Eric Drexler on Learning About Everything

30 Post author: Vladimir_Nesov 27 May 2009 12:57PM

Related to: The Simple Math of Everything, Your Strength as a Rationalist, Teaching the Unteachable.

Eric Drexler wrote a couple of articles on the importance and methods of obtaining interdisciplinary knowledge:

Note that the title above isn't "how to learn everything", but "how to learn about everything". The distinction I have in mind is between knowing the inside of a topic in deep detail — many facts and problem-solving skills — and knowing the structure and context of a topic: essential facts, what problems can be solved by the skilled, and how the topic fits with others.

This knowledge isn't superficial in a survey-course sense: It is about both deep structure and practical applications. Knowing about, in this sense, is crucial to understanding a new problem and what must be learned in more depth in order to solve it.

This topic was discussed intermittently on Overcoming Bias. Basic understanding of many fields allows to recognize how well-understood by science a problem is and to see its place in the structure of scientific knowledge; to develop better intuitive grasp on what's possible and what's not; and to adequately perceive the natural world.

The advice he gives for obtaining general knowledge feels right, even for studying the topics that you intend to eventually understand in depth:

Don't drop a subject because you know you'd fail a test — instead, read other half-understandable journals and textbooks to accumulate vocabulary, perspective, and context.

Comments (13)

Comment author: JoeShipley 28 May 2009 01:27:56AM -2 points [-]

I think it's all too typical in geek culture for someone to consider just one topic worthy enough of his or her intellect. I run into this sort of person a lot... any given programming convention for example, but it's certainly not everybody. Still, a lot of people scoff;

"Philosophy? It's all total bs, who knows the answers to that stuff anyway?" "Literature is for english majors. Don't make me gag." "Economics is guesswork, at least programming follows defined rules for sure" "Physics and chemistry is for newbs, biology is where it's at."

One field that gets disregarded repeatedly is feminism or women's studies. Lots of geeks want to look at it like a solved problem, but anybody who has worked in the industry knows the ridiculous sexism that continues to pop up without the geeks-in-charge even noticing it. Understanding why these issues are important helps increase your total understanding and helps you tackle more difficult problems.

Interdisciplinary understanding at least some basic points in many different fields gives you more than just a hammer in your toolbox to handle problems that aren't nails. I'd agree that it's essential to solving the Big Problems. The payoff of specialization in things like agriculture and industry is obvious. With difficult problems requiring many different fields of knowledge, the clarity and bandwidth of your thoughts you can convey from one specialist in one side of the problem to another specialist in another side of the problem drops to nil without some basic understanding on all sides.

Comment author: HughRistik 29 May 2009 05:11:26AM *  15 points [-]

One field that gets disregarded repeatedly is feminism or women's studies. Lots of geeks want to look at it like a solved problem, but anybody who has worked in the industry knows the ridiculous sexism that continues to pop up without the geeks-in-charge even noticing it.

It's true that feminists make some correct empirical and moral claims that are prematurely discarded. Yet this mistake doesn't mean that Women's Studies isn't rightly looked down on as a real academic field.

I've taken Women's Studies classes at a top university. Here's a quote from my Feminism 101 syllabus:

This course embarks from a few key feminist assumptions: women’s and men’s lives are thoroughly gendered, gendered dynamics of power and inequality are reproduced in and through other forms of difference (class, race, ethnicity, sexuality, nationality, religion, disability and so on), and such social inequality is unjust.

Whoa there! Instead of explaining and justifying the foundations of a discipline, this course simply "assumed" them, and "embarked" from there. And admits it! (not that there weren't many other assumptions in the course that weren't admitted) This sentence has at least 4 loaded terms: gender, power, inequality, and unjust. Feminists are constantly throwing around terms like this, and I was taking this course to try to figure out what they mean (along with "patriarchy," "oppression," etc...). Unfortunately, I was disappointed: no real analysis occurred for the latter three. "Gender," was discussed, but from a muddled anti-scientific social constructionist perspective based on the work of Judith Butler and Anne Fausto-Sterling (who dubiously believes that sex, i.e. classification of people as male and female, is also socially constructed).

The epistemic standards of feminist theory are horribly bad. This doesn't make feminism completely wrong; I actually agree with the sentence I quote from the syllabus (based on my own conceptualization of those terms, no thanks to Feminism 101) with the stipulation that "inequality" often disfavors men, not just women as assumed in the course. Many of the moral claims of feminism are correct, even when they are based on shoddy reasoning.

Some of the claims of feminism are so lacking in rigor that they aren't even wrong: for example, the typical view of academic feminists that women are "oppressed" and men are not (and if it is granted that men can be oppressed, women are still oppressed "more"). You can't evaluate the truth of this claim any more than you can say whether an oak tree is "bigger" than a pine tree: it depends on what you mean by "bigger" (height? width? mass? surface area?).

Not only does feminism contain a high concentration of thought gone wrong, but an example of its bad epistemic standards is its lack of quality control. Relatively rational feminists are notoriously bad at criticizing the thought-gone-wrong of other feminists. Mary Daly wouldn't mind if men were wiped off the face of the earth(I'm not kidding; read the entire interview and see if you can figure out what is wrong with that thought process).

You would think that other feminists would condemn Daly for giving feminism a bad name and try avoid being associated with her. Yet despite attracting some incidental criticism, Daly is popular enough that she has been invited to speak at about 12% of North American Universities over the past few decades.

There are many problems with feminist thought, in and out of academia. For a more comprehensive treatment, see Daphne Patai's Professing Feminism: Education and Indoctrination in Women's Studies and Nathanson and Young's Spreading Misandry and Legalizing Misandry. While my personal experience in Women's Studies was not as bad as some of the horror stories Patai describes, it did show me that feminists don't have rigorous reasoning (or often, any reasoning) behind feminist theory, and that it is a morass of articles of faith, self-serving arguments, circular reasoning, and already-falsified hypotheses. These problems make it questionable as an academic discipline, despite getting some moral and empirical arguments right.

Comment author: Ttochpej 28 May 2009 02:39:20PM *  0 points [-]

I am completely new to the lesswrong site but have read a few of the blogs and think that I could learn a lot from this site. I think it is a good thing to have a basic knowledge of many fields, but I also think that it is important to have a in depth knowledge in the basic principles of science first. For example a person who has a broad knowledge in witch craft and taro cards could be very smart with what they know, but all they really know is fiction and without first having an in depth knowledge of scientific principle to test what they learn against, other wise people can be lead to believe anything.

Comment author: Matt_Simpson 27 May 2009 05:30:02PM 5 points [-]

speaking of the Simple Math of Everything, that sounds like a great Less Wrong community project, no? I have no idea what the money from sales would be used for - maybe donations to SIAI or something.

Comment author: XFrequentist 31 May 2009 04:39:03PM 0 points [-]

Agreed, fully!

This should be discussed in a top level post. Such a document would be a valuable contribution to the intellectual world.

A wikibook might be a good medium for this kind of community project. The LaTeX book gives an example of how well these can turn out.

Comment author: tony_powers 27 May 2009 05:18:34PM *  1 point [-]

Yes. This is very good stuff. Interdisciplinary knowledge is becoming more important as the man made world gets more complex.

My fear is that some people just aren't wired for it. I personally get a healthy dopamine jolt when I cross connect ideas. I suspect I'm in the minority though. It seems depressingly unrelated to education level as well. I work with people that currently have many more years of schooling than me, but are intellectually stagnant and don't care. I know a post doc in astrophysics that believes in reiki and essential oils. She has a narrow knowledge band and no BS filter

Most people probably require different incentives. Like health problems leading you to study medical topics etc...But what if the medical topics you study lead you to reiki?

Is a robust BS filter a prerequisite to interdisciplinary learning? Or will the process of learning improve your BS filter. I suspect it's both. Though, the more people I know, the more I push toward genetic determinism. I'm hovering around 25/75% right now. The i's currently favoring determinism.

Comment author: MichaelBishop 27 May 2009 05:42:22PM 0 points [-]

While it is sometimes useful to say something like genetic variation explains 50% of the variance in trait X, this may often be misleading because of how genes interact with the environment and with each other.

A trait may be immutable at age ten or twenty which was very much undetermined at age two.

Comment author: Kaj_Sotala 28 May 2009 04:32:33PM 1 point [-]

Not to mention that heredity is always defined relative to a specific population being measured.

(For instance, giving everybody in a certain area access to high-quality education will reduce environmental variation and thus increase the role that genetic variation plays, pushing up the measured heredity of educational achievement.)

Comment author: steven0461 27 May 2009 01:27:46PM *  4 points [-]

Good stuff, upvoted. Related point: since learning can be far more efficient if you're intrinsically curious about the subject matter, you should collect as many "angles" as possible to interest you in new bodies of knowledge, and cherish any connection you can find between things you care about on a gut level and fields you might benefit from understanding for extrinsic reasons.

Comment author: astray 27 May 2009 05:53:09PM 0 points [-]

These are good reads; I was going to post these links later today, after I had time to write up a summary, but you have saved me the trouble.

Notice that concepts make more sense when you revisit a topic, and note which topics provide keys to many others.

I realized when reading this that I have largely been following this method for computer science. Even without any obvious gears clicking into place, I understand talk about, e.g., binary trees or closures that would have baffled me a year ago.