Harnessing Your Biases

You're not harnessing your biases, you're harnessing your pride. So do I, in various ways.

In all frankness... I would have voted up your post except for the title. I'm fatigued of such clevernesses.

Exploiting your own biases is a topic I've been wanting to write about for a while, although I've thought about it more in terms of using biases (and other unpleasant aspects of our personality) to at least partially counteract each other.

For example, if you know from experience that your initial estimate of how long a software project will take -- which you are often forced to give, even if informally, before you've had time to think properly -- is usually far too low (planning fallacy), make your initial estimate X times your real estimate. Even if you don't really believe it will take that long, the anchoring effect ensures that your final estimate will probably be higher (and thus more accurate) than if you had chosen your initial estimate 'more rationally' (not really, of course, but that's how it feels).

We are all familiar with this particular example -- and I'm sure many of us in the software field have used it repeatedly -- but the general principle of using one bias to counteract another seems more widely applicable. As long as we still have these biases, why not try to put them to positive use and to use them against each other?

I don't think that every bias can be counteracted with some other specific bias, but in a more general sense, we can also often use what we consider weaknesses (e.g., intense emotions interfering with rational thought during an argument) to re-structure or change how we think about decisions or tasks so that the 'rational' decision/behavior feels like it is the emotional or ego-gratifying one that takes no effort. I personally have found that merely thinking about certain things as (and engaging in them as if they were) games or optimization problems can be very helpful in getting myself out of habitual patterns of behavior, switching from a state that is mostly automatic and system1-like into a state that is more reflective and strategic and system2-like. What makes the alternate state so powerful is that by viewing it as a game, my strong ego and intense sense of competition compels me to put great effort into winning the game.

Although it's interesting to ask whether talking in a group about something you read in a possibly-wrong article doesn't provide opportunities for people with more expertise than you to disseminate their knowledge. And for people with worse epistemologies to insist that they have more expertise than you and are disseminating their knowledge.

Certainly I find out a lot more about all the things I classify as "probably true" by talking to other people who have a different set of "probably trues" on the topic than I do by looking up as much information as I can find about each possibly-true. Do you gain more (or contribute more to the world) by classifying everything as "unknown" unless you have investigated it sufficiently and therefore not speaking on it, or by treating your knowledge as something worthy of conversation--with caveats, like "I read it in so-and-so, and they've been known to sensationalize or get science wrong, but..." or "I haven't followed up on this in a couple of years, but I heard there was a promising claim..." Then this may spur someone who knows more (or less) to speak up and may cause some further search for the truth later.

It seems to me that the main problem arises when someone bickers bitterly for a fact they've accepted to be recognized as true even when they shouldn't feel that certain. Of course, there is damage done in miscommunication, as when someone gets the impression (we'll say it's nobody's fault) that your pure speculation or dubious source is solid fact. And the best defense against that is knowledge of how much social wisdom is BS.

Perhaps it would be better to recognize that you have another class of filing information--"I've heard no contradictory evidence, it's useful to think about, but I wouldn't build a bridge on its blueprints." Going beyond that on a few statements may very well be quite a good thing, but I doubt it will eliminate altogether your tendency to remember and mention dubious facts.

I'm not sure this really counts as a bias. It seems quite rational, unless you will actually suffer immediate and significant consequences if you are wrong about string theory.

The cost of privately held beliefs (especially about abstract truths) is quite low. If I believe the earth is a flat disk on the back of an infinite stack of tortoises, and if I'm, say, a car mechanic, I will not suffer at all for this belief. Unless of course, I mention it to my friends, because they will judge me for it, unless of course they all believe the same thing, in which case I'll probably proclaim this belief loudly and often and possibly meet up with them to discuss it on an appointed day of the week. I may suffer because I fail at epistemology, but it doesn't seem clear how trusting the wrong source on one marginal occasion will corrupt my epistemology (doubly so if I'm refined enough to have a concept of my own epistemology). Taking epistemology as exogenous, there's really no cost to a marginal false belief (that does not affect me directly).

Having a false belief about some fact that has no direct bearing on your life is way, way, way cheaper than publicly expressing belief in such a fact and being refuted. There seems to be nothing irrational about investing more energy fact-checking in the latter scenario.

Edit: Two things.

First, the turtles example was poorly chosen, as it blurs the line between belief and epistemology too much. Better examples would include, say, wrongly believing celebrity gossip, or having incorrect beliefs about unpractical science due to a lack of information or a misunderstanding of alternatives. If the car mechanic believed Newton was totally right (because he hadn't seen evidence to the contrary), this would be a very, very low cost false belief. Interestingly, "Barack Obama is a Muslim" probably falls under this category, though it blurs the epistemological line a bit more.

Second, it's also quite possible that you care more about misleading others than you do being wrong. It's easy enough to change your mind if you see contradictory evidence or reread the article and understand it better. It's rather harder to change other people's minds who have been convinced, and you'll feel like you've let them down as an authority, since they trusted you.