Meta: Should we be encouraging people to read popular books? Popular books have their place, but there is a common failure mode where people read popularizations, get their daily fix of insight porn, and go away without having learned anything substantial.
Reading popular books on the same subject have sharply decreasing marginal utility, and I’m guessing that the kind of person who reads LW is the kind of person who has already read enough popularizations that reading another is probably a waste of time.
I agree with you about popular books. But I am wondering about the term “insight porn”—does it do anything besides put a negative connotation on the process of obtaining new insights? (If “insight porn” refers to useless insights, would it be appropriate to say something like “pure math is full of insight porn”? :P)
The way I’ve heard it used is to describe activities that give you the illusion of understanding a complex subject. Examples would be Malcolm Gladwell’s books, or the high school student who thinks they understand quantum mechanics after reading The Elegant Universe.
So that usage wouldn’t fit pure maths, since we seem to generally agree that pure mathematicians have a true understanding of their fields, even if we don’t agree on the value of said fields.
Makes sense. It might be useful to score models based on how much they improve your predictions, e.g. Newtonian physics is not completely accurate but improves my predictions substantially vs not knowing any physics.
Reading popular books on the same subject have sharply decreasing marginal utility
I think there are so many separate subjects (given that you’re reading for fun, not for specific knowledge), that this really isn’t a problem.
What you said seems true for for sciences with well-ordered master theories. These obviously include physics, and to a lesser degree perhaps biology, astronomy or comp-sci. Once you’ve read a few really good books about popular cellular biology or popular evolution, there may not be much more to be gained at the popular level.
But in some areas of study, there are very many subjects which must popularized separately. History is the most obvious one—I’ve read dozens of popular history books, and will gladly read dozens more, because they’re dealing with different times and places, or with different themes such as economic history. Another example is zoology—the biology and ecology of particular organisms.
Meta: Should we be encouraging people to read popular books? Popular books have their place, but there is a common failure mode where people read popularizations, get their daily fix of insight porn, and go away without having learned anything substantial.
Reading popular books on the same subject have sharply decreasing marginal utility, and I’m guessing that the kind of person who reads LW is the kind of person who has already read enough popularizations that reading another is probably a waste of time.
I think a bigger problem with popular books is that they are often just wrong.
I agree with you about popular books. But I am wondering about the term “insight porn”—does it do anything besides put a negative connotation on the process of obtaining new insights? (If “insight porn” refers to useless insights, would it be appropriate to say something like “pure math is full of insight porn”? :P)
The way I’ve heard it used is to describe activities that give you the illusion of understanding a complex subject. Examples would be Malcolm Gladwell’s books, or the high school student who thinks they understand quantum mechanics after reading The Elegant Universe. So that usage wouldn’t fit pure maths, since we seem to generally agree that pure mathematicians have a true understanding of their fields, even if we don’t agree on the value of said fields.
Makes sense. It might be useful to score models based on how much they improve your predictions, e.g. Newtonian physics is not completely accurate but improves my predictions substantially vs not knowing any physics.
I think there are so many separate subjects (given that you’re reading for fun, not for specific knowledge), that this really isn’t a problem.
What you said seems true for for sciences with well-ordered master theories. These obviously include physics, and to a lesser degree perhaps biology, astronomy or comp-sci. Once you’ve read a few really good books about popular cellular biology or popular evolution, there may not be much more to be gained at the popular level.
But in some areas of study, there are very many subjects which must popularized separately. History is the most obvious one—I’ve read dozens of popular history books, and will gladly read dozens more, because they’re dealing with different times and places, or with different themes such as economic history. Another example is zoology—the biology and ecology of particular organisms.