“Oh, I don’t mean that on consideration I endorse purchasing one—I’m just expressing my urge to to eat a donut.”
There are two meanings of ‘want’ in common usage. A feeling of desire, and an endorsed intention. ‘I want a baby tiger!’ is not analogous to ‘I want to work more on my taxes tonight’.
These ‘want’s are basically the input of a decision process and the output. I feel desire for a baby tiger and utter ‘I want it!’, and my brain considers that desire plus some other stuff about baby tigers and my life, and decides that on reflection I do not intend to acquire one. On the other hand, I feel no positive attraction toward my taxes at all, yet my aversion to prison and lawyers and generally being disagreeable in any way, once fed through my decision process leave me ‘wanting’ to work on them.
It is hard to be confused about the baby tiger and tax cases, but other times I think this leads to genuine confusion. The donut case above was a genuine confusion, but one of no importance. I think it leads to more important genuine confusions when one talks to oneself, and lacks two distinct concepts.
Luckily I waited ages before writing this blog post, and so came across David Wong of Cracked talking about something similar, as the #1 Way you are sabotaging your own life (without knowing it), which I’ll just quote here in full (the section, not the whole post, though I don’t agree with all of):
#1. Lying to Yourself About What You Actually Want
Hlib Shabashnyi/iStock/Getty Images
Off the top of your head, say something you’ve always wanted to do. Then, follow it up with why you’ve never done it.
So, maybe you said something like, “I’ve always wanted to start a little business selling cupcakes! But I wouldn’t even know how to get started!”
Aaaaand … 90 percent of you just lied.
I know you did, because if you actually wanted to do the thing, then the second part — the obstacle — wouldn’t exist. For example, if that person up there actually wanted to start their cupcake business, they wouldn’t be confused about how to get started. They’d be a freaking walking encyclopedia of information about how to get started, because they’d have spent every single day reading up on it and calling other cupcake-shop owners for advice. They don’t do that because they don’t actually want it. They don’t have the invisible gun to their head.
BrandlMichaela/iStock/Getty Images “The cupcake is a lie.”
This, right here, is at the heart of every unfulfilled ambition in your life. We use the same word — “want” — to mean two completely different things, and the constant confusion between those definitions is why so many people are disappointed in how their lives turned out. Depending on the context, “want” can be:
A) A statement of intended action (“I want to mow the lawn before it rains.”)
B) A statement of general preference (“I want everyone to live a long and happy life.”)
It sounds simple enough, but the confusion of those two uses of the word is everything. We switch between the two definitions sometimes in the same sentence. This morning, I was driving to Five Guys to get a burger and an entire grocery bag full of french fries to go with it (that is, the “small”). I passed a guy who was jogging, shirtless, who had a torso like Matthew McConaughey. I said to myself, “I want a body like that!” And, if I’d pulled over and asked the guy why he runs and works out, he’d have said the same thing, almost word-for-word — “It’s because I want a body like this!”
Same phrasing, meaning two completely different things. I used “want” in the same way I say I want world peace — a wistful statement about something I actually have no control over. If it’s the same effort either way, sure, I’ll take the rock-hard abs — give me an ab pill and I’ll swallow it. Otherwise, no, it ain’t happening. That jogging guy, on the other hand, used “want” as a statement of intended action — he “wants” to run five miles every day because he “wants” to be fit.
desertsolitaire/iStock/Getty Images “Also because there’s a guy with a gun pointed at me. Please, call the police.”
Now look around you — look at all of the minimum-wage people who “want” to be rich and/or famous, with some vague notion of, I don’t know, being on a reality show some day or getting “discovered” for some talent they didn’t know they had. Now look at all of the MBAs working 100-hour weeks on the trading floor because they “want” to be rich. The difference in the two is night and day, but in many cases the former group doesn’t realize it. They just stay poor while the other group starts shopping for vacation homes.
And I’m starting to think that the world really is divided between those who have a clear idea of what it means to want something — including the total cost and sacrifices it will take to get it — and those who are just content to leave it as an airy “wouldn’t it be nice” fantasy. The former group hones in on what they want and goes zooming after it like a shark. The latter looks at them, shakes their head and says, “How do they do it?” As if they have a cheat code, or a secret technique.
Eldad Carin/iStock/Getty Images “That son of bitch and his Konami code.”
“What, you’re saying we should all be douchebag stockbrokers working hundred-hour weeks?” No. I’m saying that while some of you are sitting around the coffee shop talking about how you “want” the system to change, that douchebag is accumulating money so he can actually run for congress. Because when he “wants” something, he doesn’t sing a song about it. He prices that shit and makes a down payment. And when that relentless BMW-driving douche has kids, he’ll teach them, too, what it really means to “want” something — to be single-minded, and voracious, and to pursue it to the ends of the Earth. Instilling that lesson goes just as far toward preserving wealth and power in a group as the actual inheritance they’ll leave behind.
Are you scared of those people? Are you imagining them as cold-blooded stock brokers and lobbyists and swindlers, the Wolf of Wall Street types who are eating away at the world like a cancer? Well, they scare you because it’s a glimpse at what accomplishing great things actually costs. You know Steve Jobs was a fucking psychopath, right? So the next time somebody asks you if you want to be rich, really stop and think about it. Think about what it will take. Think about what kind of person you’ll need to become.
David Paul Morris/Getty Images News/Getty Images “I would literally make them from the blood of orphans if it could save me five cents on the per-unit cost.”
And that’s the point of all this — I’ve found, as time goes on, that everybody gets what they want. Not what they say they want in order to make themselves look good to others, or what they tell themselves they want so they feel better about the current state of their life. No, I’m talking about what they really want. And to find out what they really want, you don’t need to ask them. You just need to look at what they did today. You want to change, start there.
He has a more complicated thesis, and is saying things I don’t necessarily agree with, but his central point is that mistaking one kind of ‘want’ for another is something like the number one way you are messing up your life, which suggests that he considers it an important confusion.
He thinks it is related to the strange discrepancy between our imagined prospects in five years, and what we do right now. You say to yourself ‘I want to be a classical guitarist’—and it arises as an idle positive urge toward playing classical guitar, which arises due to letting your mind wander while listening to classical guitar one time. Then you just kind of figure that you do in fact want to, but don’t know how to right now or there are some obstacles or something, and hopefully you’ll figure it out in the vague future and probably one day be a great classical guitarist.
I claim that part of what is going on here is that you are observing your urge to play classical guitar, and thinking of it as ‘I want to play classical guitar’. Then you fail to distinguish this input to a decision from an actual decision. So either you sign up for a classical guitar class, but feel kind of bad about it and like you have too many things going on in your life, or you say to yourself that you want to, but you don’t, and you figure there is something wrong with you.
So maybe you ask yourself, ‘do I really want to play classical guitar?’ and you look inside your heart and see that you do really feel warmly about playing classical guitar, and you don’t notice that that is the answer to a different question than the one that is relevant to whether you should play classical guitar. You are checking that the input really is ‘classical guitar is nice’, rather than that the output really is ‘learning to play classical guitar is what I want on net, given the costs, and that it is nice’.
In some sense, if you fail to distinguish the input to your decision from the output, and confusedly use the input where you would rationally use the output, you are in fact correct that they are the same. It’s just that you are missing out on making a decision where you could benefit from doing so. Like, if you mistakenly treat ‘potato chips taste delicious’ as logically identical to ‘I endorse eating potato chips’ because you call them both “I want to eat potato chips!”, then you are missing out on a great chance to take into account considerations other than the flavor of potato chips in your diet, some of which may be important to you.
I have seen myself making this error. I remember it happening when I’m mostly thinking about something else, but idly appreciating something in my surrounds. For instance, I’m likely to see a cool startup and think ‘mmm, yeah I should have a startup’ or see a nice blouse and think ‘ooh, I should get a blouse like that’. And if I was really thinking about the issue, I would remember that things have costs, but if I’m not then the part of my brain that says ‘ooh’ at stuff also registers them as tentative decisions. I go away assuming that I now intend some day to have a cool startup.
This seems to parallel other useful distinctions I have seen people talk about in recent history. ‘Impressions’ and ‘beliefs’, and ‘what words are being said in my head’ versus ‘what I believe and stand for’ for instance. These are similarly inputs and outputs of decision processes, and by naming them as such, we can remember to actually stick a decision process between them. Much like if we label ‘raw pasta’ and ‘cooked pasta’ different things, it is easier to notice that cooking is an important step, and we are less likely to end up with pasta that is weirdly ill suited to being eaten half the time. Instead of just citing our impressions as beliefs, and then arguing with other people, or being confused that our ‘beliefs’ aren’t updating when other people tell us theirs (and we can tell they aren’t, by checking our impressions), we can just have some impressions and then consider them when deciding what we believe. To me these distinctions sound so obvious in retrospect that it is weird to even hypothesize that a moment ago you might not have made them. But relatedly, I think pointing them out has been pretty useful.
I’m not sure why we don’t make clear conceptual distinctions in these cases. Perhaps making distinctions is just hard. I think maybe these things are not so obvious because we didn’t always do so much intelligent decision-making. Presumably in goodest oldest days, ‘this food tastes good’ was closer in meaning to ‘I have decided to eat this food’ and ‘I feel like this plan is going to fail’ was closer in meaning to ‘this plan is probably going to fail’, and it is only later that considerations like health at age fifty and outside view evidence became sufficiently worth manually adding to one’s decision calculus to bother having a decision calculus to add them to, beyond unreflective feelings and intuitions.
Wong’s post also discusses people’s failure to connect their plans with the costs of those plans. They plan to learn the classical guitar, but don’t think of themselves as ‘planning to learn the classical guitar instead of spending so much time with their friends’ or ‘planning to learn the classical guitar instead of spending one of their two daily outside-of-work intentional-action-slots on reading’ or whatever the bottleneck may be. I don’t think he explicitly connects the two, though he clearly thinks they are part of a larger related structure. They are at least related in that people are wrong about what they want in part because they are not considering the costs. In his story, he wants to be fit, but only if he completely ignores the costs. Which is to say, if he just checks whether he likes the idea of being fit. He describes being confused, and going on to tell himself that he wanted to be fit in the ‘intending to do it’ sense, while mysteriously not being inclined to.
My model gives a straightforward reason for these errors to be related. Suppose you are making a cost-benefit analysis. If you confuse a single entry in the ‘benefits’ column with the output of the entire analysis, this reliably undercounts costs, to put it mildly. That is, if you ‘decide’ to learn Japanese by just using ‘liking the idea of knowing Japanese’ as a proxy for making a decision, then your decisions will be independent of what will be lost by learning Japanese. You will do the same thing whether it takes three hours or three million hours to learn Japanese. And you will find yourself with some very bad decisions that you can’t bring yourself to actually uphold.
I was told at a CFAR workshop that it is useful to say to yourself ‘if I write this paper it is going to be super annoying, but worth it’ rather than just ‘writing this paper is worth it’. This seems probably related. For instance, maybe you are just using ‘I want a tiger’ to decide to acquire one, and then if you say ‘I want a tiger, but it will be very messy, oh wait maybe I don’t’ or ‘I want a tiger, and it will be very messy, and still worth it’ then in the latter case you believe yourself more, or something.
Anyway, maybe more distinct terms would be helpful. For now I’m going to use ‘yearn for’ and ‘intend’, like ‘I yearn for a donut but I don’t intend to get one’. Better suggestions welcome.
Want like want want
“I want a donut”
“Ok, I’ll buy one for you”
“Oh, I don’t mean that on consideration I endorse purchasing one—I’m just expressing my urge to to eat a donut.”
There are two meanings of ‘want’ in common usage. A feeling of desire, and an endorsed intention. ‘I want a baby tiger!’ is not analogous to ‘I want to work more on my taxes tonight’.
These ‘want’s are basically the input of a decision process and the output. I feel desire for a baby tiger and utter ‘I want it!’, and my brain considers that desire plus some other stuff about baby tigers and my life, and decides that on reflection I do not intend to acquire one. On the other hand, I feel no positive attraction toward my taxes at all, yet my aversion to prison and lawyers and generally being disagreeable in any way, once fed through my decision process leave me ‘wanting’ to work on them.
It is hard to be confused about the baby tiger and tax cases, but other times I think this leads to genuine confusion. The donut case above was a genuine confusion, but one of no importance. I think it leads to more important genuine confusions when one talks to oneself, and lacks two distinct concepts.
Luckily I waited ages before writing this blog post, and so came across David Wong of Cracked talking about something similar, as the #1 Way you are sabotaging your own life (without knowing it), which I’ll just quote here in full (the section, not the whole post, though I don’t agree with all of):
He has a more complicated thesis, and is saying things I don’t necessarily agree with, but his central point is that mistaking one kind of ‘want’ for another is something like the number one way you are messing up your life, which suggests that he considers it an important confusion.
He thinks it is related to the strange discrepancy between our imagined prospects in five years, and what we do right now. You say to yourself ‘I want to be a classical guitarist’—and it arises as an idle positive urge toward playing classical guitar, which arises due to letting your mind wander while listening to classical guitar one time. Then you just kind of figure that you do in fact want to, but don’t know how to right now or there are some obstacles or something, and hopefully you’ll figure it out in the vague future and probably one day be a great classical guitarist.
I claim that part of what is going on here is that you are observing your urge to play classical guitar, and thinking of it as ‘I want to play classical guitar’. Then you fail to distinguish this input to a decision from an actual decision. So either you sign up for a classical guitar class, but feel kind of bad about it and like you have too many things going on in your life, or you say to yourself that you want to, but you don’t, and you figure there is something wrong with you.
So maybe you ask yourself, ‘do I really want to play classical guitar?’ and you look inside your heart and see that you do really feel warmly about playing classical guitar, and you don’t notice that that is the answer to a different question than the one that is relevant to whether you should play classical guitar. You are checking that the input really is ‘classical guitar is nice’, rather than that the output really is ‘learning to play classical guitar is what I want on net, given the costs, and that it is nice’.
In some sense, if you fail to distinguish the input to your decision from the output, and confusedly use the input where you would rationally use the output, you are in fact correct that they are the same. It’s just that you are missing out on making a decision where you could benefit from doing so. Like, if you mistakenly treat ‘potato chips taste delicious’ as logically identical to ‘I endorse eating potato chips’ because you call them both “I want to eat potato chips!”, then you are missing out on a great chance to take into account considerations other than the flavor of potato chips in your diet, some of which may be important to you.
I have seen myself making this error. I remember it happening when I’m mostly thinking about something else, but idly appreciating something in my surrounds. For instance, I’m likely to see a cool startup and think ‘mmm, yeah I should have a startup’ or see a nice blouse and think ‘ooh, I should get a blouse like that’. And if I was really thinking about the issue, I would remember that things have costs, but if I’m not then the part of my brain that says ‘ooh’ at stuff also registers them as tentative decisions. I go away assuming that I now intend some day to have a cool startup.
This seems to parallel other useful distinctions I have seen people talk about in recent history. ‘Impressions’ and ‘beliefs’, and ‘what words are being said in my head’ versus ‘what I believe and stand for’ for instance. These are similarly inputs and outputs of decision processes, and by naming them as such, we can remember to actually stick a decision process between them. Much like if we label ‘raw pasta’ and ‘cooked pasta’ different things, it is easier to notice that cooking is an important step, and we are less likely to end up with pasta that is weirdly ill suited to being eaten half the time. Instead of just citing our impressions as beliefs, and then arguing with other people, or being confused that our ‘beliefs’ aren’t updating when other people tell us theirs (and we can tell they aren’t, by checking our impressions), we can just have some impressions and then consider them when deciding what we believe. To me these distinctions sound so obvious in retrospect that it is weird to even hypothesize that a moment ago you might not have made them. But relatedly, I think pointing them out has been pretty useful.
I’m not sure why we don’t make clear conceptual distinctions in these cases. Perhaps making distinctions is just hard. I think maybe these things are not so obvious because we didn’t always do so much intelligent decision-making. Presumably in goodest oldest days, ‘this food tastes good’ was closer in meaning to ‘I have decided to eat this food’ and ‘I feel like this plan is going to fail’ was closer in meaning to ‘this plan is probably going to fail’, and it is only later that considerations like health at age fifty and outside view evidence became sufficiently worth manually adding to one’s decision calculus to bother having a decision calculus to add them to, beyond unreflective feelings and intuitions.
Wong’s post also discusses people’s failure to connect their plans with the costs of those plans. They plan to learn the classical guitar, but don’t think of themselves as ‘planning to learn the classical guitar instead of spending so much time with their friends’ or ‘planning to learn the classical guitar instead of spending one of their two daily outside-of-work intentional-action-slots on reading’ or whatever the bottleneck may be. I don’t think he explicitly connects the two, though he clearly thinks they are part of a larger related structure. They are at least related in that people are wrong about what they want in part because they are not considering the costs. In his story, he wants to be fit, but only if he completely ignores the costs. Which is to say, if he just checks whether he likes the idea of being fit. He describes being confused, and going on to tell himself that he wanted to be fit in the ‘intending to do it’ sense, while mysteriously not being inclined to.
My model gives a straightforward reason for these errors to be related. Suppose you are making a cost-benefit analysis. If you confuse a single entry in the ‘benefits’ column with the output of the entire analysis, this reliably undercounts costs, to put it mildly. That is, if you ‘decide’ to learn Japanese by just using ‘liking the idea of knowing Japanese’ as a proxy for making a decision, then your decisions will be independent of what will be lost by learning Japanese. You will do the same thing whether it takes three hours or three million hours to learn Japanese. And you will find yourself with some very bad decisions that you can’t bring yourself to actually uphold.
I was told at a CFAR workshop that it is useful to say to yourself ‘if I write this paper it is going to be super annoying, but worth it’ rather than just ‘writing this paper is worth it’. This seems probably related. For instance, maybe you are just using ‘I want a tiger’ to decide to acquire one, and then if you say ‘I want a tiger, but it will be very messy, oh wait maybe I don’t’ or ‘I want a tiger, and it will be very messy, and still worth it’ then in the latter case you believe yourself more, or something.
Anyway, maybe more distinct terms would be helpful. For now I’m going to use ‘yearn for’ and ‘intend’, like ‘I yearn for a donut but I don’t intend to get one’. Better suggestions welcome.