Failing to Fail

March 22, 2025

After college, people experience a large drop off in how much they learn. And for a lot of people, this is a great loss. The reasons to learn are varied: some want to pick up a hobby and get good at it, others want to be better at their job, and yet others want to simply grow and develop as people. But despite the near-infinite store of knowledge at their fingertips, they struggle to move past the first search page. Why?

I contend there are two parts to learning: background and ability to navigate metacognitive burden. When people talk about their inability to learn more after college, they often ascribe the challenge to some lacking prerequisite knowledge. They open a physics textbook a few months after graduating and realize that they’ve forgotten how to calculate a derivative; they can’t hold a bar chord on the guitar and are forever doomed to play Riptide by Vance Joy. But this doesn’t seem to be the core of the problem. While knowing more things or having more skills before starting a new project is helpful, being able to play the B minor chord is not the biggest barrier to becoming Jimi Hendrix.

I think these are all problems of navigating metacognitive burden. The people in my life that can plow through a textbook and join a research project in a field they’ve never encountered are olympiad kids. Having been forced to accumulate a broad yet targeted store of knowledge, they found out how to pick through a textbook for the information they’re interested in, to prioritize their course of learning, and to not rely on practice problem solutions. Hidden behind this knowledge of how is hours of aimless wandering – wasting an evening reading an irrelevant chapter, trying a confusing memorization strategy, failing another practice exam. These negative experiences fleshes out the possibility space for them: they know not only what is good and what is bad for executing their goals, but why. And knowing why is the basis for knowing how. Years after the pressure for winning gold is gone, they can apply this knowledge to fields and experiences they’ve never touched before.

And this isn’t just about academic learning. It’s also about cooking for the first time; about sticking with a new year’s resolution; about making new friends; about developing a novel crochet pattern; about trying to get ChatGPT to highlight the difference in rows between two spreadsheets; about writing a thesis.

Execution is the art of directing attention and generalizing from failure. It’s noticing when you’re confused (a cognitive skill), managing your uncertainty (an emotional skill), and understanding what your failures imply (an experimental skill). Developing this art requires both previous experiences struggling with the frontier and the confidence that experience inspires. This latter component is the difficult part. To give confidence, the problem needs to be challenging enough to force failure and exploration, but easy enough to eventually solve. The correctly balanced problem forces you to experience uncertainty without letting it overwhelm you. But once you have that confidence, it generalizes to related tasks; you will know how to overcome the new set of metacognitive burdens imposed on you.

This implies an interesting tradeoff: Advice improves your ability to execute a given task, but mediates your ability to execute a similar task. After all, advice allows you to avoid inefficiencies but can’t teach you viscerally why the inefficiencies are inefficient; it removes uncertainty and productive failure. This is not to say that you should never seek advice. It’s just important to tune the amount of advice you seek to make the problem you face the right amount of challenging.

This is all to say: go! You are not being held back by your skills or knowledge or intelligence or natural ability. You are being held back by how little you have failed before. By how much more you will need to fail again.