Some things I've learned in college

post by aaronb50 · 2021-03-25T21:30:05.899Z · LW · GW · 1 comments

This is a link post for https://aaronbergman.substack.com/p/some-things-ive-learned-in-college

Contents

      Signalling
      Learning the wrong things
      Inefficiency: time
      Inefficiency: money
    for the post
  What I’ve Learned
    Year
      Ethnicity, Race and Nation
      Seeing the World Through Different Lenses
      Intro to Ethics
      Linear Algebra
    Year
      Intro to Proofs
      Religion and Secularism
      Intro to Econometrics
      Mind and World (with some influence from a later philosophy course called “What Am I?”)
    Year
      Data Visualization and Graphics
  Conclusion
      Footnotes
None
1 comment

In my corner of the internet, there is a healthy contrarian skepticism about education, and about college in particular. Here are some of the points or critiques of the learning that (perhaps) goes on in college, many of which I largely agree with:

Signalling

(i.e. any learning that occurs is minimal or useless, as it is merely a side effect of school’s primary function. I agree about 75%)

Bryan Caplan’s The Case Against Education indeed makes a convincing case against education: school is about signaling, not learning things. Doing well in college signals to employers or graduate admissions departments that you are intelligent, conscientious, and otherwise capable as functioning in as a knowledge worker in modern society.

To some extent, this can be an important function of college since employers and other institutions do need a way of figuring out who is competent. However, the cost is that a huge proportion young adult’s lives are spent “learning” things that they will rapidly forget and/or never use.

Learning the wrong things

(I agree about 30%)

Maybe people are learning things, but they’re useless things. This is largely the old “What’s the difference between a philosophy degree and a large pizza? The pizza can feed a family of four” or “We don’t need art history/gender studies/classics majors, we need engineers!” or “Go to trade school!” critiques.

Inefficiency: time

(I agree with this about 50%)

Instead of spending dozens of hours a week for four years going to class and completing assigned coursework, smart and driven students could learn the same amount that they actually learn in a fraction of the time.

Inefficiency: money

(I agree with this about 99%)

Instead of spending $200,000+, smart and driven students could learn the same amount that they actually learn with a tiny fraction of the money. I can’t find the source for this, but I distinctly remember reading that, for a fraction of college tuition, you could hire a team of PhDs to personally tutor you in their subject of expertise for the amount of time students generally spend in class (edit: this is at least one of the sources). And, of course, many could learn just as well with a few textbooks and cheap or free online courses.


Inspiration for the post

As my “I agree with this X%” statements indicate, I too am skeptical of the direct educational value of college. To Brian Caplan’s credit, his book does concede that around 20% of the value of education comes from “human capital accumulation” rather than signaling.

Recently, though, I’ve read part of the very wholesome “college is (or can be) good actually (even if elite schools suck right now)” book Excellent Sheep. I have my issues with the book, and might even write a review at some point, but it does provide an interesting contrast to some of the contrarian takes of the internet-intelligentsia.

So, I started thinking a little more about what, if anything, I’ve actually learned in my first 2.75ish years of college. Indeed, I think I’ve taken a bit too much of the “college coursework provides little or no intrinsic or direct educational value” red pill, rounding this 20% down to ~0% and assuming away the intrinsic value of any learning not captured by “human capital accumulation” Upon reflection, I think that I actually have learned few things that are one or more of the following:

  1. Likely to be useful in my career.
  2. Currently or likely to be useful in normal life.
  3. Intrinsically valuable to me by giving me interesting ideas, perspectives, or knowledge.

Importantly, I’m also claiming that I would not have learned these things if it were not for college, although of course this is impossible to know for certain.

Without further ado, here are a selection of classes and what I think they taught me, along with some other commentary on the courses. Indeed, I’m not listing every class—not even all those in which I learned something—both because that would be boring and long. The following are all recalled directly from memory, without checking notes or assignments of any sort. That would be cheating!

What I’ve Learned

Freshman Year

Ethnicity, Race and Nation

Seeing the World Through Different Lenses

Intro to Ethics

Linear Algebra

Sophomore Year

Intro to Proofs

Religion and Secularism

Intro to Econometrics

Mind and World (with some influence from a later philosophy course called “What Am I?”)

Junior Year

Data Visualization and Graphics

Conclusion

None of this is much evidence against Brian Caplan’s hypothesis or any of the other three critiques I described before. I have no doubt that much of this (with the probable exception of Stata and R) doesn’t really contribute to my “human capital” in the labor market; even the few things I find intrinsically interesting aren’t of much practical use, and the whole shebang is definitely massive waste of money (from a social perspective; it seems individually rational to dish out massive sums to get the college stamp of approval).

Even still, I thought this was worth writing up because, for all the higher level commentary about the utility of college and other forms of education, there seems surprisingly little discussion about the actual, object-level things that people learn in school. After all, this is the purpose of education—ostensibly, anyway.

Footnotes

1 Short version of my exasperation: everyone seems to assume that a given ethical decision comes with an unambiguous “maxim” on a silver platter, but any action can be formulated as following any of numerous maxims. Also, I am sure that many brilliant people have addressed this concern in the literature, but so far have been too lazy to dig in.

2 If you count making a presentation as a research assistant as “real life”

3 I tried earnestly to convince my professor of this, but he remained unconvinced that a realism could be true if God does not exist. My argument and intuition centers around a thought experiment:

Push a button, and you make a person suddenly and permanently experience excruciating pain. Nothing else about the world changes. I have a strong intuition that suffering is intrinsically—even tautologically—bad. Therefore, how could the world with this suffering person not be objectively worse than the alternative? If so, it would strongly seem that pushing the button is an objectively “bad” thing to do.

1 comments

Comments sorted by top scores.

comment by Gerald Monroe (gerald-monroe) · 2021-03-26T07:39:26.420Z · LW(p) · GW(p)

Another topic is that many students take the wrong degrees to be useful at all.
I majored in computer engineering.

My school covered various theories about how a computer in general is usually architected.  How a specific example computer's bytecode operates.  Various labs and assignments where you had to write small programs in assembler and a half dozen languages.  How to analyze where the slow step probably is.  How to translate a truth table to boolean expressions to logic gates.  

Ironically, pure math procedures like "truth table to boolean" is the most useless because a software tool will always exist I can just get to do that, for as long as I am employed.  

I would say that all that linear algebra stuff they had you do will turn out to be similarly useless, they should have spent the time having you write programs in python to solve the problems. (or JUST teaching the one structuring of the problem step that scipy/numpy doesn't do for you, etc)

Some useless courses in topics like theater were required as part of the 'core', hey I got to see a live play but yes, there went a tuition payment and some hours of my life. 

But in general I would say about 2/3 of the topics are used somewhere in my actual field, and maybe 1/3 of the knowledge I have definitely used as an engineer.  Even if it ended up being just a starting point.

Could a set of compressed 'boot camp' like courses have gotten me almost as good for a fraction of the time and money?  Maybe.  I would want to see data on this, not just average joe opinions on whether or not these various tech skill boot camps work or not.  

Certainly a more efficient structure exists.  If as a society we actually cared we could shorten college and make it optimize towards quantifiable outcomes.  Though one issue is that we don't know what those are.  It may be that in my career everything I learned stops mattering because we develop an AI agent that can architect computers better than any human alive.  

Similarly, 30 years ago electrical engineering was obviously superior to being a programmer but today it seems that most of the jobs are in software, and most of the $.