of leaking into your consciousness and being accessible if you actually pay attention to it.

SPENCER: Right, subtle signs that you may easily miss normally,

KAJ: Yeah. And then if that makes some of your, like, incorrect predictions or assumptions, accessible to you then Just noticing that may trigger that kind of like, the transformative process where you notice that actually, taste doesn't make any sense. Or after it feels like it makes sense, but then you realize that it doesn't.

SPENCER: It reminds me of one of my favorite little techniques, which I call the inner why technique, which is very simple. It's just when you notice your motion suddenly shift, you just immediately ask why. And I find that if you can get that y question in, like, within a few seconds of your emotion shifting, it's often really obvious why it shifted. Because like only one thing has happened in the last few seconds. But if you wait five minutes, it can be really hard to figure it out. Because you just can't pinpoint this precise thing that causes your emotion to shift.

KAJ: Yeah. And I would also say about like things like mindfulness and introspection in general, is that a big part of a kind of transformative change that's possible in therapy and otherwise is that like a big component of it is just having different scales for digging up the subconscious assumptions and making them explicit enough that you can actually do the transformation on time. Because as long as you have no idea of why you are acting in this way, can you also can't bring in any contradictive knowledge because you don't know what you even should be contradicting? Like if you just have some vague anxiety in some particular situation then the first step is to try to figure out why exactly are having that anxiety in the first place.

SPENCER: Right. So maybe it's like step one, notice that you're having behavior or belief that seems to be not leading to your goals or causing problems for you. Step two, try to understand what you're trying to achieve with that, or why you're running that strategy. Is that right?

KAJ: Yeah.

SPENCER: Like what you're trying to avoid?

KAJ: Yeah.

SPENCER: And then step three, try to find examples that seem to contradict that. And then you can kind of hold those examples in mind, and just kind of just feel both of those things being true simultaneously. For example, a memory that seems to contradict that, and then let those integrate. Is that accurate?

KAJ: Yeah, basically, with step two of figuring out what's going on is that often if you try to intellectually figure out why exactly am I doing this, it's a very authentic case. That is because the sort of conscious part of the issue isn't actually the only reason then if you are just trying to intellectually think about why might I be acting the way I'm acting, you very likely don't actually have access to ulterior motivations. So a big part of why I have been interested in these things is that, some of these therapists, such as internal family systems, or focusing oriented psychotherapy, where you can basically teach you techniques for sort of getting more access into the actual, underlying motivations. And they are like are somewhat mindful, like in that they tend to involve taking some experience, and holding it in your attention, and sort of asking questions about that experience or feeling that you might be having, but not trying to intellectually analyze them, but sort of just poke at the feeling and investigate what kinds of thoughts and associations are popping up when you hold that feeling in consciousness. And then gradually, putting more details emerge to your conscious mind.

SPENCER: Right. And I think one of the insights from the kind of focusing idea that seems really useful is that sometimes you could have a thought pop into your head, like an explanation for your behavior or something like that. But it doesn't exactly ring true, if you like, think about it. Whereas other times, you're like, oh, no, that seems exactly right. Or another example would be like, let's say you're trying to explain how you feel. In one case, you might give an explanation for how you feel, but then it doesn't quite ring true. And then you need to adjust it and like, ah, that's it. Now I hit it. So it's kind of that loop of like, you kind of spit something out. And then you kind of do this evaluation say, is that really right? Ah, yes, that's really it, or oh, that's not quite right. Let me tweak it. So you're kind of using this process to hopefully make it more likely that the thing is real things are real motivation, or really the right explanation or it captures what you're really feeling. Because obviously, there's a danger with these techniques of coming up with like, sort of BS explanations.

KAJ: Yeah, exactly. It's sort of hard to balance, especially at first, when you don't necessarily quite know what you're doing in that you might get sort of what you might come up with some intellectual explanation that feels really convincing, and feel like, oh, this must be it. But it's not actually the sort of intuitive answers that you would get if you investigated it at sort of intuitive sense of his feels right, or this feels slightly off, which is a different feeling of feeling right from this sort of the feeling. Right. And it takes some practice to actually start, be able to tell us apart.

SPENCER: Before we finish up, I just want to ask one more question, which is, what makes you feel that these techniques are more effective than a sort of transformational approach?

KAJ: What makes me feel they are more effective than the counteractive methods. Is that what you mean?

SPENCER: Yeah, counteractive, or let's say there are alternative approaches.

KAJ: So my strongest reason for me personally, is a personal experience, and also the personal experience of other people I know who have been getting these kinds of methods to work is that I had various kinds of emotional issues or issues where I don't quite get myself to do what I won't or can't quite get myself to feel whatever I want to feel, and whatever. And for quite a lot of different things by applying these construction methods, I've actually managed to create kind of a change that has felt like it has been really significant and basically has had a permanent effect. So that I've never needed to deal with that particular issue again.

SPENCER: Does that come about from doing it just one time?

KAJ: So for some issues, just doing it one time has been enough, can there have also been sort of deeper issues in like someone's case issues seem to come in layers. For instance, if you, as a very young child, have some experience, and pick up some particular belief about things when you are, say two years old then probably, this belief is going to pop up in your life in quite many different areas. And you can sometimes get access to some version of a belief that you developed later, like, maybe you are 10, or 16, and sort of gradually work your way deeper and deeper versions of the issue. So in those cases, it might be that you need to do quite a bit more work and transformation. Then if it was some relatively superficial issue, that you just developed from having one experience at some time.

SPENCER: Awesome. Kai, thanks so much for coming on. This was really fun.

KAJ: Yep. Thank you for having me.





Click here to return to the list of all episodes.


Sign up to receive one helpful idea and one brand-new podcast episode each week!

Contact Us

We'd love to hear from you! To give us your feedback on the podcast, or to tell us about how the ideas from the podcast have impacted you, send us an email at:

Or connect with us on social media: