Popular Culture as Distributed Lookup Table

“It’s hard for me to understand what Americans are talking about because they are always referring to TV shows or movies.”

I’ve heard this same thing, both from my wife –  who did not grow up in America, as well as from coworkers who emigrated.

There’s an episode of Malcolm in the Middle, where Malcolm’s mom says that if he watches too much TV, he’ll become stupid. Malcolm responds, “TV doesn’t make you stupid. TV makes you normal.”

I am hardly the first person to observe that TV shows like ‘Saved by the Bell’ present us with scripts for how a person could be. Far from being ‘just mindless entertainment’, these shows imprinted a generation with scripts for life. There’s the jock, the nerdy guy, and the cool guy who’s always getting into trouble.  As adults, they became the sales guy, the tech guy, and the CEO. Check out anything by Chuck Klosterman for more to this effect. What I think is a novel claim is the idea that we can view mass media as being a flow of information that defines and controls culture, with that information content effectively serving as a lookup table.

For generations, we learned much of how to behave in life by watching others and imitating them. My daughter does all kinds of things that she’s seen us do, including ‘vacuuming the floor’, ‘cooking in her kitchen’, et cetera. She’s never done these things herself, but she’s seen us do them many times. She is instinctively imitating our behavior.

Deciding how to be in the world is hard. The space of possible human interactions is insanely massive, and most moves are just bad.  By watching what other people do, and occasionally imitating them, we can effectively learn from other people’s experiences.

The P / NP  distinction surfaces again, here.  It’s very difficult to find new ways of interacting with the world that work well. But if someone shows you a way of interacting, it’s much easier to verify for yourself if that way of interacting with the world is effective.

Each of us could learn, entirely on our own, how to navigate the massive, tangled space of possible social interactions. Each of us could learn, entirely on our own, what works and what doesn’t.  That might be possible, but it would be massively, insanely inefficient from an energy perspective. The main shaping force that has guided evolution is energy efficiency. From that perspective, it makes more sense to have people do all kinds of little experiments on their own, and then share the results of these experiments with each other.

Thus, pop culture makes perfect sense as a solution to a certain computational problem: cheaply enabling people to learn from others how to navigate the massively complex space of human social interaction.  When I started learning how to do all of this stuff around age 25, I quickly realized how difficult it was to learn social skills through a combination of trial and error, and first-principles reasoning. I was baffled by the fact that everyone around me seemed simultaneously:

  • Able to perform these incredible feats of human social interaction
  • Completely unable to explain why they did what they did.

When I asked people questions like ‘how they knew the right thing to say in a given situation’, they usually didn’t have an answer. Often, they’d be come slightly irritated, as if this were an unreasonable question to ask.

Eventually I realized that most people interact with the world like self-taught javascript developers who don’t know a thing about what’s going on in the memory. They know “If I do this, I get that response”. In other words, most people have internalized a sufficiently rich “action->outcome” lookup table, that they are able to act as if they have an O(1) algorithm for computing the ideal strategy in any situation. This works great, as long as they only get into a limited set of situations.  

Perhaps travel broadens minds because it continually causes you to encounter situations that you don’t have in your lookup table.

So most people don’t actually understand what they are doing, or why they are doing it. They just know, instinctively, that it’s the right thing to do in the situation. They’re unwilling to explain why, because they can’t.

This kind of knowledge is the sort of thing that works brilliantly, until it fails.  That’s how a lookup table is – it’s the ideal solution for any question with an answer that’s present in the lookup table. It’s also abysmal for any question whose answer doesn’t lie in the lookup table.

The limits of the lookup table: that’s where we’re running into a ton of problems as a society right now. Older adults are all running into the limits of their own lookup tables, and the younger people are left making things up as they go along, in response to all kinds of things that didn’t exist before.

There are two problems with us having this giant, shared lookup table of embedded cultural knowledge. The obvious problem is that this lookup table it isn’t any good at dealing with unfamiliar situations.  The second, non-obvious problem is that the lookup table looks like knowledge, but isn’t.  Our culture rewards people with fast, large lookup tables far more than it rewards people who can figure out a solution to a complex problem from first principles.  

Of course, fast, large lookup tables are useful and I don’t want to claim otherwise. The problem here is that when someone says things like “I don’t know”, or “it’s not clear to me”, we tend to tune those people out, in favor of people who loudly proclaim their confidence in specific solutions. It’s like we’re evolutionarily predisposed to the idea that the computational complexity of our biological niche is not bigger than our lookup tables. We instinctively rally around people whose lookup tables seem to be the richest and most complete. Instead, we ought to listen (but not necessarily follow!) people who can help us compact our lookup tables, throwing out redundant information, and erasing things that used to be true but no longer are.

Historically, the role of ‘thing which helps the human network to forget’ was death,  but it seems that we are moving too fast into the future for death to properly do its job.

We generally think of smart people, “thought leaders,” as being people who say things that sound interesting and insightful, that make us go ‘hmm’.  We don’t generally think of smart people as people who say things which sound confusing, unintuitive, and frightening. And yet a historical glance at how paradigm shifts have been received should make it clear to us that complex truths have generally been opposed most heavily by the intelligent, educated members of the old guard.

If an idea simply resonates with you and make you go ‘hmmm’, it can’t be nearly as groundbreaking as an idea that sounds simultaneously confusing and terrifying. It’s easy to thumb our noses at people who found the historical ideas of evolution terrifying. It’s much, much harder to realize when we’re doing the same, and defending the contents of our lookup tables instead of reasoning about reality from the ground up.

So our current state of political affairs, from a computational perspective, looks like this: We’re all running around with outdated caches of information, heavily trained to follow and listen to people who have quick cache access. Most of us haven’t learn how to interact with the modern world, but instead have learned how to interact with a world that no longer exists (and in some cases, arguably never did.).  

The same teachers who told us “you won’t carry a calculator everywhere with you” also said things like “don’t change jobs too often, it’ll look bad on your resume.” The world is changing, rapidly, and our lookup tables are almost all out of date.

For similar reasons that many young adults are drowning in student debt and unable to find gainful employment outside of a few large cities,  most nursery rhymes and fairy tales are incomprehensible to my daughter. The first one she really took a liking to was “wheels on the bus.” This is probably because a bus is something she’s actually seen, whereas farm animals aren’t something she encountered on a daily basis, and yet they are heavily overrepresented in the songs she hears.

She has never once encountered: A king, queen, prince or princess, a castle, houses in a forest, a woodcutter, a wolf, or anyone who has ever been threatened by one. She has never seen a tuffet, and has no idea what curds and whey are.  She’s seen a bus, however, even been on one  (Which was a huge source of excitement for her!), and so “Wheels on the Bus” is her favorite song.

If we grew up a few hundred years ago, however, these stories would have been far more relevant to her day to day life.  Even when I was a kid, these things kind of made sense, because Grandma and Grandpa’s house was kind of in the woods, so it didn’t’ seem that unlikely that we might go on a long walk there.

Yet my daughter’s grandmother doesn’t live in a house in the woods, that we’d get to only after a dangerous journey. She lives in a faraway land called Ohio, accessible only via airplane. She sees her grandparents on a screen far more often than she sees them in real life.

Are these stories we read to her doing her any good? Or are they preparing her for a world that no longer exists?

I love a series of children’s books called ‘Elephant and Piggie’, in part because they help my daughter develop her social emotional intelligence. She’s starting to understand why, in given situations, people feel the things they feel. She has learned to recognize the subtle difference between ‘determined’ and ‘angry’. She has recently become interested in the concept of ‘blushing’.  These books are great for populating the ‘situation->emotion’ cache of a young primate.

And yet one of these books, on one page, shows why everything seems to be breaking down in society right now.  A character has an idea, and one of these pops up over her head:

I wonder if, eventually, this image will come to represent “An idea that someone had during the presidential administrations of George W Bush, or Barack Obama.”  We don’t have any of these in our house, and I doubt her school has any either. We’re familiar with symbols of old technology representing bygone eras, but we haven’t yet internalized the idea of ‘a bygone era that lasted about a decade’. A horse and wagon could be a symbol for thousands of years of history. This lightbulb above lived for maybe 10. “The past” used to be older than we were; soon a number of important pasts will be younger than my kids.

We’ve all heard the idea that most kids growing up today will never see a floppy disk, and yet understand that this image is a symbol of saving a file. I haven’t yet heard anyone link this phenomena to the election of President Trump, and one goal of this essay is to attempt just that. The rapid pace of technological advancement should be seen as a tectonic force with immense political ramifications.

Speaking of politics, here are a few bets I’ll happily take:

  • Most people have seen reproductions and images of courtrooms far more times than they have seen actual courtrooms.
  • Most people have seen far more fictional interactions with police officers than they have had actual interactions with police officers.
  • Most people have seen far more fictional battles than they have actual battles.
  • Most people have seen far more fictional negotiations than they have participated in actual business negotiations.

This list could go on and on, but I think you’ll get the point.  The reason these things are relevant is that courtrooms, police interactions, business negotiations and battlefields are simultaneously

  • Incredibly important causal drivers of large-scale human outcomes
  • Understood, by the vast majority of persons, only through facsimiles created strictly for the purpose of entertainment.

You might consider that this state of affairs is kind of odd, or maybe even a bad thing. Is it really good if the systems which determine large scale human outcomes are not really understood at all by the majority of people?

Of course, the obvious response to this state of affairs is the fact that most people don’t need to have this knowledge, in order to live their day to day lives.  So maybe it’s fine?

I think this dichotomy is not good for democracy.  I wrote a tweetstorm to this effect (edit: i’ve since deleted my twitter account, so this tweetstorm has dissappeared) that went viral in Brazil, because it apparently resonated with people there.  That country recently elected a Trump-like politician of their own, after the establishment candidates were brought down in a series of corruption scandals.

If most people are ignorant of the ways power is used and exercised, and have no first-hand experience of it, I don’t see how a democracy can function.  Our lookup tables for these important situations aren’t calibrated based on real world experience, but merely on stories. They’re calibrated not based on what works, or what doesn’t, but based almost entirely on what stories we’ve already been told.

For example, what would happen to support for the war on drugs, if every single police drama on TV was about police violence ruining people’s lives with botched drug raids?  Of course, that’s not an accurate representation of what police do all day, any more than stories where the police are always right. What would happen to people’s attitudes towards police if most police dramas were about the  mundane, boring day to day reality of dealing patiently with obnoxious, ridiculous persons, and occasionally having terrifying encounters in the dark where the difference between life and death hangs in a split-second instinctive reaction? Or would anyone even watch those shows?

When we consume popular culture, we’re populating and updating these internal lookup tables, and changing our beliefs about reality, even though we know these are just stories.  When you see a police drama, you understand that the high level plot is, of course, fiction. But your brain doesn’t apply that same ‘reality/fiction’ distinction to the way the weapons sound and kick, the motives of the criminals and techniques the police use.  

When we do see accurate depictions of things we’ve never encountered, we often reject the realistic depictions in favor of the facsimiles we are more familiar with. Try to imagine watching a movie about a battle in space, without any sound.  Most people would reject this movie – a more accurate representation of the real thing – in favor of a facsimile that beras much less accurate a representation of reality.

If we are doing this – rejecting reality in favor of simulations that match our expectations – to situations like ‘explosions in space’, why wouldn’t we be doing the same thing when it came to ‘decisions made by persons in power?’

If the only stories that get mass traction – and thus influence mass beliefs – are those which are gripping, entertaining, thrilling, and scary – is it any wonder that society now feels like a reality television show?   If everyone is walking around, calibrated to be a startled primate in a scary world, where A BAD THING COULD HAPPEN AT ANY MOMENT, is it any wonder the world feels like it does today?

What would entertainment look like if accurate depictions of reality were more economically viable than facsimiles? Is that even possible?  Or are we soon bound for a visit from Death, to help flush the caches of billions of primate-meat computers, by performing a hard reset on hardware which doesn’t support such an operation? Or am I, the author, just a frightened primate, trying to make sense of a complex reality in terms of stories I’ve already been told?

To recap: popular culture is a distributed lookup table, containing answers to questions about reality.  Our evolutionary training has us confusing glib soundbites and quick answers with deep understanding of reality. We’re voting for people to make decisions that we don’t understand ourselves. What we see in the external world generally has more to do with what we already believe than with the truth.

The world continues to change, faster and faster. If we’re going to survive, we need to see our internal lookup tables as being hardware caches of a certain size, and not repositories of truth that, once learned, need not be unlearned. We need to get much better at learning how and when to unlearn facts, rather than simply insist that learning more is the answer. We need to get better at finding people who have better answers, and ignoring – or at least, listening less, to people who have routinely gotten things wrong.  I believe these changes are entirely feasible with a combination of cryptography and quantified social networks. I believe these new technologies are just around the horizon. I’ll write more on those later.

10 thoughts on “Popular Culture as Distributed Lookup Table

  1. I wish you’ll publish a book with your writings. I know you already published one book, and I admit not reading it (yet), but it is dated further back than these writing, so I think it’ll be awesome to have those recent thoughts of yours that you publish here in your blog, in a book as well, or at least having an organized PDF file, like for example Eliezer Yudkowsky “Rationality: from AI to Zombies”. Your ideas are quite similar in a way to the ideas of Joscha Bach (there are some very interesting lectures on YouTube). I wanted to email you this message, but I could not find your e-mail address.

    1. Hey, thanks for this comment. I can see i’ve responded much later than you wrote it, but life has been super crazy for me, even before the pandemic. I’ll probably get around to putting out another book, and i’ll try and ping this thread if/when that happens 🙂

  2. If what you’re saying is true (which seems really plausible) then my mental hardware is pre-optimized for recall using lookup tables. This makes me more tempted to double down on my own use of lookup tables, but simply curate them more often to make sure all the entries are accurate. I also have this other mental tendency to use the availability heuristic to trim down what I need to recall into the most-often-used entries, which would be useful if I needed to trim the entries down to just one answer to slot away in the lookup table and then delegate automatic responses to a subconscious process. If I need to find it quickly in the future, a plan that worked at least once before is quickly and automatically suggested.

    How does this help society? No idea. Very interesting to mull over as a self-optimization technique though.

    1. > but simply curate them more often to make sure all the entries are accurate

      YUP that’s super important. Just highlighting that ‘i am using a cache often’ made it clear to me that i need to be mindful about when i’m relying on a cache, and at least asking if the cache is up to date.

      > I needed to trim the entries down to just one answer to slot away in the lookup table and then delegate automatic responses to a subconscious process

      This is even better than a lookup table!

      > How does this help society? No idea.

      I think society benefits best when there’s a large number of people with different values and different abilities, each striving to optimize their own capacity to accomplish goals and nudge the world forward. People who are great with their lookup caches should find people who have similar values, who maybe aren’t so great with a cache, are great at producing models.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.