A useful post from CFAR co-founder Andrew Critch’s personal blog describes how to get the most out of graduate school — especially if you want to maximize your positive impact on the world.
Sample tip (this is 1 of 8):
Find a flexible program. PhD programs in mathematics, statistics, philosophy, and theoretical computer science tend to give you a great deal of free time and flexibility, provided you can pass the various qualifying exams without too much studying. By contrast, sciences like biology and chemistry can require time-consuming laboratory work that you can’t always speed through by being clever.
Last month over 100 CFAR alumni from around the world converged on Asilomar State Beach, on the edge of Northern California, for our second annual reunion. And I know I wasn’t the only one whose frequency of exclaiming, “Whoa! Interesting. I hadn’t thought of that before…” was unusually high that weekend.
So I polled some alumni reunion alumni and asked them: What’s something you took away from the CFAR reunion weekend?
If you don’t have any Grand Ideas or Ambitious Plans, one of the best ways forward is simply to do ALL of the obvious, small things. “Shut up and do the impossible” is a great meme to have in the community, but all too often I find myself skipping over the step “shut up and do the possible,” which is almost as valuable and just as rarely done.
In the same way the knobs & dials on a device form a user-friendly interface that lets you modify the device’s internal parameters, good habit formation techniques form a user-friendly interface that lets you modify your behaviors. So don’t make your plans things that the virtuous superhero version of you would do, instead make them like good user interfaces in the sense that you don’t need to think hard about them for them to work.
If you can measure accomplishment by how much you’ve changed for the better, instead of how much you have left to change, the reward curve comes out the helpful way.
Instead of sitting down to write an entire essay, start out by writing an email or personal message. [Julia’s note: I’ve already used this, to good effect!]
I think I left with the distinctly new impression that not only are the individual technqiues we’ve learnt from CFAR useful, but that we can learn to piece them together into larger techniques. I went in thinking of them as complete units, (albeit ones where some had others as prerequisites) and came out thinking of them as an initial set of building blocks we can add to, and piece together to make other things.
I thought I was good at introspection (especially listening to system 1), being an introvert. But now I see that there is so much room to grow that skill, and that my lack of skill in that area was making some of my habit building and rationality interventions fail.
… and here’s one of mine:
When people within a community disagree with each other — for example, about ethics — they should limit themselves to using only persuasion tactics that are more effective for true ideas than false ones. For example, anyone could threaten social ostracization if you don’t adopt Policy X, but this tactic could be used indiscriminately regardless of whether Policy X is good or bad. Public betting and Aumanning, by contrast, are more effective when you’re right then when you’re wrong.
Bayes’ Rule is a theorem from probability theory about how to adjust your confidence in your beliefs, as you learn new information. It’s often featured in CFAR workshops.
In a recent video blog, Julia uses animations to explain how Bayes’ Rule works, and how she uses it in her everyday reasoning and decision making:
We spend a fair amount of time at CFAR workshops tackling “bugs” in our daily lives. But does learning to solve small problems really make a big difference?
Recent CFAR alum and Oxford student, Ben Albert Pace, posted this thoughtful discussion of the “debugging” mindset on his personal blog. Excerpt below; read the whole thing here.
One criticism of the utility of getting better at solving these, is that they are all small problems. This is an important criticism, that all of this ‘rationality’ is only marginally useful. The first counter-argument that might be offered, is that little problems build. Lots of little inefficiencies, from morning to night, can really add up over the course of a lifetime, losing you years of happiness and productivity. But I don’t think this is the strongest counter-argument. The fact is, dealing head on with the biggest problems in life requires the same skills as dealing with the small ones.
The mental move by which you try not to think about your dissatisfaction with the tidiness of your house, is the same mental move in which you try not to think about your dissatisfaction with the course that your career is taking. The dissatisfaction is of greater magnitude in the latter, but it is the same unhelpful skill of ‘not thinking about dissatisfaction’ that you are practicing. If you can’t do addition and multiplication, you can’t do research in pure mathematics, and if you can’t resolve the small problems in life, you will not be able to improve on the big, important ones. Bugs aren’t defined by the size of the problem, but by the cognitive algorithms that cause them to be problems.
So the concept of ‘bugs’ is really useful: once you’ve labelled something a bug, it is now in the category of ‘problems that I can practice solving to get better at life’. The staff helpfully emphasised this in classes, with talk about “Keep your eye on the ball” and “You are not here to learn the techniques, but to solve your problems, don’t forget that.”
Main take-away: If I have a problem in life, I think “Okay! Here is an opportunity for me to get better at life. Where’s my pen and paper?”
Ben’s original post here.
We’re back from running our first-ever rationality workshop in Boston. (Well, actually, a rustic retreat in Harvard, MA, about 50 minutes outside of Boston).
We usually try a few new things each time we run a workshop, while keeping most of our tried-and-true features constant. This workshop was no exception. For one, we had an unprecedentedly large cohort of participants: 38 instead of our typical 25, which seemed to work pretty well even though the classes were bigger than average.
But one of the most pleasantly surprising experiments this time around was our Day of Tutoring. Basically, we’ve come to suspect that the way people really learn a new idea — and the way we internalize that idea, to a degree that will shift our behavior — is to teach it to someone else.
So we spent half of Saturday having the participants tutor each other in the material from Thursday and Friday, in a chaotic-but-fun rotating wheel of tutor-pupil pairs. Throughout the exercise I overheard people noticing things they didn’t understand about the classes while trying to explain them; thinking up real life examples and applications; and putting our ideas in their own words, all in a high-bandwidth way that I haven’t really seen occur within the context of a traditional “teacher lecturing students” class before. (And nearly everyone rated the tutoring process as useful, after the fact — a higher approval rate than any of our standard classes.)
We’ll certainly be experimenting with more ways to help people learn by teaching. And in the meantime, why don’t you try it at home? Teach someone something today, and let us know what you learn.