How Filter Bubbles Will Save The World

A few weeks ago, I spent a weekend at an obscure conference for a wildly ambitious, way out-there software project. Everyone was united by one weird belief (that the project could, against all odds, succeed). And what’s surprising is that they had a lot of other crazy beliefs, too — on politics, social issues, religion, you name it. I’d never seen a civil discussion between a Trump supporter, a communist, and a Buddhist monk, but this X, Y, and Z walk into a bar joke setup really happened.

Odd beliefs correlate. An easy explanation for this is that people with high openness to experiences don’t spend all their Openness Points believing one crazy thing; they believe a disproportionate share of the crazy things they hear about. But when you look at extremely weird subcultures, you start to notice that they’re often transgressive in exactly the same way: it seems like half of the world’s anarcho-capitalists eat exclusively carnivorous diets, and 80% of the anarcho-communists are vegans. The apotheosis of this is Burning Man, where everyone’s Open to exactly the same Experience. Wow, drugs and nudity, where do you get all your bold new ideas?!

There’s something more here: bonding over a single odd idea widens your Overton window for all new ideas. Why don’t people like to share crazy ideas? Because they’ll get laughed at or shouted down. If you’re in a group where everyone shares one extremely laughable/shout-downable viewpoint, everything else is relatively normal by comparison. It’s like a Costco for heterodoxy: you pay a big upfront membership cost, but once you’ve paid, further deviance is cheap and it’s available in bulk.

Eli Pariser writes about the Filter Bubble, the idea that online media show you the stuff you already agree with, causing increased partisanship in general and an increase in partisan social bubbling. That’s probably corrosive at the political level — although I’d note that in a US context, when we use party identification as a proxy for beliefs, we underestimate how rigid political identities used to be: a huge cohort of establishment Republicans and Southern Democrats switched party affiliations without changing their minds about any important issues. But at a smaller level, a filter bubble is more like a incubator that lets an idea grow safely until it’s robust enough to survive in the wild. What is a startup, if not a group of the only three people out of seven billion who are so terribly filter-bubbled they think that their company can succeed?

The Stakes

Bold ideas are fun, but they’re not merely recreational. If you don’t have ideas you’re afraid to share, that’s dangerous: every good idea starts out as somewhere between ridiculous and anathema. And long-term economic growth is ultimately dependent on new ideas. If everybody stops producing them, it’s the end of a multi-century run of productivity growth — and it will take us a long time to adjust. Our political and economic system depends on entitlement spending and retirement savings, and both of those are predicated on sustained economic growth.

In the best case, a continued slump in productivity means we have to drastically cut consumption in order to invest more in productive assets. (The digitization of the white collar workplace still has a good decade or two left, and there have been analogous historical periods: in A Great Leap Forward, I learned that a substantial proportion of economic growth in the 1920s could be attributed to factories switching from centralized steam power to decentralized electric motors. The idea of factory electrification was older, but the implementation took about a decade.)

But the 1920s were also a time when life expectancies at birth were 54 for men, and government spending was 12% of GDP. No Social Security, Medicare, or Medicaid, and the big entitlement obligation of the previous generation, pensions for Civil War veterans, had finally run off. Now, we have to run fast just to stay in place, and an economy where GDP growth skews towards fixed investment rather than consumption is a tighter-belted economy than what we’re used to.

If we don’t have geniuses like John McCarthy inventing new programming paradigms, our progress will just consist of refining other people’s good ideas. (McCarthy, incidentally, also devoted copious time to arguing on Usenet about the merits of capitalism and nuclear power, and the demerits of communism and organized religion. And let’s not even get started on Isaac Newton’s loony passtimes, or how Keynes kept busy when he wasn’t reinventing economics.)

A Black Market in Heterodoxy

How do you find a good filter bubble, and how do you build one? I have a few highly disparate ideas:

  1. Be Straussian. Go meta online, and object-level in person. If you have some specific, dangerous premise, you can cast it in sufficiently abstract terms that it’s a signal fire to potential allies and a smokescreen to everybody else. Does anyone really know what Paul Graham thinks you Can’t Say? Yes, probably: the people who read the essay, guessed the answer, and asked him. The Internet is a great way to meet friends, but it’s also a great way to make enemies, and the power-law distribution of enemies is much more extreme.
  2. Speaking of enemies, Conspire. Earlier this year, I read Ryan Holiday’s wonderful book, Conspiracy, about the efforts to litigate Gawker into submission. Regardless of your feelings about the matter at hand (strong, I’m sure), what was striking about the book was that it was a long-term, secretive plan that was in place for years before it had any effect, and which didn’t get outed as a conspiracy until a year after that. Operating on the working theory that anyone you dislike probably is a colossal jerk to other people, too, conspiracies are a good way to meet people with interesting and crazy ideas. Perhaps there are many more Conspiracy-style conspiracies that dissolve once the protagonists realize that, having formed a cohesive team with a variant viewpoint, they really ought to get busy making a movie or curing cancer or something.
  3. Find a reliable way to break taboos. There’s a hidden downside to general open-mindedness: without taboos, you lose the ability to signal a willingness to violate taboos, so you can’t tiptoe into dangerous intellectual territory. (This may be why everybody at Burning Man talks and votes the same way: if everybody’s a rulebreaker, nobody is.) Alcohol and marijuana prohibition have their downsides from a policy standpoint, but they produce a surprising sort of upside: they give everyone the opportunity to show that they flout norms. Marijuana is particularly effective here: not only is it relatively harmless as far as drugs go, but it stinks: you can literally smell the transgression the minute you walk into the room.

The Weirdness Shortage

How many people do you know who have worked on something 1. Unconventional, and 2. In relative secrecy for a year or more? People pursue conventional schemes all the time: get that promotion, get into grad school, escape from grad school, buy that house, etc. But starting something unconventional and then finishing up is relatively rare. The only examples I can think of in the last year are stealth-mode startups and a Facebook friend who took up powerlifting but didn’t mention it until his deadlift hit 500. (Lots of people are “working on a novel,” but there’s a tendency to lose steam as soon as you tell your friends that’s what you’re doing.)

Maybe my experience is atypical, but I doubt it. It really seems like America is suffering from a dire shortage of extreme weirdness. Even 2018’s political environment shows it: the big norms-violation isn’t saying totally new stuff, it’s saying stuff that isn’t conventional wisdom now, but was a few decades ago.

I can think of a few reasons: there may just be fewer good and weird ideas left. Maybe we really are at the end of history, although we’ve still structured our economy as though history is going to keep on happening. The Internet is also a factor, in an inverse filter-bubble way: everybody suffers under the tyranny of Grep Risk, the fearsome possibility that someone will trawl through your old Facebook and Twitter posts and find something you said that must have struck you as very funny back in 2011, but that will now be the first search engine result for your name forever. The Social Media Savonarolas are relentless: they need clicks every day, and every new burst of pageviews spreads the same display ad budget thinner and thinner.

You might think that rich people can escape this, and to some extent they can. Past a certain net worth, you really can cry all the way to the bank. But money isn’t everything, and as you get richer incremental dollars matter less, but social status remains important, so being a pariah becomes a bigger risk than losing a paycheck. Wealth correlates with name recognition, especially in today’s media environment, and a wider audience means you run a bigger risk whenever you run your mouth. As it turns out, “Fuck you” money correlates with “Wow, a lot of people just told me to go fuck myself” fame. I don’t have the data to prove this, but a fair amount of oddball-billionaire-projects coverage has cropped up in the last few years, and seems to fixate on things that the billionaire has been doing for a decade but that they’re just getting infamous for now.

Among those of us not lucky enough to have billions of dollars, or even millions of dollars, there’s a different kind of pressure: the costs of housing, healthcare, and education are rising faster than inflation, and (with some exceptions), those things are all non-negotiable expenses. As those costs rise, everyone’s margin for error shrinks.

Across the economic spectrum, weirdness is getting squeezed. We, as a country, are running down our Strategic Weirdness Reserve. So, please: Form a secret society. Start a conspiracy. Build a weird contraption in your garage. Try out an all-celery diet. Our way of life depends on you.