looking for gold

i suppose this is a continuation of the previous post

at my job we do a lot of brainstorming and prototyping, and L often reminds me to look for gold - to try my best to uncover the value hidden in new ideas before thinking about their flaws. it is the kind of simple request that is unreasonably hard for me, in part because i have not had to think in this way before and in part because i am very opinionated about concerns like user experience and practicality and self-consistency, and fresh ideas are usually lacking in one of these categories

a few months ago i wrote about how there are two dominant modes of listening, understanding and evaluating. i mentioned this in the context of nonviolent communication, but i’ve since realized this distinction is present in many other areas. my coworkers love that i am great at evaluation and providing honest feedback and do it essentially 24/7 but i think this is also one of the traits holding me back as a researcher. most ideas in their infancy are badly formed and not being able to wholeheartedly engage with bad ideas makes it difficult to discover other ideas. or, to phrase this differently - reward signals are too sparse to be picky about which ones to listen to

i thought it would be easier to switch freely between understanding and evaluating, and i still believe this is a worthwhile personal goal, but it’s proven to be more elusive than i imagined. to be more specific: once i made it an explicit priority i actually found it somewhat easy to be less judgemental in social contexts but basically impossible in technical contexts. probably this is the result of spending eight years of my life on math contests - you become excellent at a very constrained and short-term type of decision-making problem - which of these two techniques should i use? which approaches are too complicated to work? which approaches don’t have enough power behind them to solve the hardest version of the problem? it is a very persistent bias in how i interpret the world which i did not appreciate the extent of until recent conversations with L. of course it’s good to be able to reason in this manner; as with every other skill the risk is that if you’re very fluent you can get far enough with one type of reasoning that you never develop the ability to reason in other ways

there are many problems in my life but one of the more frequently-recurring ones can be summarized as: i am good at choosing between goals and good at executing on goals but terrible at proposing new goals. with my job this means i naturally get funneled towards solving hard engineering problems which enable my teammates to run their experiments. with clothes this means i have a lot of opinions on why different outfits look bad but no real path towards assembling outfits i like. all this reminds me of that miyazaki interview: When young, nearly all of us want to be taken seriously, as soon as possible. Perhaps because of this we tend to overemphasize technique… There are innumerable examples of people making films with a very high level of technique, but only a very fuzzy idea of what they really want to say. And after watching their films, viewers are usually completely befuddled. Yet when people who know what they want to say make films with a low level of technique, we still greatly appreciate the films because there is really something to them

recently i’ve realized that i enjoy getting negative results in research. part of this is because i am lazy and negative results (this method definitely won’t succeed) mean there’s less work to do than positive results (this method might succeed if we implement it better). another part of it is because i like having confidence in the choices i am making and it’s much easier to reject a hypothesis with high confidence than it is to accept it. when i told this to L they pointed out that confidence is not really the right metric to be pursuing because opportunity is usually improbable. pruning bad options is a helpful skill to have in the initial phase of exploration but when you’ve actually found something good you won’t need to prune any longer because it will be obvious what to do

on a related note i’ve been thinking about how many decision-making problems initially present themselves as tradeoffs: partner intelligence vs attractiveness vs personality, product feature speed vs learnability vs capabilities, job salary vs impact vs fulfillment, and so on. of course tradeoffs exist in the sense that there is always a pareto front, but i think we tend to exaggerate the degree to which you need to balance different considerations. eg. nowadays i think my friends are very smart and cute and kind people! though admittedly it took quite a few years to find them. what i am really trying to say is that with these kinds of problems the first few data points you get often give off the impression of strong negative correlations and needing to consider tradeoffs but that’s just because you need a stronger sense of self / taste / direction before you can find options that satisfy all your requirements. and the issue with being purely evaluative is that you examine the early data and see the tradeoffs and begin optimizing along the tradeoff (eg. “i’m going to pick these two categories to focus on and ignore the third”) and miss out on things that would’ve made you very happy as a result. unfortunately i think this is sort of the story of my life - a tendency to give up early and choose between things i love instead of trying harder to choose all of them. (i am working on it)