4 ms·

With all due respect, what are 99 percent of people gaining from a book on, say, convex optimization? Yes, these are great choices if you're looking to become a

by Elepsis 17y ago

With all due respect, what are 99 percent of people gaining from a book on, say, convex optimization? Yes, these are great choices if you're looking to become an expert on, say, physics. But not everyone is a physicist, nor does everyone need to be. The point of the books on that list is that they are applicable to _many_ situations. You're not going to get oodles of "hard knowledge" from them, and I think people who go into them with that expectation are setting themselves up for disappointment. Rather, they present some anecdotes -- some interesting, and some not -- and say "Okay, now think about these, and see if any of them can affect the way you do things." If you read a business book and don't bother to do the thinking/applying afterwards, you are, indeed, wasting your time. The books on your list are ones that you go through, finish (though honestly, if you're just reading those books cover to cover and enjoying yourself, you're a better person than me... I'd use them as a reference at best) and can say "okay, I've learned something." The books on the original article's list are ones where once you've finished the book, you've only started learning.

TriinT 17y ago
"With all due respect, what are 99 percent of people gaining from a book on, say, convex optimization? Yes, these are great choices if you're looking to become an expert on, say, physics. But not everyone is a physicist, nor does everyone need to be." Convex Optimization for physicists??? Whaaaat? OK, consider Game Theory, Machine Learning, Operations Research. All have practical value and all require optimization. In fact, Machine Learning is mostly applied optimization theory. The Simplex algorithm is used all the time, in the most various industries. Yet once again, the need to know a bit of optimization theory is made evident. Physicists don't really require the kind of optimization on Boyd's book. Unless one is an experimental physicist and needs to analyze data sets. Theoretical physicists usually work with infinite-dimensional optimization, i.e., variational analysis. I entirely agree that not everyone is a physicist, nor does everyone need to be... but picking Convex Optimization as an example was a poor choice. The list is my personal list. It's a personal thing, and as such, it is non-transmissible. I am not trying to preach. I couldn't care less if the masses read junk...
- michael_nielsen 17y ago
  Convex optimization is a standard tool for theoretical physicists doing quantum information and quantum computing. And B & V is the standard reference. (Don't want to engage in the main debate, but I think this is an interesting tidbit of information. It's an example where a part of mathematics has recently become useful in an unexpected way.)
  - TriinT 17y ago
    Wow! I didn't know that! Thanks for the info. I know nothing on QIT / QC, but it's interesting and intriguing to see finite-dimensional optimization being used by physics people.
  - eugenejen 17y ago
    Sometimes I wonder the case is when we meet any new problem, we first try to fit it into any existing framework in order to understand it, until the framework is broken. This doesn't debase the beauty of math. But I just feel maybe we just need an excuse to fund mathematicians.
  - newsio 17y ago
    This is why I love Hacker News. Where else could I expect tp see this sort of debate in the course of casual browsing? (well, maybe Metafilter, but I'd have to pony up $5 to leave a comment)
mquander 17y ago
"With all due respect, what are 99 percent of people gaining from a book on, say, convex optimization?" The headline doesn't say "reading list for 99% of people." It says "Highly recommended reading list for those who love to learn." If you love to learn, then (I haven't read the book, so grain of salt) you're getting a well-written introduction to an interesting field of study.
timwiseman 17y ago
You have some good points, but many of those books on that list help people learn to truly think deeply, not just read and briefly consider anecdotes. It is that deep, rigorous, logical thought that is valuable to virtually everyone (and certainly everyone who claims to love to learn) that is more important than the contents of those books. Also, some of them (game theory in particular) are broadly applicable to and by most people who put the time and effort in to properly learning them.
dinkumthinkum 17y ago
Most of those books are from the Jibber Jabber industry with a couple exceptions, perhaps "Don't Make Me Think." I agree with another commenter that the title is about books for those who love learning not books that would be applicable to many situations if you what you need are vacuous anecdotes. Vacuous anecdotes don't build bridges, solve concurrency problems, enrich anyone's philosophical depth or anything like that.

by Elepsis 17y ago

TriinT 17y ago

"With all due respect, what are 99 percent of people gaining from a book on, say, convex optimization? Yes, these are great choices if you're looking to become an expert on, say, physics. But not everyone is a physicist, nor does everyone need to be." Convex Optimization for physicists??? Whaaaat? OK, consider Game Theory, Machine Learning, Operations Research. All have practical value and all require optimization. In fact, Machine Learning is mostly applied optimization theory. The Simplex algorithm is used all the time, in the most various industries. Yet once again, the need to know a bit of optimization theory is made evident. Physicists don't really require the kind of optimization on Boyd's book. Unless one is an experimental physicist and needs to analyze data sets. Theoretical physicists usually work with infinite-dimensional optimization, i.e., variational analysis. I entirely agree that not everyone is a physicist, nor does everyone need to be... but picking Convex Optimization as an example was a poor choice. The list is my personal list. It's a personal thing, and as such, it is non-transmissible. I am not trying to preach. I couldn't care less if the masses read junk...

michael_nielsen 17y ago
Convex optimization is a standard tool for theoretical physicists doing quantum information and quantum computing. And B & V is the standard reference. (Don't want to engage in the main debate, but I think this is an interesting tidbit of information. It's an example where a part of mathematics has recently become useful in an unexpected way.)
- TriinT 17y ago
  Wow! I didn't know that! Thanks for the info. I know nothing on QIT / QC, but it's interesting and intriguing to see finite-dimensional optimization being used by physics people.
- eugenejen 17y ago
  Sometimes I wonder the case is when we meet any new problem, we first try to fit it into any existing framework in order to understand it, until the framework is broken. This doesn't debase the beauty of math. But I just feel maybe we just need an excuse to fund mathematicians.
- newsio 17y ago
  This is why I love Hacker News. Where else could I expect tp see this sort of debate in the course of casual browsing? (well, maybe Metafilter, but I'd have to pony up $5 to leave a comment)

mquander 17y ago

"With all due respect, what are 99 percent of people gaining from a book on, say, convex optimization?" The headline doesn't say "reading list for 99% of people." It says "Highly recommended reading list for those who love to learn." If you love to learn, then (I haven't read the book, so grain of salt) you're getting a well-written introduction to an interesting field of study.

timwiseman 17y ago

You have some good points, but many of those books on that list help people learn to truly think deeply, not just read and briefly consider anecdotes. It is that deep, rigorous, logical thought that is valuable to virtually everyone (and certainly everyone who claims to love to learn) that is more important than the contents of those books. Also, some of them (game theory in particular) are broadly applicable to and by most people who put the time and effort in to properly learning them.

dinkumthinkum 17y ago

Most of those books are from the Jibber Jabber industry with a couple exceptions, perhaps "Don't Make Me Think." I agree with another commenter that the title is about books for those who love learning not books that would be applicable to many situations if you what you need are vacuous anecdotes. Vacuous anecdotes don't build bridges, solve concurrency problems, enrich anyone's philosophical depth or anything like that.