Meta-advice

Good advice is always contextual.

There is no such thing as a piece of advice that is good for everyone. No matter how pure you think your ideology may be, there is someone out there who has taken it too far and needs to chill. No single principle will lead you unerringly. Wisdom means finding a balance between good things.

Some people need to learn the value of hard work. Other people need to learn not to invest their entire lives in their job.

Some people need to learn to be less materialistic and more spiritual. Other people need to learn about the world that exists outside their church.

Some people need to be less cynical and allow for the possiblity of good. Other people need to be less idealistic and engage with reality.

Be wary of untargeted advice. The best advice comes from wise people who know the specific situation they’re speaking into. Barring that, it falls on a wise listener to determine whether a piece of advice applies well to their own context. This is harder than it sounds.

One of the known bugs in human psychology is that we overemphasize inputs we already agree with. If an optimist and a pessimist each hear a piece of good news and a piece of bad news, the optimist will become more optimistic and the pessimist will become more cynical. This means that issuing untargeted advice may actually do more harm than good, because the people most likely to follow it are the people who need to hear it the least and vice versa.

If you tell a thousand people that the key to success is hard work, the slackers will ignore you, and the ones who are already killing themselves will work even harder.

If you tell a thousand people that God can forgive any sin, the ones already wracked with guilt will assume you mean any sin but theirs, and the ones who are molesting children will nod and say “Amen!”

Getting good results from flawed materials, part 1 of infinity

On the receiving end, how much value you can get out of untargeted advice depends on how good you are at countering your own biases, a skill that’s notoriously difficult to self-evaluate.¹ The best you can really hope for in this regard is to notice your mistakes and reduce them over time. A better choice is to get advice targeted to your situation (or a similar one), ideally from people with a track record for making good choices in a similar context to yours.

TL;DR: Advice becomes less accurate the more broadly it’s given. Making broad advice useful is hard, and guessing whether it will make you in particular better or worse off is harder. Target your advice narrowly, and stop looking for a unifying Theory of Everything.

1. Developing this skill is a central goal of the Rationalist community, which I’ve been peripherally following for several years. The writing of Scott Alexander was my intro point, but I particularly recommend Harry Potter and the Methods of Rationality as an entertaining introduction. ↩