Advice on advice

Advice is a dangerous thing. You most need it when you’re most naive, which just happens to be when you’re least able to distinguish between the bad and the good. And there’s no shortage of terrible advice out there: stuff that’s well-meaning but idiotic, stuff that’s communicated badly, and the occasional actively malicious advice. But I think the pitfall that people are most prone to is taking advice meant for someone else.

The thing is, figuring out what’s good for you in particular requires self-awareness, and self-awareness is hard. Not the pedestrian, moment-to-moment sort that judges some relevant trait against other people in your immediate vicinity, we’re quite good at that. What’s hard is taking millions of those little judgments made all through your life, against all manner of people, and integrating them into a coherent self-concept the describes your position on a curve with the rest of humanity. And knowing how that trait interacts with other traits, with different environments, with different people.

Your brain does some of this for you automatically, and you probably have an intuitive “feel” for what you can do and where you are in many domains. Often in these cases you can “see” your way through problems that are beyond you by watching others tackle them.

But advice is a step removed from implementation–it’s harder to “feel” your way around, to push off of where you know your experience ends and grab just beyond it. Instead, advice asks you to leap almost blindly in the direction it suggests. In many cases of advice-taking the stakes are relatively low, so trial-and-error isn’t a catastrophic method to adopt. However, a more efficient strategy will probably require a representation of your starting point that you can manipulate symbolically, which is a higher bar than your intuition can clear on its own.

An example: humans are great at ranking ourselves amongst our peers by intelligence, but we’re terrible at predicting our own IQ scores.(1) Unfortunately, certain career paths are quite dependent on absolute intelligence level. The smartest kid at a tiny rural high school might have good prospects overall compared to his friends, but that doesn’t mean he’s cut out to be an astrophysicist. Similarly, the dumbest kid at a very competitive school may drop out to become an electrician and find that he’s made a huge mistake. Other fun consequences of taking advice meant for people smarter or dumber than you include crushing student loan debt you’ll never pay off, wallowing in remorse over opportunities you weren’t confident enough to take, and children.

So if you want to take advice in an informed way, how should you go about learning your relevant properties? Try considering the domain you’re working in, and ask yourself how the people in it vary. Are there many different approaches represented, or just a few? Are there distinct groups that you can see? Take any of these styles or abilities that you can identify and find out if there’s a name for them (there probably is). Alternately, work backwards and reason about what might be required for the task at hand. Once you have a handle on where your strengths and weaknesses lie, you can start interrogating advice for its suitability: whether it assumes skills you lack, whether it will address your address your weak spots, whether you’re a representative member of the audience it targets.

Often major constructs with psychometric measures attached to them (Big Five, IQ) account for some amount of individual differences in outcomes, and these have the advantage of giving you a quantitative sense of where you are versus everyone else. Psychiatric diagnoses, especially those with cognitive components (autism, ADHD, OCD, etc) can also be helpful to bear in mind. But a great many things are not spelled out so neatly. Sometimes an important concept lacks a name, or is actually two concepts. Often there are no official statistics to dredge up. In such cases, you may want to strike out on your own and name a concept or two, which allows you to not only reason about it more explicitly, but also to discuss it with others.


(1) I think this was the actual (frequently misrepresented) result from Dunning-Kruger, but there’s a chance it was from elsewhere and I’m too lazy to look right now. If anyone can refute that and is itching for a cite, let me know.

Advertisements