Propagating Facts into Aesthetics

Epistemic status: Tentative. I’ve been practicing this on-and-off for a year and it’s seemed valuable, but it’s the sort of thing I might look back on and say “hmm, that wasn’t really the right frame to approach it from.”

In doublecrux, the focus is on “what observations would change my mind?”

In some cases this is (relatively) straightforward. If you believe minimum wage helps workers, or harms them, there are some fairly obvious experiments you might run. “Which places have instituted minimum wage laws? What happened to wages? What happened to unemployment? What happened to worker migration?”

The details will matter a lot. The results of the experiment might be weird and confusing. If I ran the experiment myself I’d probably get a lot of things wrong, misuse statistics and forget to account for some confounding factors. But I don’t feel confused about how to learn better statistics, account for more confounders, etc.

But there’s a problem that seems harder to me, which is how to change my mind about aesthetics. Sarah Constantin first brought this up in Naming the Nameless, and I’ve been thinking about it ever since.

I think a lot of deep disagreements have to do with “what is beautiful, and what is ugly?”, and inability to directly address this is part of what prevents those disagreements from resolving.

In the case of the minimum wage example, you might run an experiment, and find overwhelming evidence that minimum wage helps or hurts workers. But because there’s lots of confounders, the evidence might be mixed and confusing. How you interpret it will depend on how it fits into your existing worldview.

Part of this has to do with your ontological frame. But I think a lot has to do with aesthetics judgments, such as:

  • Is capitalism ugly and/​or distasteful? You might have very salient examples of how capitalism can result in exploitation, pollution, or people becoming trapped in unhealthy power structures.

  • Is capitalism beautiful? Alternately, it might be salient that capitalism creates supermarkets, gains from trade, and vast surplus. Economic efficiency isn’t just pretty numbers on a graph, it’s real value being created.

These don’t directly bear on the minimum wage question, but might make it harder to resolve.

In some cases, your aesthetic taste might make it harder to update on new information properly. In other cases, your aesthetic taste might help you to notice important patterns more readily.

Why ‘aesthetics?’

I’m using the word aesthetic in a nonstandard way. When people do that, I think it’s important to be clear and about what they’re doing and why.

There’s a few different words I might have used here, including “feelings”, “ontologies”, “frameworks”, and “values.”

Most obviously, I could have asked ‘is capitalism good/​bad?’ instead of ‘is capitalism beautiful or ugly?’.

I’m making a fairly strong claim (weakly held) that “is it beautiful or ugly?” is at least one of the important questions to be asking, in addition to “is capitalism good/​bad” and “does raising minimum wage help or harm workers?”. Not because it’s how a flawless AI would think about it, but because it’s how humans seem to often think about it.

What is an aesthetic?

An aesthetic is a mishmash of values, strategies, and ontologies that reinforce each other.

The values reinforce “you want to use strategies that achieve these values.”

The act of using a particular strategy shapes the ontology that you see the world through.

The ontology reinforces what values seem important to you.

Together, this all creates a feedback loop between your metagoals and subgoals, where the process of using this cluster of value/​strategy/​ontology makes each link in the chain stronger.

In humans (who have messy, entangled brains), this cashes out into feelings, felt senses. The original goal and the metagoals blur together. I think “this helps me achieve my [generic] goals” might reinforce “these particular subgoals I have are good goals to help with my overall flourishing.”

This might be implemented via evolution over millions of years, or via human brains over decades. A Just So Story I’m not sure I endorse but hopefully gets the point across:

“A flower is beautiful, you say. Do you think there is no story behind that beauty, or that science does not know the story? Flower pollen is transmitted by bees, so by sexual selection, flowers evolved to attract bees—by imitating certain mating signs of bees, as it happened; the flowers’ patterns would look more intricate, if you could see in the ultraviolet. Now healthy flowers are a sign of fertile land, likely to bear fruits and other treasures, and probably prey animals as well; so is it any wonder that humans evolved to be attracted to flowers?

Here are some things that you might find beautiful, or distasteful:

  • Mozart

  • Punk Rock

  • Readable, well-written code

  • Clever hacks that got the job done quickly

  • Cities built on rectangular grids

  • Winding alleyways in villages where nobody has consistent names

  • People being physically affectionate in public

  • A harsh, barren desert

  • A lush valley with a river

  • Swamps /​ wetlands

  • Nature in general

  • Manicured gardens

  • Books

  • Throwing away books

  • People speaking in languages different from yours

  • Dense spreadsheets laden with accurate data

  • Minimalism

  • Frugalism

  • Patriotism

  • People going out of their way to be kind to their neighbors

  • People going out of their way to solve small-but-common problems using math

(If you’re like me, you might find it distasteful when people make moral arguments that seem rooted in distaste… and then feel kinda self conscious about the contradiction)

Sometimes you’re doublecruxing with someone, and they’ve explained their model. And their model… makes sense. But the conclusion just seems so damn ugly. You want to take 5x the time to write beautiful code, and they just want you to get the job done and ship it.

One thing you can do is push aside your aesthetic judgment, shut up and multiply. This may be useful for expediency.

But sometimes, I think the correct thing is for one or both people to backpropagate facts through their aesthetics.

I do not think you should rush or “force” this. Your sense of beauty is there for a reason. But I have a sense that figuring out how to do this well is a key open problem in applied rationality.


Examples

Are Swamps Beautiful?

Compare the swamp with a verdant forest.

If you’re like me, swamps seem ugly. Forests seem pretty.

My associations with swamps come largely from stories (and perhaps most concretely, from the game “Magic the Gathering”), where they’re often presented as places of disease, murky horrors and corrupt magic. In person, swamps are physically hard to walk in (sometimes solid ground turns out to be algae), and full of mosquitos that bite me.

Are these associations accurate?

Well, the solid ground and mosquito issues are definitely real.

Swamp Thing is not real, life-stealing magic is not real.

Are there additional facts I can learn? My sister evaluates land for construction projects. She says that swamps often serve important roles as a natural way to filter water, and when you naively drain swamps, water quality in an area gets worse. James Scott in Against the Grain claims that early Sumerian civilization developed in swamps, where food and resources were plentiful and life was fairly leisureful – until empires arose and subjected people and forced them to switch to easily-taxable crops instead of the ones that grew naturally. (Speaking of which: is civilization beautiful or ugly?)

I probably find forests beautiful, in part, because they represent a lot of resources that I understand how to make use of. If swamps also supply those resources, maybe I should respect them more?

I also find forests beautiful because my experience stems from a) enchanted forests in fairytales, and b) relatively manicured national parks. If I remind myself that the last time I walked through an untamed forest, it was dense with brambles that cut me ’till I bled. It wasn’t actually a much nicer experience than the last time I explored a swamp. (It also had non-trivial numbers of mosquitos)

In this example, simply mulling over the facts naturally re-organizes my feelings about them. I still find swamps ugly, but less ugly than before. I expect that, if I reflected on this periodically, over time, it would shift a bit more.

Are Harsh Deserts Beautiful?

I am in fact confused by this. My answer is “yes”, and I don’t know why. Deserts don’t have much in the way of resources. Their stark beauty is more like the way a statue is beautiful than the way a forest is beautiful.

I mulled this one over for a while, am still confused and I note it here because “noticing the limits of a model” seems important.

[Edit: this was discussed more in the comments.]

Is Helping Nearby People Other Beautiful?

The first experience I got with aesthetic doublecrux was debating “hufflepuff virtue” with Oliver Habryka.

I had a strong sense that “people helping each other out” was good and right and virtuous. There was a beauty to the sort of community where everyone notices when someone is hurting (and reaches out to help), or when a space is messy (and cleans it up). There was a cluster of attributes that seemed to fit together in a way that was stronger than the sum of its parts.

And this was visibly lacking in the Berkeley community, and it was resulting in people feeling alienated and distrustful of each other, and many spaces being either messy, or burdening a single person with cleaning up everyone else’s mess.

This seemed concretely harmful. But it also just seemed… ugly and bad.

Oliver had a different view, which I summarize as the “systemization and specialization” approach. (previously discussed here)

If everyone has to pay attention to their environment and notice things that need doing, this is a lot of cognitive overhead. If people only have seven working memory slots but they’re spending one of them on tracking the environment, that’s a dramatic cost on their ability to think. For a community that specializes in thinking, this could be quite bad.

Moreover, “everyone pitch in” is just a really inefficient way of getting things done. A better solution is to streamline and automate as much of the work as possible, hire cleaning services, and whatever remaining work needs doing, simply pay one one of the people something commensurate for their time and effort. Specialization is how things get done when you’re doing them seriously.

We argued about this over the course of three days.

I still think there are some things habryka was missing here. But eventually my worldview shifted in some significant ways:

  • I updated that the “everyone pitch in” way of keeping spaces clean doesn’t make sense for longterm organizations with serious funding. Specialization is real, cognitive bandwidth is precious, and it’s generally better to just hire a cleaning service if you can afford one.

  • I updated a bit (talking with Satvik) that my model that “helping each other out in low-key ways builds trust which later enables more extensive projects” wasn’t as strong as I thought. Satvik asked something like “do you think startup cofounders tend to team up because they’ve helped each other take out the trash? I feel like it’s more about sharing a clear vision and principles or something.” And I thought back to some experiences and… yeah that seemed maybe more accurate.

  • I gained a better understanding of where and why the “everyone pitch in” approach is useful.

    • Cleaning services are expensive, and if you’re a fledgling organization or a typical household, it’s probably not worth hiring a cleaner more than once a week or so. Meanwhile, people make messes much more frequently than once a week. If you want your space nice, you have to clean it yourself.

    • There’s a value that comes from having community spaces use the “everyone pitch in” method, in that it creates a stronger sense of ownership and buy-in for the space. It also is a mechanism by which people can relate to each other more easily. While this might not be that important for a company, it seems important for a community that’s aiming to meet community-shaped-needs.

But this all left me with a nagging, frustrated sense that something important and beautiful being lost. I want to live in a world where people help each other out in small ways. It’s the particular kind of beauty that a small town in a Miyazaki movie embodies. It feels important to me.

Under what circumstances should I change how I feel about that?

There’s a sense in which aesthetics can’t be proven wrong, or at least “trying to prove it wrong” isn’t really the right frame of mind.

But… I have an aesthetic preference for consistency, and for believing true things (whether this is good is another question, but I’m taking it at face value for now), which informs my other aesthetics. Aesthetics can turn out to be built out of contradictory pieces, and they can turn out to hinge of false beliefs.

“Trying on” another aesthetic

While talking to habryka, I tried to get a sense of what it’s like to live in the world where systemization and specialization are obviously good and right. What was it like to be habryka? How did this fit together with his other beliefs and values?

Then, once I had a good handle on that, I tried to inhabit “what would it be like to be a Raemon who found systemization and specialization good and right?”. Without actually adopting the aesthetic, I tried fitting it into my existing model. This was a bit of an aesthetic process of its own – like trying on a new outfit and seeing how I reacted in the mirror.

I’m not sure if habryka endorses considering those as an ‘aesthetic’, per se. But I found this process valuable.

I gained some ability to see systemization as beautiful. My sense of hufflepuff beauty became more nuanced and caveated.

Clean Code vs Quick Hacks

Humans have (an instinctive? Learned? I’m not sure) sense that when you smell fecal matter or rotting flesh, there is probably disease nearby. It’s digusting.

Dogs… well, I’m not 100% sure what’s going on with dogs but I think it’s something like “strong odors that mask my scent are more useful than disease is bad”, and for some reason fecal matter is joyful to play around in.

Programmers often learn that spaghetti code is evidence of bugs, even if they don’t know exactly what the bug is yet. It acquires a bad code smell.

Young programmers often do not have this sense of distaste, and it is important for them to acquire it.

On the flipside: there is also a thing where, well, sometimes you’re rushing to ship an Minimum Viable Product and you don’t have time to do everything right. It can be legitimately hard to figure out how much effort to put into “doing things right.” But it seems at least sometimes, experienced coders either need to learn to “hold their nose” and do the quick fix, or to develop alternate aesthetics that they can shift between depending on circumstances.

Knobs to Turn

There are a few different directions this kind of process might go:

  • You could shift to find something more beautiful than you did before

  • You could shift to find something less beautiful than you did before.

  • You could shift to find something more distasteful than you did before.

  • You could shift to find something less distasteful than you did before.

I have some sense that these are subtly different processes, although not much evidence to back that up. I also feel like in each case, going from Zero to N, or N to Zero, is different than dialing an existing aesthetic response up or down.

Gaining a new appreciation for why something is beautiful feels different than gaining a categorically new form of disgust. In particular, gaining a new form of beauty mostly makes my life feel nicer, whereas gaining a new form of disgust increases the unpleasantness

Why Does this Matter?

In Naming the Nameless, Sarah Constantin references this comment by Scott Alexander:

Sometimes I can almost feel this happening. First I believe something is true, and say so. Then I realize it’s considered low-status and cringeworthy. Then I make a principled decision to avoid saying it – or say it only in a very careful way – in order to protect my reputation and ability to participate in society. Then when other people say it, I start looking down on them for being bad at public relations. Then I start looking down on them just for being low-status or cringeworthy.

Finally the idea of “low-status” and “bad and wrong” have merged so fully in my mind that the idea seems terrible and ridiculous to me, and I only remember it’s true if I force myself to explicitly consider the question. And even then, it’s in a condescending way, where I feel like the people who say it’s true deserve low status for not being smart enough to remember not to say it. This is endemic, and I try to quash it when I notice it, but I don’t know how many times it’s slipped my notice all the way to the point where I can no longer remember the truth of the original statement.”

Sarah notes:

Now, I could say “just don’t do that, then”—but Scott of 2009 would have also said he believed in being independent and rational and not succumbing to social pressure. Good intentions aren’t enough. [...]

I think it’s much better to try to make the implicit explicit, to bring cultural dynamics into the light and understand how they work, rather than to hide from them.

Scott’s comment gets at what I mean by “An aesthetic is a mishmash of values, strategies, beliefs, and ontologies that reinforce each other.” He starts with a belief, then adopts a strategy for how he relates his communication to that belief, and then ends up with a vague sense that the belief is “cringey”, and later collapsing it to “cringey and wrong”.

This quite worrying epistemic horror.

I think most of what needed saying, Sarah already said, but it’s worth concluding with here:

If you take something about yourself that’s “cringeworthy” and, instead of cringing yourself, try to look at why it’s cringeworthy, what that’s made of, and dialogue honestly with the perspective that disagrees with you—then there is, in a sense, nothing to fear.

There’s an “elucidating” move that I’m trying to point out here, where instead of defending against an allegation, you say “let’s back up a second” and bring the entire situation into view. It’s what double crux is about—“hey, let’s find out what even is the disagreement between us.” Double crux is hard enough with arguments, and here I’m trying to advocate something like double-cruxing aesthetic preferences, which sounds absurdly ambitious. But: imagine if we could talk about why things seem beautiful and appealing, or ugly and unappealing. Where do these preferences come from, in a causal sense? Do we still endorse them when we know their origins? What happens when we bring tacit things into consciousness, when we talk carefully about what aesthetics evoke in us, and how that might be the same or different from person to person?

Unless you can think about how cultural messaging works, you’re going to be a mere consumer of culture, drifting in whatever direction the current takes you.

I’m hoping this post gives some nuts and bolts on how to actually make progress on that goal.

Again, I don’t know that the specific techniques I list in this post are the best ones, or how often exactly aesthetic concerns are most relevant. I think it’s usually good form to start with an attempt to take arguments at face value, and debate about concrete beliefs.

But, if that isn’t working, I think digging into aesthetics is one of the tools that’s important to have in your toolkit.