it's hard not to learn from experience: Nonlinear Function
Created: July 10, 2020
Modified: February 25, 2022

it's hard not to learn from experience

This page is from my personal notes, and has not been specifically reviewed for public consumption. It might be incomplete, wrong, outdated, or stupid. Caveat lector.

It can be very hard to hold onto a positive self-image and an optimistic worldview, even if you intellectually 'know' these to be the right way to think.

When I was young, I believed that I had immense potential and could accomplish great things. Nothing in my life so far has disproven this, exactly, and it's a very powerful and helpful self-image. Nonetheless I've found it very hard to maintain in the face of (hopefully temporary) setbacks.

My experience in grad school 'taught' me that I'm not smart enough to do good work, and not socially fluent enough to fit in to mainstream society. Intellectually, I know those are bad lessons to learn, and that in many ways I was just unlucky. If you roll a die once and get a low value, you shouldn't immediately conclude that your die is loaded to lose. It's true that the low roll provides some Bayesian evidence that low rolls might be more likely than high rolls, but it's very weak evidence. It shouldn't lead you into any kind of certain belief.It may be evolutionarily advantageous to amplify weak signals: making a bunch of decisions each of which has 51% probability of being right can give you an edge, in aggregate. But the cost to you as a person in not believing in yourself is way higher than the cost to your genes (which are shared with many people), so even if this logic were sound on an evolutionary level, it would still lead to behavior that is individually irrational.

Still, it's hard to maintain a contrary belief in the absence of evidence. It was one thing, when I was young, to believe that I had immense potential. And if my experience had demonstrated that---I had quickly grasped new ideas, joined intellectual communities, and made serious research contributions that led to academic respect and new opportunities, then I would have the empirical basis for continued self-confidence and an expanded sense of the possible. But I didn't have those experiences.

A belief isn't just an abstract thing that we can change at will. One can infer beliefs from actions, and in a sense, a belief about yourself is a model of your policies. It reflects how you know you'll act in different situations: how you approach problems, how you engage in conversations, what tasks you'll take on and which ones you won't. These policies are themselves learned from experience---many subconsciously, as most learning is by demonstration---and you can't just make them different by wishing you'd had different experiences.

Belief in our own capabilities is also reinforced by social bonds and relationships. A lot of our 'learning' is held externally in the minds of people we are close to. If you've succeeded at building relationships with people you respect, their support can help you believe in yourself. If you haven't, then you don't have that advantage.

Put really generally: a bit of bad luck can lead to a genuine lack of growth, in skills and relationships and achievements. It's irrational to generalize from small bits of bad luck. But the holes left by the bad luck are real.

That said, everyone experiences bad luck at some point. Learning to overcome bad luck is an important life skill. All the things above are true, but truth is a low bar. Most true things aren't relevant. You still need to learn not to overgeneralize or wallow in despair.