I'm a retired professor of philosophy and so am used to s-l-o-w reading. However now I'm retired and read for pleasure. I keep track and since retirement I've been reading approximately 200 books per year. Keep in mind I don't golf or garden.
I’m a retired physician with a serious reading habit, and in most years I get through 150-200 books. However, some of these are re-reads, some are beach books, some are graphic novels. I read the Bible regularly, and can get through most of it in a year; when I finish the Book of Revelation, I start over with Genesis. I’ve been told that John Quincy Adam’s did this, and got through the whole Bible at least six times before he died.
What a good post! I love fermi estimates, and the farther I get into my career as a physicist, the more and more I do them.
In case anyone is interested, this is a pdf of one of my favorite textbooks ever, that goes a long way towards teaching you how to think like a physicist, mainly by doing Fermi estimates:
Great post! This type of thinking used to be part and parcel of management consulting interviews. I had a friend who was asked during a McKinsey interview how many taxi medallions she estimated there were in NYC. The idea was to see if you could figure things out logically using known information and reasonable assumptions. (She actually just happened to know the exact number.)
I wonder if the ease with which people can just search online for an answer has stunted their ability to do this.
You’re probably right about the online searching. The problem is that when someone can’t immediately see that some claims are obvious hyperbole, if not a joke, but can’t be bothered to look it up, then they wind up believing them. Like the sushi.
Good to know there’s a term for this. I was told at a recent work seminar that the average person makes 35,000 decisions per day, which is absolutely unbelievable when you consider the number of minutes in a day. And I felt like the crazy one for stopping to do the math. I’m so sick of people inventing figures, which seem to so easily become urban myths. They know that the more they inflate a number, the more attention they’ll get.
But I'm just not sure that a facility with Fermi problems is really going to make a difference, because it assumes something that I'm pretty sure is demonstrably not true: it assumes that the student spreading the word about all those DOGE savings cares whether or not that anecdote is true.
You didn't finish your story - after you discovered that the claim was nonsense, did you share that with him? And if so, was his response "If they are making such obviously falsifiable lies to promote their position, maybe I should rethink my views"? Or was his response "yeah, whatever...."
The problem is (I don't claim this as an original insight) that people aren't interested in forming beliefs from evidence. They are interested in finding evidence that justifies their beliefs. That's how we are rigged, and not very many of us get trained out of that reflex.
This particular story did have a happy-ish ending. I told my student that the sushi thing was a satire and never actually happened, and emailed him a link to a Reuters fact-check. He was surprised to hear this and thanked me for setting him straight about it. Whether he made any further inferences I don’t know.
I just want to make sure I understand this comment: you are saying that even though you know the figure wasn't true and was intended as satire, it ... what, should be true? might as well be true? ought to be true? is true enough for you? is true enough for you that you are prepared to double down and invent your own false story about a sushi project manager to bolster it?
"It's false, but you should believe it because it bolsters the arguments for other claims that I'm ideologically required to believe and that are also probably false and that I'd like you also to believe."
While I'm a huge fan of Fermi estimates, I think anecdotes of this form "famous person has interesting outlier stats" are more plausibly true than you're giving credit for:
1. There's some cherry-picking going on, where people report the most interesting true(ish) things about their own behavior, and then the media amplifies it. If Bush eg eats an average amount of food and golfs an average amount of holes, that is uninteresting and ignored; it's the unusual, outlier things that Bush or his friends first bring up in conversation, and will make it past the reporter & editor's filter of interestingness.
2. People who are outlier on some dimension (eg being the president of the US) are likely to be outlier on some other dimensions (eg reading a lot). In fact, I'd think the specific trait of "reading much more than population average" is well-correlated with success, famously people like Bill Gates and Patrick Collison and Barack Obama are also huge infovores.
I think it amounts to a misuse of Fermi estimates, though, which instead of being a general-purpose method to calculate the plausibility of a claim, just assesses the order of magnitude to do a gut-check on its plausibility. It looks like the number of books Bush reads is actually plausible to an order of magnitude, which means that if your calculations are a bit off, then it could absolutely be right. For example, I read a heck of a lot more than sixty pages an hour, and if he has a similar reading speed to me - not at all implausible, I am unusual but by no means a freak - and most of the Bush book count is relatively short, well, it could be very much true. It seems to me that you took a wrong turn in the analysis.
Very useful in daily life. One has to allow for black swan events but overall applying reasonable approximations to test an assertion is good when it comes to facts.
In personal relationships it is always useful to look out for categorical “always” and “never” especially in accusations they almost always hide an underlying emotion or resentment. “you never do the dishes” - a statement that is provably false in most cases. But I wouldn’t challenge it directly unless you were ready with the statement- “You need to calm down” as the coup de grâce. 🥸
«2. Just rely on some heuristics like, “I think the government wastes a bunch of money and my family’s all Trump voters, so this sushi story fits right in with my worldview. I’m going to believe it is true.” The risk with this option, of course, is that you might have some pretty bad heuristics.»
It's not a bad heuristic. Of course, it's just inaccurately calibrated here. Like at some point, you should also remember the base facts of "a year has 365 days" and "600/365 ~= 1 million per day", and "a million dollars is 1000 * 1000 dollars".
Ground truth is irrelevant in a loose, happy-clappy or unreflective-confrontational social context, where political discussion is vibes-based, directional allegiance signalling, not epistemic ground truth. Your student failed to notice, that your specific academic context incentivized rigor (not that this is necessarily the norm in an academic setting).
These sorts of claims almost never require it. 99% of all claims to that nature he so far has heard, repeated or if he's particularly inventive, made up himself will be conveniently-framed half-truths, outright fabrications, misleading numbers, irrelevant click-bait equivalents.
But that doesn't mean he's incapable of it, if forced to bet $1,000 on the spot on the claim either being actually true or not, with five minutes of thinking and no internet allowed.
I'm sure he could make the correct choice. And if he chose wrong, it would likely have been a painful lesson, that sears some numerical heuristics into him. And one about thinking before you speak.
By the way, it's a good heuristic to never use rationalwiki as a source for anything. It's a platform for online-leftist wanting to imbue their political advocacy with rigor, they have no claim to. It's a normie trap, for those new to discussing online rationality or sometimes for those, that do not notice that the rationality-thing always seems to be "correct" about politics, if they're lefty/progressive themselves. Use LessWrong instead. You don't have to believe me, since even stating that, I overtly seem to betray a belligerent political bias. But please consider it.
The verdict of science is as follows: using a scientific calculator, I calculated that if you spent $200 on each sushi meal, and if you ate three sushi meals each day, you would have to live 2739.7260273 years to spend down the $600,000,000.
As this is pop science (I am a member of the pop rather than a scientist), I am assuming that the effects of inflation and the effects of compound interest offset one another well enough to be safely ignored here.
In high school and college I read (for pleasure, not classes) 60 pages per hour, for roughly 28 hours per week, or 1680 pages per week. 1680/63=26.66666667 times what GWB allegedly read. This was long before YouTube or Substack, so I’m down to a few dozen pages per week, amortized over a year (punk-eek like evolution).
I'm a retired professor of philosophy and so am used to s-l-o-w reading. However now I'm retired and read for pleasure. I keep track and since retirement I've been reading approximately 200 books per year. Keep in mind I don't golf or garden.
I read about ten books a year…why not more? Because I am addicted to reading stuff on Substack. I gotta stop…maybe tomorrow?
I’m a retired physician with a serious reading habit, and in most years I get through 150-200 books. However, some of these are re-reads, some are beach books, some are graphic novels. I read the Bible regularly, and can get through most of it in a year; when I finish the Book of Revelation, I start over with Genesis. I’ve been told that John Quincy Adam’s did this, and got through the whole Bible at least six times before he died.
What a good post! I love fermi estimates, and the farther I get into my career as a physicist, the more and more I do them.
In case anyone is interested, this is a pdf of one of my favorite textbooks ever, that goes a long way towards teaching you how to think like a physicist, mainly by doing Fermi estimates:
https://www.inference.org.uk/sanjoy/oom/book-a4.pdf
It’s called “Order-of-Magnitude Physics: Understanding the World with Dimensional
Analysis, Educated Guesswork, and White Lies”
Wonderful book, and a pleasure to read casually too, at least till the fluid dynamics chapters, which is a subject I’ve always disliked.
Great post! This type of thinking used to be part and parcel of management consulting interviews. I had a friend who was asked during a McKinsey interview how many taxi medallions she estimated there were in NYC. The idea was to see if you could figure things out logically using known information and reasonable assumptions. (She actually just happened to know the exact number.)
I wonder if the ease with which people can just search online for an answer has stunted their ability to do this.
You’re probably right about the online searching. The problem is that when someone can’t immediately see that some claims are obvious hyperbole, if not a joke, but can’t be bothered to look it up, then they wind up believing them. Like the sushi.
Good to know there’s a term for this. I was told at a recent work seminar that the average person makes 35,000 decisions per day, which is absolutely unbelievable when you consider the number of minutes in a day. And I felt like the crazy one for stopping to do the math. I’m so sick of people inventing figures, which seem to so easily become urban myths. They know that the more they inflate a number, the more attention they’ll get.
I am all for solving Fermi problems.
But I'm just not sure that a facility with Fermi problems is really going to make a difference, because it assumes something that I'm pretty sure is demonstrably not true: it assumes that the student spreading the word about all those DOGE savings cares whether or not that anecdote is true.
You didn't finish your story - after you discovered that the claim was nonsense, did you share that with him? And if so, was his response "If they are making such obviously falsifiable lies to promote their position, maybe I should rethink my views"? Or was his response "yeah, whatever...."
The problem is (I don't claim this as an original insight) that people aren't interested in forming beliefs from evidence. They are interested in finding evidence that justifies their beliefs. That's how we are rigged, and not very many of us get trained out of that reflex.
This particular story did have a happy-ish ending. I told my student that the sushi thing was a satire and never actually happened, and emailed him a link to a Reuters fact-check. He was surprised to hear this and thanked me for setting him straight about it. Whether he made any further inferences I don’t know.
"This particular story did have a happy-ish ending."
Yeah, whatever....
Tbh could completely believe the sushi figure. Besides, it’s government: 600 million spent on sushi in no way means 600 million of sushi.
And probably could foresee people complaining that doge laid off the poor sushi project manager millionaire
I just want to make sure I understand this comment: you are saying that even though you know the figure wasn't true and was intended as satire, it ... what, should be true? might as well be true? ought to be true? is true enough for you? is true enough for you that you are prepared to double down and invent your own false story about a sushi project manager to bolster it?
Not quite sure what you are doing here....
That it’s false, but also metaphorical for a lot of verifiable problems you see in such things as railway projects to anti homelessness grants
That it is false, but that we should ... pretend it is true for the greater good?
"It's false, but you should believe it because it bolsters the arguments for other claims that I'm ideologically required to believe and that are also probably false and that I'd like you also to believe."
“We’re snowed with numerical claims every day, sometimes they offer faux precision, sometimes they are outlandish on their face.”
I just read that 72.4 percent of numerical claims are outlandish on their face…
While I'm a huge fan of Fermi estimates, I think anecdotes of this form "famous person has interesting outlier stats" are more plausibly true than you're giving credit for:
1. There's some cherry-picking going on, where people report the most interesting true(ish) things about their own behavior, and then the media amplifies it. If Bush eg eats an average amount of food and golfs an average amount of holes, that is uninteresting and ignored; it's the unusual, outlier things that Bush or his friends first bring up in conversation, and will make it past the reporter & editor's filter of interestingness.
2. People who are outlier on some dimension (eg being the president of the US) are likely to be outlier on some other dimensions (eg reading a lot). In fact, I'd think the specific trait of "reading much more than population average" is well-correlated with success, famously people like Bill Gates and Patrick Collison and Barack Obama are also huge infovores.
I take your point, but I also think people valorize the famous and are prepared to believe just about anything about them.
I think it amounts to a misuse of Fermi estimates, though, which instead of being a general-purpose method to calculate the plausibility of a claim, just assesses the order of magnitude to do a gut-check on its plausibility. It looks like the number of books Bush reads is actually plausible to an order of magnitude, which means that if your calculations are a bit off, then it could absolutely be right. For example, I read a heck of a lot more than sixty pages an hour, and if he has a similar reading speed to me - not at all implausible, I am unusual but by no means a freak - and most of the Bush book count is relatively short, well, it could be very much true. It seems to me that you took a wrong turn in the analysis.
Very useful in daily life. One has to allow for black swan events but overall applying reasonable approximations to test an assertion is good when it comes to facts.
In personal relationships it is always useful to look out for categorical “always” and “never” especially in accusations they almost always hide an underlying emotion or resentment. “you never do the dishes” - a statement that is provably false in most cases. But I wouldn’t challenge it directly unless you were ready with the statement- “You need to calm down” as the coup de grâce. 🥸
Borges' Labyrinths is a great beach read because it's a great read anywhere.
«2. Just rely on some heuristics like, “I think the government wastes a bunch of money and my family’s all Trump voters, so this sushi story fits right in with my worldview. I’m going to believe it is true.” The risk with this option, of course, is that you might have some pretty bad heuristics.»
It's not a bad heuristic. Of course, it's just inaccurately calibrated here. Like at some point, you should also remember the base facts of "a year has 365 days" and "600/365 ~= 1 million per day", and "a million dollars is 1000 * 1000 dollars".
But bad!=inaccurate. Beliefs have to pay rent.
https://www.lesswrong.com/w/making-beliefs-pay-rent
Ground truth is irrelevant in a loose, happy-clappy or unreflective-confrontational social context, where political discussion is vibes-based, directional allegiance signalling, not epistemic ground truth. Your student failed to notice, that your specific academic context incentivized rigor (not that this is necessarily the norm in an academic setting).
These sorts of claims almost never require it. 99% of all claims to that nature he so far has heard, repeated or if he's particularly inventive, made up himself will be conveniently-framed half-truths, outright fabrications, misleading numbers, irrelevant click-bait equivalents.
https://www.lesswrong.com/posts/9weLK2AJ9JEt2Tt8f/politics-is-the-mind-killer
But that doesn't mean he's incapable of it, if forced to bet $1,000 on the spot on the claim either being actually true or not, with five minutes of thinking and no internet allowed.
I'm sure he could make the correct choice. And if he chose wrong, it would likely have been a painful lesson, that sears some numerical heuristics into him. And one about thinking before you speak.
By the way, it's a good heuristic to never use rationalwiki as a source for anything. It's a platform for online-leftist wanting to imbue their political advocacy with rigor, they have no claim to. It's a normie trap, for those new to discussing online rationality or sometimes for those, that do not notice that the rationality-thing always seems to be "correct" about politics, if they're lefty/progressive themselves. Use LessWrong instead. You don't have to believe me, since even stating that, I overtly seem to betray a belligerent political bias. But please consider it.
That place is absolute cancer.
If Elon Musk gifted me his fortune I could easily spend $600 million on sushi. I await the verdict of science and I'm happy to be proven wrong.
The verdict of science is as follows: using a scientific calculator, I calculated that if you spent $200 on each sushi meal, and if you ate three sushi meals each day, you would have to live 2739.7260273 years to spend down the $600,000,000.
As this is pop science (I am a member of the pop rather than a scientist), I am assuming that the effects of inflation and the effects of compound interest offset one another well enough to be safely ignored here.
In high school and college I read (for pleasure, not classes) 60 pages per hour, for roughly 28 hours per week, or 1680 pages per week. 1680/63=26.66666667 times what GWB allegedly read. This was long before YouTube or Substack, so I’m down to a few dozen pages per week, amortized over a year (punk-eek like evolution).
Enjoyable article. I always need a reminder to tune my BS detector a little
This sounds like a VERY fishy story to me.