Metacognition

A transcript of Episode 303 of UX Podcast. James Royal-Lawson and Per Axbom are joined by Steve Fleming to discuss metacognition. Thinking about thinking and thinking about other minds.
.

This transcript has been machine generated and checked by Stephanie Janeczek.

Transcript

James Royal-Lawson
Ambition and power is a professional education programme is directed towards design leaders and UX professionals interested in upping their game through a continuous learning journey that engages you in small chunks every week during your membership. Find out more by visiting UX podcast.com/empower.

Computer voice
UX podcast episode 303.

[Music]

Per Axbom
Hello, I’m Per Axbom.

James Royal-Lawson
And I’m James Royal-Lawson.

Per Axbom
This is UX podcast. We’re in Stockholm, Sweden and you’re listening to us all over the world from Macau to Bulgaria. Steve Fleming is a professor of cognitive neuroscience and runs a neuroscience lab dedicated to the study of self awareness at University College London. His team, the metacognition group, is one of several working within the Welcome Centre for Human neuro imaging, located in an elegant townhouse in Queens Square in London.

James Royal-Lawson
This whole thing with saying neuro is quite hard, isn’t it? Anyway, the basement of their building, and I love the fact it’s a basement, that houses large machines for brain imaging and each group in the centre uses this technology to study how different aspects of the mind and brain work, how we see, hear, remember, speak, make decisions, and so on. His research focuses on the science of self awareness and last year, Steve published the book, Know Thyself, the new science of self awareness.

Per Axbom
Set the stage, because this is a design podcast and as designers, we talk a lot about behavioural science, we talk about cognitive biases, and some designers love applying the ideas of nudging while others like in this sort of to the removal of autonomy, but can you help us out with an explanation of what metacognition is and how it fits into this broader idea of how we as humans make decisions?

Steve Fleming
Sure. So metacognition literally means cognition, about cognition, or thinking about thinking. So it’s the capacity we have this weird and wonderful capacity, we have as humans to reflect on a know something about the operation of our own minds. So an example of metacognition would be realising that you might not know enough when you’re revising for an upcoming exam, for instance. So that’s a metacognitive assessment of how much you know. And that has consequences for how we behave. So if we think we don’t know enough, we might then continue to study rather than put down the books. And we think metacognition operates across many different domains or areas of the mind. So we can think about not only our memories, but also how well we’re performing. In different areas of life, we can think about whether we’ve made a good or bad decision, we can think about whether our physical or cognitive capacities might be changing and older age or in disorders, brain disorders, for instance. And in relation to decision making, so an autonomy, we think that it plays a big role in what it means to be a self aware agentive human being because, in a sense, be able to endorse reflectively endorse the courses of action that we have chosen, and say not only have I made this choice, but I want to be doing what I’m doing right now, that we think is foundational to autonomy. It’s not just me that said this, quite a few philosophers have made this point that reflectively, endorsing our choices might be central to what in folk psychology we refer to as free will.

James Royal-Lawson
So this is this is a personal performance evaluation that we we do constantly, I guess, with ourselves.

Steve Fleming
Exactly, exactly. So we think that it exists on a local timescale. So you might kind of think to yourself, Oh, I’ve just forgotten to do something that I should have done this morning, for instance, that’s a realisation, self realisation, that we failed it on some local level, but then also you can have metacognitive beliefs that extend over a longer timescale so we might think to ourselves, hang on is, you know, is my vision failing? Do I need a new pair of glasses, that’s a metacognitive assessment about a physical capacity, so there’s various timescales of metacognition, but they all share this element of being a self assessment.

James Royal-Lawson
So I guess we’re not aware of some of them then because like, I mean, I’ve just got new glasses because for the first time in a long time that my kind of base eyesight has kind of changed and, and that’s taken me a long time to actually accept or understand, I guess the concepts of it, so I suppose there’s this stuff that’s happening constantly, like I reach out for something. Does that mean that if I kind of miss reaching out, then I might not reflect on it at first, but I have kind of some where deep down reflected on it.

Steve Fleming
Yeah so I think the the distinction you’re drawing there is only between what we refer to as implicit or unconscious monitoring. And some people there’s debate over terminology here some psychologists refer to that as unconscious metacognition. Other people say, no, that’s not what we mean by metacognition, it’s just something that’s still monitoring what we’re doing, but at a lower unconscious level, I think that’s just a semantic debate, really but essentially, the idea which I outlined in the book is that we have self monitoring processes going on at multiple levels within the system. So you, your actions are constantly being monitored for flow and accuracy so when you, an example I like is when you step on a movie escalator, your body automatically corrects your posture for that movement and then if you do that, that correction can be exposed as being very automatic because if you then step on a stationary escalator, your visual system still perceives the escalator and your body starts preparing to make that correction and yet, it’s not moving anymore so then you get slightly knocked off balance.

James Royal-Lawson
Which is that is actually a really fascinating thing, because, you know, you can see, I mean, it’s one of those frustrating things it’s like, come on body. I mean, I can see it’s not moving, but you’re you’ve so deeply trained to use that tool as an escalator.

Steve Fleming
Exactly, exactly. So we’ve built up this association over the years with that postural correction of stepping on an escalator and the you can see the learning process there. You know, when I take my three year old son on an escalator, he hasn’t learned that yet so he stumbles when it’s moving, whereas I would stumble when it’s stationary. So there’s that kind of exposure of the learning of self monitoring at a lower level. And that’s what I refer to as essentially autopilot for the mind, we have this self monitoring level, going on below the surface, it keeps our mental planes flying straight and level and then at certain points, that self monitoring process becomes conscious when we start thinking to myself, hang on, I’ve made an error there, or I’ve gone off track.

Per Axbom
Oh, you actually need glasses. It’s not it’s not that they changed the type and the magazine. And I think in the book, you talk about this as making and updating predictions. And you This is related to dopamine, which I also found fascinating because there’s this myth about dopamine that I’m hoping you’ll help us kill now.

Steve Fleming
Yes, so dopamine has often got popular press for being the reward chemical, but that that has some truth to it, but it’s it’s oversimplifying what dopamine is doing in the brain, I should say, first of all, that we’re still learning a lot about what dopamine and other neurotransmitters are doing in the brain so the story is not finished yet, but what we think dopamine is doing is signalling some kind of error on our predictions of what the world is going to be like. And so an example of this would be, if you go into a new coffee shop, you don’t know what quality of coffee, they’re going to serve you. And so your expectations are relatively low, they serve you really good coffee, you would get a positive, what’s called a prediction error signal. So that is a reward.

In a sense, it feels good to get that good coffee, but it’s and it’s better than you expected it was going to be so then you get a positive dopamine spike. But then over time, you learn that that’s going to be the type of coffee you get from that shop, and there’s no error in your expectations anymore. So it’s still rewarding to go and get that cup of coffee, but your dopamine won’t be signalling any change in your predictions anymore. And this was shown in really amazing experiments by a neuroscientist called Wolfram Schultz, who’s now in Cambridge, who recorded from dopamine neurons and showed that they showed this prediction error like signal. And what’s really even more fascinating is that that prediction error like signal is exactly what computer scientists have been saying for many years is needed to train reinforcement learning systems. And now these kinds of RL reinforcement learning systems are used all the time in artificial intelligence, programmes that Learn To Play Chess, and go and so on. So there’s a real consilience there between AI research and neuroscience research.

James Royal-Lawson
I think this is absolutely fascinating to me, because from a design perspective, you know, for years now, we’ve been there’s been this whole talk of delight, and kind of like, you know, making experiences that delight the users. And and dopamine has come up with those kind of discussions. And what you’re saying to me there, though, is because it’s a reward for discovering an error, then you’re getting dopamine, because something is kind of not what you expected, and that we’re painting as a delightful thing. Oh, God. Exactly, you said the good coffee. But at the same time, you could get just a bigger hit from a poor experience that is as surprising.

Steve Fleming
So it’s still signed. So if you get a, you can get a negative prediction error. So if I thought it was going to be really good, but now I get poor coffee, my dopamine neurons will then decrease in their response. So it’s still a sign thing. So positive is good and negative is bad. But what I think exactly as you say, like I think it has implications for how we think about the subjective experience of joy and, and, and, and well being. And there’s some work that a colleague of mine who used to be at UCL, he’s now at Yale, Rob Rutledge has been doing to look at people’s moment to moment, subjective ratings of how happy they feel, in an experiment, where they’re receiving different small rewards, where he’s tightly controlling, in a reinforcement learning framework, the expectations they have in that task. And what he’s found is that their happiness in experiment is not linked to the long run wealth that they’re accumulating. So they accumulate accumulate more money over time. Instead, it’s very tightly linked to this prediction error signal. So essentially, if you have a certain expectation, and that you get more than you expect, then you get a bit more happy and, and vice versa. And so I think this, this matches with everyday experience that we might accumulate more worldly goods, but we rapidly then reset our expectations. After a while they don’t make us happy anymore. And we have a certain baseline level expectation, then we need something that violates that in a positive way to get that kind of spike again.

Per Axbom
So what would be the benefits of improving our I don’t know metacognition skills, per se, would that actually improve our well being or our relationships to other people?

Steve Fleming
It’s a good question. So what link this might have to well being I don’t think has been fully unpacked yet. I think that there are certainly functional benefits for daily life in terms of how metacognition helps us make better decisions, because it helps us to realise we might have made an incorrect, call and revise that revisit that it helps us to temper the rigidity of our beliefs by realising that we might not have all the answers, and we should listen to other people’s opinions. It helps us to work together with others because we can collaborate better, if we are appropriately sensitive to when we might not know the answer to a problem we’re working on. And we seek the advice of others, for instance, so so it has all these functional benefits that are reasonably well worked out.

And we can see in a lot of the work we do in the lab is in collaboration with clinical colleagues working with patients with psychiatric and neurological disorders, who often have profound deficits in metacognition, and self awareness. And they’re not always very detectable in the clinic, but they have a profound impact on their daily on people’s daily lives because, say, in the case of dementia, for instance, you might, a patient with dementia may be unaware that they’re losing their memory, and that has real impact on their decisions to continue to drive or go out to the shops and so on. So the extent to which you can compensate for, say, changes or deficits in cognition and performance is very dependent on metacognition, so that clearly has a functional benefit. But then there’s also potentially another side of this coin, which is that if I’m very aware of the potential for getting things wrong, then in theory that could have a negative consequence, maybe I engage in more regret or rumination on whether things are going well in my life and so on. And I think that connection hasn’t been fully explored, it might be there. But it hasn’t been as well unpacked in the sciences as the benefits of metacognition for for function for daily life function.

James Royal-Lawson
So that’s, that’s the whole thing where you’re observing the thing. You are the thing being observed. So kind of, you’re impacting yourself with your own observations of how you’re performing how you are.

Steve Fleming
Yeah, so I mean, we do know that in is well established in depression and some forms of anxiety that people engage in a lot of rumination. So the way I think rumination is connected with metacognition, it relies on similar psychological machinery. But the way I see those being different is that you can think of rumination as being the quantity of metacognition you engage in to how much do I tend to think about myself and my own mind, whereas the functional benefit of metacognition is not necessarily doing it all the time, it doesn’t mean you have to sit around like Rodin’s Thinker and just introspecting all the time. It means that when you need to, you kind of take that reflexive stance, on how things are unfolding, in at work at home, and so on, and then change course, accordingly.

Per Axbom
Because I know you read about how the gains of metacognition aren’t really nothing compared to the downsides of losing it. And that’s what you sort of touched upon. Now, you’ve also talked a bit about how it helps you understand other people, which also creates this platform for a better, more fair society, perhaps. And at the same time, what you’re also saying a bit now, if you’re, if you’re ruminating, that can also lead to losses. And at the same time, as we we’re actually criticising society today and saying that we don’t have enough time for reflection, because it’s this appears to be a luxury. So there’s a balance going on here. That’s really difficult.

Steve Fleming
Yes, no, there is a balance. I mean, just to pick up on one thing you said there about understanding others. We think that that’s another very important component to the way that metacognition provides a platform for social interactions, because there’s growing evidence that the ability to reflect on and think about ourselves might share neural and cognitive machinery with the capacity to think about other minds. So what psychologists refer to his theory of mind. And it’s clear that when you want to engage in a nuanced to person interaction, when you’re collaborating with someone, for instance, you want to be appropriately sensitive to what you bring to the table. So that’s metacognition, but you also want to be aware of what the other person’s skills and failings might be. So you can navigate that interaction. Well, I think this happens all the time in a in a workplace context. So the effective people are the ones who can kind of go with their own view, when they think it’s the right thing to do or defer to others, when they think that someone else has got a better handle on it.

James Royal-Lawson
This is something I really wanted to bring up. Now, the use of theory of mind, or I think you refer to is mind reading, as well. And from, you know, from a designer perspective, I mean, I was I was kind of struck by that whole idea of mind reading and reflected on how, how much mind reading, I have to do as designer because it feels like that it’s not just kind of collaborating in a team, that when you’re designing services, digital services, and websites, and so on, then I’m going I’m having to do a level of mind reading, which is well outside of my my mind, and you know, even outside of the realm of people I can come in direct contact with.

Steve Fleming
Yeah, so I feel like in, especially in creative work, where we’re trying to think through how someone else will see what we’re creating. So their work of fiction, or even even the extent to which I find this happens when I teach that I need to think really hard about what my student’s view of the world is, and try my best to get out of what I know. And I find it incredibly difficult. And I think a lot of people do that, essentially, you need to lose your knowledge, you need to build a new model of how the world is being seen from someone else’s perspective. And the facility to do that allows you to then, I guess, unlock new experience in other people because you can then create that new experience the new way of seeing the world, uncontaminated by your own pre existing perspective. And I think, you know, the Great, the great novelists clearly do this all the time, because they they’re able to actually construct a world that is brand new from the readers perspective that is relatively uncontaminated by the writers own is the other thing.

James Royal-Lawson
I think I’d say the thing there about the feedback as well because it connecting mind reading and then the errors or updating our models of of how well things worked. And when you’re doing that was like a work of fiction, you create an entire book. So you’re doing that entire book and putting it out there. And it’s like the feedback loop. It must be kind of same thing I suppose with some of the designs we do. We try and do research and testing and so on, but you you know, is it really can you really create that loop at such a pace that that means you are doing better and better stuff, or is it we just failed, or we just kind of doomed to fail? But at regular intervals?

Steve Fleming
Yeah, I mean, I think this is a fascinating issue on how the human mind is able to cope, essentially without much feedback. And this is where it often separates out from current AI systems. There’s lots of progress going on this in terms of what’s known as credit assignment in, in artificial intelligence, like how does an AI system play an entire game of chess, and then only at the very end, do they find out, you know, whether they want to last? Now, the way to do that is essentially to backup all the values of the individual moves and start attaching value to intermediate steps along the way. And I think metacognition is doing something quite similar, in cases, in open ended cases, where we’re not getting feedback until the end of a whole sequence of actions. But still, there seems to be these internal prediction errors. So I might set off writing a paper and then get a few paragraphs in and realise “hang on this is this is rubbish”, but that’s my self assessment now was told me it’s rubbish. But I’m kind of assessing that. And so I might scrap that and go and do something else. And there’s been some lovely work showing that the same types of dopaminergic prediction errors that we talked about earlier, that are driven by external rewards, like drinking the coffee, or getting the drop of juice in the experiments are also being driven by internal errors that we make.

So this has been studied in songbirds. So when a bird is producing a song, if it makes a mistake, then it gets this little negative dopamine spike. So and then the way that’s that’s experimentally studied that the the experiments is kind of amazing setup, they can play mistakes to the bird. So the bird thinks like, Hang on, I’ve just screwed up there, and you get a little negative dopamine drop. So it seems like there’s a common system that is trying to predict whether we’re going to get things, whether we’re going to get rewards. And those rewards include internal feedback that we’re giving ourselves. And I think that’s exactly what you need. When you’re embarking on something. Over there’s going to take a very long time, like writing a book, where you’re not going to get feedback from the outside world for a long time, but you’ve got to constantly be thinking through, “Hang on, is this on the right track? Am I gonna get this positive response to my reader for that?”

James Royal-Lawson
Yeah, so yeah, so we’re running predictions constantly. So even though the longer game, maybe we can’t get an end result, we’re looping, mind bogglingly large number of times all the time, and getting feedback and correcting.

Steve Fleming
Exactly, yeah. And in the case of something that requires theory of mind, so that that’s then layered on top of that. So yeah, you know, in the case of writing out a lecture or writing a book, you’re not only creating a prediction about whether you think this is good, you’re trying to predict whether someone else is going to think this is good.

Per Axbom
I do want us to touch a bit on AI. Because I know, you allude to it throughout the book actually about how AI is unconscious and inexplicable. And you have the suggestion of working with, for example, self driving cars to actually make the AI have confidence levels. So it can actually doubt itself, which is something is really intriguing when you think about a system that has no self awareness.

Steve Fleming
Yeah, so confidence is something that in AI and robotics research is being pursued with with vigour at the moment, including in the realm of self driving cars. And the aim there is not necessarily to replicate human metacognition, and to somehow create a self aware robot, instead, it’s to give it this minimal implicit form of confidence estimation, and to do that in a more graded way. And so it is a difficult problem because especially in AI systems where you’re training on some dataset, and then you see a new problem, say a self driving car goes into a new neighbourhood, it’s never driven in before, then it’s quite difficult to make well calibrated predictions about what you’re seeing. So it might rather than saying, you know, I’m not sure what to do at this intersection.

A lot of the problems and the concerns around self driving algorithms today has been that they often over confidently classify something and make a mistake, and just plough on regardless And so the the aim there will be to create systems that essentially are able to recognise when they don’t know what’s going on. And then either just shut down or hand back control to the, to the human. And so that notion of kind of collaboration based on confidence is something that we’re pursuing with colleagues at the Oxford robotics Institute at the moment. And they’re heavily involved in developing these kinds of technologies for self driving cars. So it’s not the goal is not to make a self aware robot, although, you know, maybe one day that will happen, but more to build in a minimal form of metacognition, into these systems.

Per Axbom
It made me realise that I mean, we don’t even have to talk about AIs. But actually, as we build websites and digital technology, we are making assumptions all the time about what the user wants next. But we could be much better at actually designing in doubt in that. I believe this is what you want next, but always allow the user to communicate back that no, that’s actually not what he wanted. So the website itself becomes almost like a chat interface, but in a different way. So always be able to, to be aware that what you assumed about the user isn’t always right. Well, really what James was talking about, because we’re making assumptions and mind reading constantly.

James Royal-Lawson
I mean, we do actually are thinking about it, the permanent with the rest, some situations where we already do that kind of, maybe it’s not kind of doubt of what we want to do. But a classic thing would be like a buy button. And, and maybe you’re underneath, you would offer another way of dealing with that interaction. Maybe someone’s not ready to buy something. But they kind of want to learn more. So you’d put like, maybe you’d put like a link or something underneath in proximity to that button to kind of give a second choice of thing. So I suppose that some whales mopping up that they’re maybe not everyone’s ready to buy, and that maybe there’s a second level thing there, or I guess that’s related what you’re saying.

Steve Fleming
Yeah, I think that the idea that providing a little bit of groundedness, in our interactions with autonomous technology, in terms of sharing confidence and uncertainty, we think could be quite powerful in in enabling trust. This is work that we’re doing at the moment, we haven’t got much data on it yet. But the hypothesis would be that one concern people have at the moment is that recommender systems and and assistants and this kind of stuff feels quite brittle. And we’ve seen that in the debate over the large language models recently that they seem to do some smart things, but then you can break them quite easily, often. And allowing a bit more, allowing that system to admit it doesn’t know. And that it’s uncertain, could then help us trust when it does produce an answer. Rather than just thinking hang on could be another example of producing nonsense. If it can signal that, in many scenarios, it will tell you that it doesn’t know the answer. But when it does tell you the answer, then you can be confident that it’s giving you a reasonable piece of advice that we hope could in increase trust.

Per Axbom
So interesting. I love that because I think that’s how we interact with humans as well. Because if someone is overconfident and always saying that they’re right and correct, you trust them less.

Steve Fleming
Exactly. So yeah, so we already have data from on human metacognition, where people do tend to, you know, be very sensitive to admissions of doubts in others that helps you trust statements of when they are competent.

James Royal-Lawson
I mean, I know, I know, I shouldn’t ask him a question, but I can’t resist it. But you have this whole thing though. This this dilemma, we’ve got, I guess that mean, the the work you’re doing, Stephen, and the things we’re talking about here, metacognition just shows how far or how not far we’ve come, I guess, in understanding that lump of grey matter inside our schools. And at the same time, now we’ve got global computing systems and global algorithms and stuff that’s, that’s steering our lives. You know, again, getting a bit nervous when we kind of ran I’m excited. And I think it’s fascinating the stuff to do messy condition of mind reading theory of mind and so on. But oh, my god, like kind of dilemma now, when we’re facing rolling out all these massive algorithms when we were but they don’t have the feedback mechanism. They don’t have the metacognition.

Steve Fleming
Yeah, no, I mean, I know. So yeah, we haven’t talked so much about the, the human brain, but it’s absolutely true that I mean, a lot of work we do in our lab is basic science on using brain imaging technologies to try and understand how metacognition works at a relatively coarse scale. So we use non invasive imaging like functional MRI, to look at different regions of the brain and the pattern of activations there when people are engaged in these kind of self assessments. And it’s complicated. There’s, you know, it’s a distributed system. It’s not straightforward to understand what different parts of that network are doing. And that’s a lot of the ongoing work we’re doing in our lab. But I think what is increasingly becoming clear, is that facility faculties like metacognition, just the capacity to have conscious experience and to communicate that experience to others. This is not the same thing as being smart and being intelligent.

So, consciousness is an intelligence used to, we used to think back in, you know, 50 years ago, a lot of psychologists would have put those two things hand in hand. And that was the origins of say, the Turing test, the idea that, like, you know, you can, from Alan Turing’s proposal that machine compasses is essentially met the bar for being human, like if you can have a conversation with it, and it can mimic being human. Now, large language models can essentially do that now. But that bar has now shifted. And I think part of the reason that bar has now shifted is that we recognise as a field as a society, that what it means to be conscious, and to be self aware, is actually very different to running a lot of numbers, and then coming up with a plausible answer. And so I think the big frontier in AI and neuroscience, going forward is going to be to understand what is missing from that picture? And, and why, and I don’t think we should necessarily be seeking to build that consciousness aspect into the machines if we find out what it is. But I think we need to be sensitive to that difference.

James Royal-Lawson
Can I think I’m gonna finish up I think this is actually taken from your, your book, our perception of the world is a controlled hallucination, a best guess, of what is out there. And I think that that is a really nice kind of reassurance that, you know, control to adults, maybe some people find it scary, but a controlled hallucination, but that’s kind of what we’re living through, isn’t it?

Steve Fleming
Yeah. So that, uh, yeah, that’s a quote from Anil Seth’s book “Being You”, which is also highly I would highly recommend, it is more focused on consciousness, and essentially, how, you know, what constructs this, this experience we have of the outside world, but I think it’s increasingly becoming clear from neuroscience that we don’t just passively perceive when the brain is not just a passive recipient of sensory data from the eyes, and from the ears and so on, we actively construct what is out there based on our experience. And I think that does provide a pretty humbling picture of the world, it means that we only see a very narrow slice of reality, and that reality is different for everybody.

Per Axbom
I love how that made me feel. It was like when I’m lying outside looking at the space, and it’s just infinite. And that what you just said, made me feel like looking in towards my brain. And that’s infinite, which is fantastic. Thank you so much, Steve, this has been wonderful.

Steve Fleming
It’s been a pleasure, thanks very much.

James Royal-Lawson
I still reflect on I just, I’m struggling to put things into words, because of the fact that I just think it’s really, really mind blowing. Mind blowing, when we’re talking about minds, that’s the whole thing as designers just not only I mean, the amount of self reflection we do or should be doing, but this whole thing of, of how much mind reading we do, how many simulations we we produce and run all the time with all the design work that we do. I mean, we’re constantly having to project and imagine how our, our users will think and behave when they meet our designs.

Per Axbom
And even when we see how they meet our designs we are interpreting were their behaviour and their thoughts and what they’re saying constantly all the time as well.

James Royal-Lawson
Yeah. So we, just because it, okay, it’s one thing when you’re doing mind reading of people, that you, that, you know, you know, people in your team that you’re doing have constant amount of mind reading projections about, okay, how would they respond when I present this design suggestion to them? How would they respond when I you know, do X or Y my team, but when we move, you know, a degree away from that and we’re talking about will a manager that of some kind of key person in my organisation I’ve never met, how will they respond, and maybe my understanding of how they respond is based on other people’s projections who have been shared with me. Then we get on to another level away from that our actual users that maybe we meet, maybe we don’t meet maybe we just meet them through an A B test or or some kind of like analytics interface. So you’re projecting then how they will respond and behave with your interface.

Per Axbom
Yeah, so what I’m realising there we, we are aware that the more we learn, or the more I know, the less I know, because so we’re diving into this topic with Steve. And he’s actually explaining a lot of the stuff of how the brain works. And he goes into depth on that in the book. But what that makes me realise is I know less. I know more about that area. But I now know less about how other people function. Because it’s, I’m realising so much becoming so aware of how all of the things I’m doing constantly, are just my interpretation. It’s just what I’m seeing is not reality, it can only exist because I’m thinking it.

James Royal-Lawson
Yeah, the reality you experience pair is individually to your reality.

Per Axbom
Right? So thinking about it, it’s, it’s not always good to become this aware. Because what it means is, I’m supposed to in like in a client situation, or when I’m selling my services, be really, really confident in that if we do this, based on research, this, this other thing will happen. That’s sort of what I’m promising. That’s sort of my profession. But the more I understand about how our brains work, the less confident I become.

James Royal-Lawson
And the more you realise that, you know, we’re looking for things that the the other, the person we’re communicating with, will expect to be heard. So we’re constantly, you know, circling around a shared thing, or trying to get to a shared thing.

Per Axbom
Exactly. And I think a really good example of this is something that he also has an example in his book is about, we sort of know, because we talked about this in UX about eyewitness reports. So if you have an eyewitness report, trusting that becomes really, really difficult depending on the situation and context and the amount of time that has passed since the event happened. And you have to realise that what we are experiencing with users as well, are eyewitness reports, they are telling us about their experiences, but how true are they? And how are they affected by our questions and our reasoning? And if there are other people involved? Are they affected by those people as well?

James Royal-Lawson
I reminds me of all the eye tracking, I used to do years ago that, you know, the way it works, the way I did research by tracking was that you record someone using or doing a task, but you wouldn’t talk to him or wouldn’t interrupt them while they were doing it. And then you play the recording back to them. So you could do they could see their own gaze across a website. And they they’ll be surprised many times of what they looked at, and how many times or how often or how much attention they paid certain things on the page, because their subconscious or their subconscious or dealt with it, not their conscious. So it was always interesting to have a dialogue around. I suppose they were theorising of why they looked at that thing and what their subconscious was doing. But so much at work here that we’re not fully in control of.

Per Axbom
Yeah, I just realised that this is something that he also has written about in his book, this paradox wherein to trust someone and see them as a leader and inspiring others to follow your lead means that you have had this image of really highly confident, assertiveness. But that’s not what makes you a good leader. Because what makes you a good leader is the ability to reflect and and understand these things that we’ve been talking about today. Which means that you need to sort of hide that part of yourself that you are always doubting everything. And in your outward profile. I’m sort of hoping that based on our episode talking about imposter syndrome talking about also about being introverts, and how that’s a good leadership skill, that perhaps this image of how you have to be to be a good leader is changing, really hoping. Well, it has to.

James Royal-Lawson
But based on this layers and layers and maybe even centuries of expected behaviour that’s been baked in to our our minds. That needs to be unpacked. God being a designer is a hard job. Maybe being a brain surgeon is easier. Recommended listening though.

Per Axbom
Oh, you’ve picked a good book here for a recommended listening. I mean, I mentioned impostor syndrome and the other one I’d ever just forgot which one I mentioned. But you have picked out 245 cognitive bias with David Dylan Thomas.

James Royal-Lawson
Yeah, wonderful chat with David and the and yet this episode is explaining more of things related to cognitive bias on a neuroscience level. And that conversation with my David is practical. I think for designers, yes, dealing with biases.

Per Axbom
Exactly. That’s a good one to feel more confident that it is you can do the right thing.

James Royal-Lawson
We can work with it. And if you’d like to contribute to funding or producing UX podcast, then visit UX podcast.com/support or just email us or something and say hey, I’d like to help out.

Per Axbom
Remember to keep moving.

James Royal-Lawson
See you on the other side.

[Music]

James Royal-Lawson
How does a brain say hello?

Per Axbom
I don’t know James, how does a brain say hello?

James Royal-Lawson
With a brain wave.

James Royal-Lawson
Being a designer is a hard job.


This is a transcript of a conversation between James Royal-LawsonPer Axbom, and Steve Fleming recorded in November 2022 and published as episode 303 of UX Podcast.