Amplifying a point from Dan Gross
Dan Gross, during Shruti Rajagopalan's excellent Talent interview, made a point worth amplifying:
GROSS: We’re certainly not saying that horsepower and IQ don’t matter. They certainly do, but there is a clearing bar at which point, for most roles, I think people tend to overvalue it and don’t realize the logarithmic nature of the curve and assume it’s linear. But I actually think, earlier than many people realize, you want to switch to caring about just how vigorous and how energetic that person is.
I think that might be because--I’d be curious if you’d agree that it’s because, for most tasks, doing gleans far more information than thinking.
TYLER COWEN: Exactly.
The point mostly speaks for itself, but here are a few notes:
- When I worked in machine learning every day, I found a close analogue of this to be very true. A slightly better architecture tended to matter much less than getting better data. Or, at least, once I hit on a reasonable model architecture, I tended to do much better by cleaning and gathering more data than by optimizing the architecture.
- I was reminded of this by yesterday's issue of Byrne Hobart's newsletter, specifically the bit about forecasting: "The market message is 'don't forecast'!" Byrne gives what I think of as a variant of the Gross point: successful people need a good sense of what's going to happen, but spend a lot more time doing other stuff. (Or: "...spend a lot more time doing stuff.")
- It's worth remembering that Gross is a startup investor, and startup investment notoriously involves the tricky balance of finding people who are (i) solid and smart, with a clear-headed and expert view of some domain, but also (ii) ahem, "optimistic" enough to be doing a startup. He more than most people needs to be thinking about the tradeoffs between different kinds of analytic and practical apititude.
- Relatedly, there are kinds of discipline and hustle that are in many situations a good substitute for raw intelligence. Math competitions, programming interviews, and the like certainly reward studying. Reasoning under uncertainty benefits from discipline, and careful thinkers willing to do some research often do much better than top raw intellects. (This came to mind while I was writing up this post about poker and epistemic mistakes.)