I’ve written about exploration and exploitation before, but I realized recently that this may be more important than I thought. Talking to a friend about what different people thought about what role effort vs. innate ability played in success, I went down my list of successes and ended up realizing that the only “difficult” part (in the sense of “part I was closest to not doing”) was actually getting up and doing things outside my normal routine.

For example, here’s a list of the things that have led to my biggest wins so far:

I’d say these were about 25% effort, 25% ability, and 50% “luck” from doing lots of things outside my normal experience.

At a guess, the reason doing lots of different things is so important is that it simply exposes you to more potential random “strokes of luck”. When you do a thing, often lots of the exposure to luck comes from the first relatively small bit of effort you put in. For instance, if I turned out not to like contra dancing, I could have stopped going, so it was a pretty small cost for a large potential upside.

This stuff compounds, too. The more you do stuff, the more “luck” you get, which helps you do more stuff, etc. In other words, it seems like the main factor in success is a preferential attachment process, which explains why there’s such a power-law distribution and why similar people often end up with such wildly different outcomes: relatively small differences in initial conditions, caused by noise, can blow up to huge effects later.

Notes on this theory:

  1. Some predictions: people who experience more different things will ceteris paribus succeed more. For example, for college graduates, life outcomes later will be correlated with diversity of the things they did over their summers. Success might be correlated with number of different places you lived growing up (there are lots of confounders here of course). A supporting piece of evidence here is Satvik’s assertion to me that “Statistically, CEOs of large companies tend to have worked in multiple areas during their early career at a rate far higher than baseline.”
  2. I’d guess that the preferential attachment process starts gaining significance almost as soon as the person starts getting agency, which can be as early as, I don’t know, three or four? There’s a famous story about Elon Musk when he was maybe six years old and loved to go visit his cousin who lived across town. One day he was grounded and his parents wouldn’t take him to visit his cousin, so he escaped and wandered around the town himself until he found his cousin’s house. That’s the kind of doing-things-ness that can compound ridiculously.
  3. I think I’m probably better (comparatively) at Actually Doing Things than I am at putting in effort, but that even still, the returns from spending more attention on Actually Doing Things are probably higher than spending that attention on putting in effort. Unfortunately, this is hard to convince myself of in a way that actually compels me to act, for reasons I’ve mentioned previously.
  4. If this theory is true, it has some implications for my behavior: I should be spending less time on classes, taking more variety of classes (this will be easier in the future as I’ve basically finished my major), try to get rid of my most time-consuming commitments (e.g. get someone else to run HHIP next year), and generally say “yes” to more things. The college Harvard’s environment is set up pretty poorly for this, inasmuch as I have to spend 20-30 hours a week thinking about math.