#probably-obvious

Often, I feel the desire to write something, but it feels too "obvious". And then I wait a while and see that if I had taken the time to write it out I could have pointed people to my writing when it was useful, and probably would have developed the idea better just by having written it. So I'll start an intermittent series where I describe something that may or may not be obvious, and I'm sure will occasionally be blatantly wrong.

Let's start with a real softball: the choice between whether something is a "rule" or "exception" in any kind of system is purely aesthetic.

If I say "there are no even primes except 2" that's fine. If I say, "anything is a prime that has at least two unique divisors, one and itself, therefore 2 is a prime, but no other even numbers are" that's also fine. These are the same, in terms of the precise system they describe, but of course they encourage different kinds of thinking. The first one gives a general rule people can use to easily test if something isn't a prime in half of all integers, while the second gives implies the causal reason why 2 is prime and other even numbers aren't.

Aesthetics aren't meaningless. If a teacher says, "everyone who got at least an A-" can go to lunch now" that has a very different impact than if they say "everyone can go to lunch now, except Johnny" (implicitly, because he's the only one who apparently couldn't ace this easy test). Obviously the second one signals that an authority is singling Johnny out, by name.

I think that a lot of people confuse the idea that the fact that aesthetics have an impact means they describe something differently. It's all silly definitions and I'm not tied to these exact words, but the point is that I can select the same formally recognizable options while connoting my way to all kinds of different ends, but these two processes operate at different planes of the human sphere, planes that don't always interact. I can't connote my way to new citizenship, I have to submit paperwork even if the country is happy to have me. I could possibly connote my way into a new relationship, but it depends on the other person's recognition and use of that implicit language, much as Rao describes "Powertalk".

Why do we have the grain rye? We have it because there was a weed, rye’s ancestor, that looked enough like wheat that it was hard to separate out. We tried and tried, and eventually it become something delicious enough to eat: rye.

There’s another way this could have worked. Rye could have been super inedible, no matter what we did to it, be impossible to filter out with the technology of the time, and caused significant starvation.

Metaphors are like this: they can go eitherway. The dynamics that connect a given object to metaphorical dynamics may carry on their resemblance past the metaphor, or they may just be skin deep. We may not need an airplane to flap its wings, but we might need to understand humans better to make a reasonable AI dialogue agent. As usual, the conversation is ruled by the abuse of strawmen.

People generally praise proactive measures, because the commonsense intuition is that people tend to be lazy and not take these measures when they should. I’m not here to argue with that; it’s probably right, but I don’t even know how I would figure out how to prove it.

One thing I want to point-out, though, is that we generally only praise as “proactive” measures we see as advantageous. There are many, many things where (a) they simply don’t have the right effect if you do them too far in advance so we don’t think of doing them in advance as good or (b) we don’t model the system correctly and we think we can do something in advance to good effect, but we cannot.

An example of (a) is cleaning your house to a high standard a month before guests come over. It might help you keep it clean, but there are plenty of things people clean for guests they’re happy to have dirty all the rest of the time, so in all likelihood you’ll really just be making more work for yourself.

An example of (b) is doing your laundry earlier than you used to, because you feel bad letting so much laundry build-up. Obviously, there’s a balance to be found that fits your hygiene aesthetics, but I would like to point-out that if you just feel bad because you “should have done it earlier” from some vague background work ethic, then you’re doing yourself a disfavor. The earlier you do your laundry, the more time you spend doing laundry total, because you can generally fit more into the laundry machine, so all you’re doing is incurring higher overhead per garment washed.

I think in America these days there’s a weird mix of what Douglas Hofstadter calls “the paranoid style” and flat earth level conspiratorial assumption. (I link to Hofstadter’s piece hesitantly, as I think he was too confident that the world isn’t more full of conspiracy than he thinks it is. conspiracy in small groups is incredibly common!)

This post is not about where you should be on this spectrum, it is about how to think about “conspiratorial pressure”. Let’s say there’s something that’s uncomfortable for a lot of people to admit, e.g. masturbation still being relatively frequent even in communities that look down on it. Well, basically, all that needs to happen for this to be the case without most people properly accepting it is that:

(1) Public messaging, especially between individuals in public spaces, revolves around one narrative, e.g. everyone says masturbation is a sin.

(2) There aren’t many opportunities to share counter-the-norm information that would help the speaker without endangering their relationships.

Lot’s of things are like this! How much do people clean various parts of the body? How much do people use birth control when they should? How often do people say racial slurs alone in their home? we will never know

Often people use this kind of thinking to say “Everyone knows X but everyone is just too scared to say it.” in order to attempt a preference cascade. If they fail, they often blame others for bullying people into the currently acceptable opinion, which definitely does happen.

However, let me offer you a tool for pushing back against the “preference falsification emergent conspiracy”, which is often over-used:

For you, as an individual, to have evidence that people might really believe X, all you need is to have known one person who wasn’t that careful and kinda just blurted out something that implied they might be X adjacent. If you talk to a reasonable number of people a day, it is crazy unlikely that you will never have been exposed an X adjacent opinion if it’s super common.

There are many exceptions, here are a few of the most important ones:

- Non-Uniform Preference Distribution—it might be that you’re not exposed to the kind of people who believe X. This is common, but you should really consider whether you would have really had zero exposure, or X is just a strawman.

- The Way You Act (or Are) Dissuades People From Talking—pretty self explanatory. Most girls aren’t going to tell dudes about their periods because of cultural stigma towards the broader subject, rather than any specific opinion or belief.

-You’re Not Interpreting X Correctly—a classic example of this is misinterpreting sarcasm because you’re already sure what someone thinks.