Why Fiction Beats Truth

> Your truth is not the only truth.

I don't think this is a helpful way of saying this. There is, in fact, only one truth in terms of what the facts are. All of us have different models in our head of what the world is like; these models are incomplete and often wrong. But the weakness of our models of the world won't stop reality from coming and biting us when we make bad decisions.

It is true, that sometimes when people disagree on a political idea, that what's happened is that one person believes A (which is true) and hasn't considered B; and another person believes B (which is true) and hasn't considered A. In that case, the best thing is for both people to learn both A and B (see a "perspective [which] is not false just because it opposes yours"), and revise their plan accordingly.

But in the case of polarized opinions, what normally happens is that at least one side believes X, which isn't true, (or is a gross exaggeration). In that case, while it may help to understand where the other person is coming from, it remains the fact that the right thing is for that person to stop believing that X is true.

So take immigration. Some people claim immigrants "take jobs away" from the native population; others claim that they 1) take jobs that the natives don't want, and 2) create as many jobs as they take by being customers as well as workers. All of these individually can be true; but at the end of the day, there will be a "number of jobs" and "number of unemployed" which is a fact, and will or will not be higher. That's the number that actually counts, not these hand-wavy predictions.

> (e.g. "should we introduce a carbon tax, and if yes, in what form?"), you should be able to reach common conclusion in a group of honest people.

Really? Even assuming that everyone agrees on the model of policy responses - introducing tax X has result Y - people will still have strong opinions about the distributive effects of those and even the relative importance of the environmental effects. People don't usually come out and say "I'm not willing to spend a dollar to prevent Miami and Bangladesh being submerged", because that sounds bad, but they come very close to it.

> some generalizations are wrong, and some specializations are wrong.

On this we agree - some models are just delusional "motivated reasoning"; Sandy Hook "truthers" etc.

> Even assuming that everyone agrees on the model of policy responses - introducing tax X has result Y

This is a purely objective question, and can be discussed between honest parties until a common conclusion is reached.

> people will still have strong opinions about the distributive effects of those and even the relative importance of the environmental effects.

This is a mix of potentially subjective values and objective statements about how those values are affected by proposed solution; I argue that even if a common conclusion cannot be reached, we can usually come pretty close to it.

> People don't usually come out and say "I'm not willing to spend a dollar to prevent Miami and Bangladesh being submerged", because that sounds bad, but they come very close to it.

People don't usually think in these categories; between lack of information, misinformation, opportunity costs (no one has time to be up to date on everything, or even to think through everything) and psychological discounting ("I worry about securing food for my children tomorrow, not about some uncertain future 50 years from now"), you end up with people who believe in things that are wrong even when taking their own subjective values into account.

I feel the problem we have isn't with subjectiveness - most political issues are objective enough in principle. We have a computational problem - we can't get enough people to think and talk through issues deeply enough. Instead, people fall back to computationally efficient heuristics - ideologies and soundbites. Pattern matching everything into beliefs like "capitalism is bad", or "government regulation is bad" is a faulty generalization, but it saves on thinking.

It may be that at the scale of our current problems, this computational barrier is in practice insurmountable. If so, we're fucked, and I'm not sure what to do. But it's probably why there's so much pushback against democracy from the "intellectual elites" - because they realize that this problem only scales with the number of people you need to involve to make a decision.

(N.b. trust is another computation-saving hack humanity has used since forever, and what enables us to have societies as large as today. The trust that other people don't try to hurt me, that they have my interests in mind. It's damn effective, and that's why I'm particularly hateful towards individual, company and government activities that erode this trust.)

I was thinking more along "being correct about change you're supposed to get in a shop" than "being correct about what to do in a fire", but still, I said people learn. Life-or-death situations are pretty rare and an individual is unlikely to experience them enough times to learn from them (much less if any small mistake leads to the tragic outcome). So the learning happens at the institutional level (fire departments are pretty smart about what to do in a fire), and best individuals can do is get trained.

"Real life is like bad television" (paraphrased woody Allen)