mardi 16 mars 2021

Breakfast Breaks

If you want to figure out and apply truth in complicated situations, there's a technique. There are two parts, and everyone can improve at both.

First, seek out evidence. Focus as closely as possible on what are especially more likely to be facts - the "facts" that have the best chance to be factual. When you get a few, you will find that many of these are quite boring. It's easy to overlook them.

The secret is to find the most certain (ie, least uncertain) of relevant facts. You know belief is all a gamble, but you are playing the long game and looking for the best gambles. How? So, do you really want to identify the best facts? The trick is extreme sensitivity and openness to the possibility that a "true fact" isn't precise, accurate, or true. That's the only way you can find the really good bets.

But when you do, the really good bets will (usually) look to others like any other statement. Often they are boring. The difference is that on careful examination, you could find no cracks. You strove to break these eggs, began to make an omelette over and over, but you couldn't break one or two. Those may be qualitatively different, more resilient. From a distance, without careful examination, such pleasant tints of nothingness usually don't appear. All the statements will seem like eggs, and most of the time, eggs aren't breaking. Most of the time, eggs are eggs. Most of the time, you can break them and make an omelette if you want. Most of the time, facts are facts. Here, though, you want to make a non-omelette with the rare eggs that don't break.

Second, find the most basic possible logic. The simpler the better. This, too, will seem boring. Like the egg-sorting, it will seem meaningless to the uninitiated, or to those who aren't applying the technique in this context. Simple logic isn't impressive. Simple logic may make you look naive.

So you have to stop to ask yourself. Are you trying to impress people, or do you want to make a non-omelette with the rare eggs that don't break? Making an omelette - impressing others - is a different pastime with different demands, sometimes quite opposed to what's needed for the non-omelette - ie, for finding and applying truth as well as possible. You have to decide what you're trying to do, discern/predict or persuade, and remember not to forget that decision.

It's easy to make an assumption for today and then misplace the memo-to-self that this was an assumption. Everyone does this. Not doing this is very difficult, and you need to choose your battles. In trying to persuade others, we often wind up forgetting about the things we didn't really know. We believe we know them. It's inefficient to consider all the ways we don't know what we're taking about when we're trying to push someone else to decide right now. And on the other side, it's inefficient to worry how others will take things when we're trying to discern and predict as accurately/precisely as possible.

Logic is brittle. It's both fragile and immensely strong. Your best bet with logic is to use the simplest logic. Just as you did for the facts that feed logic, you want to put aside how boring or clever or hurtful or nice or impressive or awkward this is, and look for the ordinary eggs (or ordinary cardboard egg cartons now, if you like) that turn out to be extraordinary by not cracking. The unambiguity of logic can make this logic-soundness egg-carton-sorting much easier than in the preceding "true or false facts" step. Just keep the logic ultra-simple, and if the facts it's working with really don't break, then the result of the logic is guaranteed. The trick is to make sure the logic is correct. It is not sophistication. It is correctness. Nature does not award bonus points or put stickers on more complicated solutions: logic that works works. Humans are not good at complex logical chains. We can only manage this by breaking it all down and using the simplest pieces.

(I can hear some of you thinking, as I am, that this is really my own limitation, which I'm trying to impose on others. Maybe. But I'll bet heavily that it applies no less if you have the highest IQ in the world, and if you do, I bet you already know it.)

This approach to expanding what you know runs against the social intuition that says objectivity and prediction are sophisticated and should involve impressive facts and mind-boggling reasoning. Good technique turns out, very often, to demand quite the opposite.

Let's recap:

- One, the "facts" must be unassailable or as close to that as possible (which means they often won't be surprising or appear any different from other statements that most people take as givens).

- Two, the logic must be unassailable (which means that every piece of logic must be made as simple as humanly possible).

This level of care over certainty is crucial in math, science, engineering, etc. It's how you make sure Mars orbiters don't crash into Mars after assuming the wrong measurement unit in some software recess. It's a lot of what's implied by "rocket science," yet it's accessible to everyone with a bit of practice.

It won't tell you everything. But it will give your perspective on a topic a solid skeleton. With the best skeleton you can get, you'll find a lot else works better. Muscles can do a lot with a skeleton. The sequoia of intuition begins to grow in fertile mud; speculation, while often fun, may not be so idle now, thanks to all the homework. And you may find yourself understanding more of what people tell you, and remembering it better.

We tend to separate "rocket science" and "mathematical proof" from everyday life. In complex real-life scenarios, it doesn't get any less important to find the core, most probable facts. And avoiding logical errors by keeping logic extremely simple doesn't get any less important. It's just that humans need initial training and continual practice to manage that. And most of us, at best, get that training in one or two areas, and don't apply it in other scenarios where we don't recognize it's needed.

What are the core, least fallible facts? What is the simplest, least fallible logic? Start there, if you can. Then take all the feedback you can get.

Either that or recognize that you are not discerning truths to use in understanding how stuff works and make accurate predictions. If you want to relate to other people and perhaps convince them, it's about relating to other people. Then the thing to understand is how they feel, and how you feel, and what stories led or can lead to those feelings. Sometimes it's also about using specific persuasive techniques, either to trick people into agreeing, or to help them find an idea - one you want them to remember and then act on - to be especially memorable. There are times when all of us appreciate the latter (an illustration that hits home), and occasionally most of us also can appreciate the former (just the right push), if it's done with care and concern for our own amusement or best interests, or at least society's.

Grokking versus convincing - they're at odds, but sometimes you're lucky and they can happen half-decently at the same time. Think of both kinds of influence on others (or the world) as writing, for a moment. Writing often involves drafting "hot," without much filter, and later editing "cold," which is almost all filter and correction. We can to some extent combine these modes. It's possible to write a thing once, even creatively, and move on, because it doesn't particularly need editing, or we have more to work on.

I can't comment much on that combining. It applies in writing, and an analogous duality (synthesis of opposites) applies in debate and influence. Some people (I'm one) will find you much less convincing if you don't spend much time in the more objective mode. So to persuade them (me, often), don't be persuasive. Put objectivity and uncertainty first, for us, despite wanting and hoping to convince. (I'll give you a chance. I'll give you many. You're welcome. And thank you.) And on the other hand, sometimes intuition or inexplicable feelings guide us to truths. I'm not going to be a rules Nazi about the advice I'm giving. And I know that there's a lot I don't understand about how these two modes combine. Sometimes they can. Often that goes wrong, though.

If you want to discern truth and apply it constructively, use the two steps. They won't banish intuition or feelings. Those will wait in the periphery and will have useful or crucial things to say. But know the difference.