Extrinsic

Trailhead: Christof Kuhbandner, Alp Aslan, Kathrin Emmerdinger, and Kou Murayama. "Providing Extrinsic Reward for Test Performance Undermines Long-Term Memory Acquisition". Frontiers in Psychology (2016-02-01).

Listen. I will not belabor this point. (I would love to belabor this point, and I have, but I have also exercised the DEL button liberally so here we are.) The point is: optimizing your learning for an exam is trash.

But that's what much of your grades are based on, so you're kind of stuck, eh?

I'm not qualified to talk about grades anymore, anyway. When you're an undergraduate, grades have Meaning. They have outsized influence on the job you get or the graduate school you get into. Even if you think grades are trash, the odds are in favor of good grades. I think grades are often trash, but the odds—can't deny it, gotta have 'em.

But in my own life: I don't have much time for grades or performance management scores or any of the pseudo-objective measurements that we're regularly subjected to. It's not a tough stance I'm taking, it's just that I'm not interested. You can optimize your performance ("performance") for what gets the grade, and still not know how the thing you're doing works. An exam tests knowledge that can be easily packaged in an exam; performance management does the same, but for things that can be packaged in a PM form.

It's an incomplete model. There's more to life and work than what gets measured on an exam. Those scores are useful, and successful people can score high, but it's those with enough resilience to solve the problem—whatever the problem at hand happens to be—who are most valuable.

Don't confuse what's on the exam with what you need to know.

Pay no attention to Caesar. Caesar doesn't have the slightest idea what's really going on.

—Kurt Vonnegut. Cat's Cradle (1963).

Knowledge cube

There are three dimensions to knowledge when talking to someone. (Obviously it's just a simple model.)

  1. Actual knowledge: don't know—know
  2. Expected knowledge: ambivalent—should know
  3. Projected knowledge: projects lack of knowledge—projects knowledge

I don't have a complete model in my head, just a few thoughts about what to watch out for—good and bad.

Someone who rates high on all three—has knowledge, should have knowledge, and projects knowledge—is solid. Is not an interesting case, but this is the kind of person you want to talk to, no?

Someone who projects knowledge but doesn't have it is dangerous—especially when is expected that they should have that knowledge, because the others are more likely to believe them.

Being low on the expectations angle isn't bad—as long as projection is low.

Someone who knows but doesn't project is either the secret weapon—people who know will know that this is someone who knows—or a wasted opportunity, because more people ought to know this is someone worth talking to.

Six weeks, six months

My heuristic for improvement is: six weeks, six months.

Need to get in shape? (Whatever "in shape" might mean, but that's another topic.) It's going to take six weeks before you notice any changes, before any benefits are apparent. It's going to take six months before you're really building strength or endurance.

Six months or six weeks aren't real, scientific numbers—they're just heuristics. When you start some habit, it takes a solid starting chunk of time to lay the groundwork for improvement. When the habit is solid, it takes another, larger chunk of time to get the results of the habit to some satisfyingly high level. Six weeks and six months are close enough to capture that.


I am relatively out of running shape now (at least compared to past exploits). I was building it up this year, but the last few weeks didn't leave much time to maintain that. OK. Let's pretend it's a start from zero. How long is it going to take to solidify the habit? Six weeks. How long is it going to take to get in competition shape? Six months.

They're both daunting numbers. There aren't any shortcuts. You just have to keep doing it until you're better at doing it. There are matters of technique, coaching, learning, each of which have their place in improvement. But the real magic is buying into maintenance —relentlessly keeping at it until you get to the threshold where you can take on something more and learn to do it better and better.

It works

NASA’s Ingenuity Mars Helicopter Succeeds in Historic First Flight

There are two important aspects of Mule Time: (1) Head down, straight ahead to the end of the row, turn around, repeat; (2) stubbornness, bordering on insubordination.

Mule Time happens when there ain't nothin' left to plan. The plan happened. The doing-to-plan happened. But there's still some work left to do, maybe to a plan, maybe not. It doesn't matter. What matters is knowing when it's time to be a mule or when it's time to be a human and find a better way to get the job done. How do you know? You don't. It's a matter of taste or temperament.

I've written enough about automation, and I've done extensive planning for complicated projects. It's real. I believe in it. I use it. But when you've got a long, boring, grinding problem, you gotta get behind the mule.