This is great! I really enjoy this sort of post where someone shares what they've learned in an earnest but humble way.
As I was reading through the sections on autonomy and competence, I was thinking to myself, "but I have high autonomy and competence, yet I have terribly low intrinsic motivation these days"...then I got to the part on relatedness and it made sense. The most social connectedness I ever felt was during school and the early days of social media. These days, although I still have friends, I feel less truly connected to people. It's a hard problem to fix. I know a lot of other people these days have the same problem.
And, yeah, I'm the same re: relatedness. I haven't found myself in a rich upswelling of intrinsic motivation for a while – maybe the relatedness variable is the issue, because – same as you – much of my current life ticks the autonomy and competence boxes pretty well. It's a real challenge in this day and age – something we probably need to be consciously working at all the time. Saying that, it's easier said than done: with the internet, Netflix, video games, porn, etc., the incentives of modern life are structured in ways that steer us away from forming the kinds of connections we need/long for.
Nice, please do more literature reviews. This adds to the flow state literature, I think. If the task itself is painful, you can't submerge yourself in it. If you do, it hurts - because in the flow state you forget about all the external stuff by definition, so you're left with the inherent pain of the task.
I think what things exactly one develops intrinsic motivation for is not clear based on this. You kind of passed over this point, but you said you studied because you were good at it. But people are passionate about things they aren't good at, e.g. dancing or painting, just because they find them rewarding. So reward must come from somewhere else, at least in part. Maybe it's just to do with one's natural neural structure or whatever; some people like highly systemizing thinking, or problem solving, or thinking about relationships or whatever, because they're just wired like that. And then maybe the skill follows from the association between dopamine and learning.
(This implies you can discover natural talents, even after you previously disliked them, by finally getting good enough to breach that skill barrier where you feel rewarded. Of course, few people pursue things long enough for that. Although many passionate people seem to have felt passionate from the beginning, they may just be the only ones who stuck with it long enough. Maybe that's why people start to like things after doing them for a while and getting good; not because of the skill, but because the skill enabled them to learn, and somehow they ended up tapping into that hard-wired enjoyment of that kind of activity. I studied something quantitative in college and really disliked it; I hoped it would get better as I developed my skill at it; I ended up graduating with honors, but still hating it. And not because of burn-out, I'm fairly confident. This probably isn't a very satisfying theory, if what you enjoy is just random. But maybe that's the case.)
NGL, i did have the same thought while writing this, r.e. how do people come to like things that they're really bad at. This is something SDT – or, at least, my read of it – struggles with. I do have a theory, though:
Returning to tennis: i think that the rewards we experience while playing tennis happen at multiple levels at the same time. They're sort of nested inside each other. So, for example, there's the reward of hitting the ball crisply; there's the reward of seeing the ball go where you want it to; there's the the reward of winning the point; there's the reward of seeing the scoreboard change when you win the game, etc.
Some of the lower-order rewards (maybe all rewards), e.g. hitting the ball crisply, are probably dispensed in a way that's relative to one's skill level/expectations. For example, if a complete novice who usually frames every shot happens to land a beautiful crisp connection, i bet they experience a massive competence signal, whereas Roger Federer would experience the same kind of thing as BAU. In other words, it's the sense of improvement that delivers the competence signal for these elements of the reward. This kind of lines up with the dopamine reward prediction error stuff as well.
I also think – in line with what you were saying – that some of the basic, lower-order rewards connect with different people differently. Some people's brains may light up when they land that 1 in 100 crisp connection with the tennis ball; other people's brains may light up when they learn their first word of french. There's definitely quite a lot of individual variation, and i guess part of life is working out what activities light you up in that way!
I read the introduction and Part 1, and am going to save the rest for later once I have more time. Really interesting stuff so far! It mirrors my experience (even the timing of when you experienced a sudden jump in intrinsic motivation) to a T. I've had times where I've worked for 3-4 months, 7 days a week for upwards of 14 hours a day on a project, and others when I can't bring myself to spend more than an hour or two (even on the same project).
As far as reading (something I never did outside of compulsion until I was 18), somehow I've become very intrinsically motivated to become an avid reader of extremely dense, and culturally important texts. I wouldn't at all be surprised if I found an exact description of what happened when I return to this article.
Hey! Interesting - i wonder how many people are in the same boat.
Funny you should mention the reading thing, because i’m exactly the same. Makes me think of all the people on reddit who say they hated reading classics until they’d been out of school for a few years, at which point they finally clicked for them. I think the autonomy component of the SDT theory does a good job of explaining this - and also gives some indication as to how we might want to structure our school system to make sure we’re not crushing the life out of people’s natural curiosity!
Most of us live in an inescapable world of rewards and evaluation: we have jobs that pay for tasks completed, kids get evalutated with grades, etc.
The challenge is how to build intrinsic motivation despite the fact that we live in a world of rewards and evaluations.
After reading your deep dive, I think maybe the answer is to engineer autonomy, competence, and relatedness for yourself, and then find a way to not be distracted from those goals.
When I was at my peak in my career, I had all three. The money felt like a bonus.
The "Mastery, Autonomy, Purpose" framework seems like a close parallel. And in fact, "shared goals" (relatedness-adjacent) are the most common backstop "purpose"
Eg Conscripted soldiers in war would commonly fight for their squad mates even if they thought the war was dumb.
The Yerkes-Dodson law also seems close to your dopamine reinforcement explanation. Eg known money reward + new puzzle task overstimulates some of the subjects, so arousal passes the peak, dopamine goes below baseline,.and performance/focus declines
Great article! SDT has fascinating parallels in consciousness science that no one to my knowledge has actually explored. This is because in consciousness research, the experiencing self is a given, it just happens to be taken for granted that there's an "I" experiencing, and the wonder and magic is focused on the experience itself. What about the self that is experiencing? On the other hand, SDT operates at a level where a biological and even experiencing conscious self is taken for granted, and the focus is on how the cognitive self operates (in many ways). And this is also where the criticism comes from. This is all in the domain of the self and motivation and whatnot articulated in language. To go deeper, we need a bridge between these two that can explain how the self is constructed. And we do have a beautiful theory/framework for that
Consciousness is a consensus mechanism by which the self is constructed. It is a recursive loop where the self emerges, experiences, and folds in the next experience to create an evolving, expanding self. With language we have the ability to freeze many of these ideas and we are able to go much further. "I can think, feel, experience and reflect on this"
And why a consensus mechanism? Because "you" are actually a constellation of cells and experiences that needs to be sufficiently decentralized but also be able to act and plan in the very short and long term. How do you get 87 billion cells (in our case) to decide as one? That is actually a pretty difficult engineering problem where you have to think about both compute (all the different data streams coming from different sources need to be digested and acted on) and commute (one cell group in the prefrontal cortex needs to immediately broadcast a danger message to other corners of the brain, and we dont have direct wiring)
Now the natural question to ask is, what do synthetic beings need to develop both? If you are interested you might want to read our book Journey of the Mind
Great article! SDT has fascinating parallels in consciousness science that no one to my knowledge has actually explored. This is because in consciousness research, the experiencing self is a given, it just happens to be taken for granted that there's an "I" experiencing, and the wonder and magic is focused on the experience itself. What about the self that is experiencing? On the other hand, SDT operates at a level where a biological and even experiencing conscious self is taken for granted, and the focus is on how the cognitive self operates (in many ways). And this is also where the criticism comes from. This is all in the domain of the self and motivation and whatnot articulated in language. To go deeper, we need a bridge between these two that can explain how the self is constructed. And we do have a beautiful theory/framework for that
Consciousness is a consensus mechanism by which the self is constructed. It is a recursive loop where the self emerges, experiences, and folds in the next experience to create an evolving, expanding self. With language we have the ability to freeze many of these ideas and we are able to go much further. "I can think, feel, experience and reflect on this"
And why a consensus mechanism? Because "you" are actually a constellation of cells and experiences that needs to be sufficiently decentralized but also be able to act and plan in the very short and long term. How do you get 87 billion cells (in our case) to decide as one? That is actually a pretty difficult engineering problem where you have to think about both compute (all the different data streams coming from different sources need to be digested and acted on) and commute (one cell group in the prefrontal cortex needs to immediately broadcast a danger message to other corners of the brain, and we dont have direct wiring)
Now the natural question to ask is, what do synthetic beings need to develop both? If you are interested you might want to read our book Journey of the Mind
This is great! I really enjoy this sort of post where someone shares what they've learned in an earnest but humble way.
As I was reading through the sections on autonomy and competence, I was thinking to myself, "but I have high autonomy and competence, yet I have terribly low intrinsic motivation these days"...then I got to the part on relatedness and it made sense. The most social connectedness I ever felt was during school and the early days of social media. These days, although I still have friends, I feel less truly connected to people. It's a hard problem to fix. I know a lot of other people these days have the same problem.
Thanks Andy! Appreciate the kind feedback.
And, yeah, I'm the same re: relatedness. I haven't found myself in a rich upswelling of intrinsic motivation for a while – maybe the relatedness variable is the issue, because – same as you – much of my current life ticks the autonomy and competence boxes pretty well. It's a real challenge in this day and age – something we probably need to be consciously working at all the time. Saying that, it's easier said than done: with the internet, Netflix, video games, porn, etc., the incentives of modern life are structured in ways that steer us away from forming the kinds of connections we need/long for.
Nice, please do more literature reviews. This adds to the flow state literature, I think. If the task itself is painful, you can't submerge yourself in it. If you do, it hurts - because in the flow state you forget about all the external stuff by definition, so you're left with the inherent pain of the task.
I think what things exactly one develops intrinsic motivation for is not clear based on this. You kind of passed over this point, but you said you studied because you were good at it. But people are passionate about things they aren't good at, e.g. dancing or painting, just because they find them rewarding. So reward must come from somewhere else, at least in part. Maybe it's just to do with one's natural neural structure or whatever; some people like highly systemizing thinking, or problem solving, or thinking about relationships or whatever, because they're just wired like that. And then maybe the skill follows from the association between dopamine and learning.
(This implies you can discover natural talents, even after you previously disliked them, by finally getting good enough to breach that skill barrier where you feel rewarded. Of course, few people pursue things long enough for that. Although many passionate people seem to have felt passionate from the beginning, they may just be the only ones who stuck with it long enough. Maybe that's why people start to like things after doing them for a while and getting good; not because of the skill, but because the skill enabled them to learn, and somehow they ended up tapping into that hard-wired enjoyment of that kind of activity. I studied something quantitative in college and really disliked it; I hoped it would get better as I developed my skill at it; I ended up graduating with honors, but still hating it. And not because of burn-out, I'm fairly confident. This probably isn't a very satisfying theory, if what you enjoy is just random. But maybe that's the case.)
Sorry, I'm stoned and rambling.
Hey! Thanks for the feedback/comment!
NGL, i did have the same thought while writing this, r.e. how do people come to like things that they're really bad at. This is something SDT – or, at least, my read of it – struggles with. I do have a theory, though:
Returning to tennis: i think that the rewards we experience while playing tennis happen at multiple levels at the same time. They're sort of nested inside each other. So, for example, there's the reward of hitting the ball crisply; there's the reward of seeing the ball go where you want it to; there's the the reward of winning the point; there's the reward of seeing the scoreboard change when you win the game, etc.
Some of the lower-order rewards (maybe all rewards), e.g. hitting the ball crisply, are probably dispensed in a way that's relative to one's skill level/expectations. For example, if a complete novice who usually frames every shot happens to land a beautiful crisp connection, i bet they experience a massive competence signal, whereas Roger Federer would experience the same kind of thing as BAU. In other words, it's the sense of improvement that delivers the competence signal for these elements of the reward. This kind of lines up with the dopamine reward prediction error stuff as well.
I also think – in line with what you were saying – that some of the basic, lower-order rewards connect with different people differently. Some people's brains may light up when they land that 1 in 100 crisp connection with the tennis ball; other people's brains may light up when they learn their first word of french. There's definitely quite a lot of individual variation, and i guess part of life is working out what activities light you up in that way!
I read the introduction and Part 1, and am going to save the rest for later once I have more time. Really interesting stuff so far! It mirrors my experience (even the timing of when you experienced a sudden jump in intrinsic motivation) to a T. I've had times where I've worked for 3-4 months, 7 days a week for upwards of 14 hours a day on a project, and others when I can't bring myself to spend more than an hour or two (even on the same project).
As far as reading (something I never did outside of compulsion until I was 18), somehow I've become very intrinsically motivated to become an avid reader of extremely dense, and culturally important texts. I wouldn't at all be surprised if I found an exact description of what happened when I return to this article.
Hey! Interesting - i wonder how many people are in the same boat.
Funny you should mention the reading thing, because i’m exactly the same. Makes me think of all the people on reddit who say they hated reading classics until they’d been out of school for a few years, at which point they finally clicked for them. I think the autonomy component of the SDT theory does a good job of explaining this - and also gives some indication as to how we might want to structure our school system to make sure we’re not crushing the life out of people’s natural curiosity!
Most of us live in an inescapable world of rewards and evaluation: we have jobs that pay for tasks completed, kids get evalutated with grades, etc.
The challenge is how to build intrinsic motivation despite the fact that we live in a world of rewards and evaluations.
After reading your deep dive, I think maybe the answer is to engineer autonomy, competence, and relatedness for yourself, and then find a way to not be distracted from those goals.
When I was at my peak in my career, I had all three. The money felt like a bonus.
The "Mastery, Autonomy, Purpose" framework seems like a close parallel. And in fact, "shared goals" (relatedness-adjacent) are the most common backstop "purpose"
Eg Conscripted soldiers in war would commonly fight for their squad mates even if they thought the war was dumb.
The Yerkes-Dodson law also seems close to your dopamine reinforcement explanation. Eg known money reward + new puzzle task overstimulates some of the subjects, so arousal passes the peak, dopamine goes below baseline,.and performance/focus declines
Great article! SDT has fascinating parallels in consciousness science that no one to my knowledge has actually explored. This is because in consciousness research, the experiencing self is a given, it just happens to be taken for granted that there's an "I" experiencing, and the wonder and magic is focused on the experience itself. What about the self that is experiencing? On the other hand, SDT operates at a level where a biological and even experiencing conscious self is taken for granted, and the focus is on how the cognitive self operates (in many ways). And this is also where the criticism comes from. This is all in the domain of the self and motivation and whatnot articulated in language. To go deeper, we need a bridge between these two that can explain how the self is constructed. And we do have a beautiful theory/framework for that
Consciousness is a consensus mechanism by which the self is constructed. It is a recursive loop where the self emerges, experiences, and folds in the next experience to create an evolving, expanding self. With language we have the ability to freeze many of these ideas and we are able to go much further. "I can think, feel, experience and reflect on this"
And why a consensus mechanism? Because "you" are actually a constellation of cells and experiences that needs to be sufficiently decentralized but also be able to act and plan in the very short and long term. How do you get 87 billion cells (in our case) to decide as one? That is actually a pretty difficult engineering problem where you have to think about both compute (all the different data streams coming from different sources need to be digested and acted on) and commute (one cell group in the prefrontal cortex needs to immediately broadcast a danger message to other corners of the brain, and we dont have direct wiring)
Now the natural question to ask is, what do synthetic beings need to develop both? If you are interested you might want to read our book Journey of the Mind
A bit about the idea of consciousness as a consensus mechanism here https://saigaddam.medium.com/consciousness-is-a-consensus-mechanism-2b399c9ec4b5
Great article! SDT has fascinating parallels in consciousness science that no one to my knowledge has actually explored. This is because in consciousness research, the experiencing self is a given, it just happens to be taken for granted that there's an "I" experiencing, and the wonder and magic is focused on the experience itself. What about the self that is experiencing? On the other hand, SDT operates at a level where a biological and even experiencing conscious self is taken for granted, and the focus is on how the cognitive self operates (in many ways). And this is also where the criticism comes from. This is all in the domain of the self and motivation and whatnot articulated in language. To go deeper, we need a bridge between these two that can explain how the self is constructed. And we do have a beautiful theory/framework for that
Consciousness is a consensus mechanism by which the self is constructed. It is a recursive loop where the self emerges, experiences, and folds in the next experience to create an evolving, expanding self. With language we have the ability to freeze many of these ideas and we are able to go much further. "I can think, feel, experience and reflect on this"
And why a consensus mechanism? Because "you" are actually a constellation of cells and experiences that needs to be sufficiently decentralized but also be able to act and plan in the very short and long term. How do you get 87 billion cells (in our case) to decide as one? That is actually a pretty difficult engineering problem where you have to think about both compute (all the different data streams coming from different sources need to be digested and acted on) and commute (one cell group in the prefrontal cortex needs to immediately broadcast a danger message to other corners of the brain, and we dont have direct wiring)
Now the natural question to ask is, what do synthetic beings need to develop both? If you are interested you might want to read our book Journey of the Mind
A bit about the idea of consciousness as a consensus mechanism here https://saigaddam.medium.com/consciousness-is-a-consensus-mechanism-2b399c9ec4b5
Thanks for this! Good read! I'd read a book on this take