I: Curators Skim

Curators and gatekeepers, because of the nature of their work, default to skimming. It's easy to be annoyed that the editor might only read the first page (or the first paragraph, or even the first sentence) of a story you spent days, months, or years writing. But most of the time, the editor only needs a few seconds to gauge the quality of the writing, and whether it’s a fit for their audience.

[ This interesting dynamic sunk in after reading The View from RSS, a behind-the-scenes view of curation by the glamorous and shockingly intelligent Caroline Crampton, who describes what it’s like to live underneath a pile of two thousand RSS feeds, possibly including this one. Full disclosure: I am plotting to use the ideas in this article to get featured by The Browser again, preferably without having to spend 6 months writing a 5,500-word essay to do so. ]

Songwriters have shortened the average length of the instrumental intro, to get to the catchy chorus hook while the goldfish are still listening. The usual story is that attention spans are declining. But is it the audience’s attention spans that are dictating this change? Or, is part of it a systemic bias for things that front-load the value, and are easily curatable?

In some cases, there’s an incentive for curators to share things they haven’t deeply read: speed. If their heuristics are accurate, skimming-only allows them to be the first curator to break good articles, increasing their status as taste-makers.

II: Skim Milking

I realized underdog independent bloggers could use curators’ skimming against them. In the spirit of Gwern’s hacker mindset essay, we could optimize our writing for these curators, by tailoring the parts that stand out to specifically catch curators' attention.

Then I realized: this is what the popular writers do. Have you noticed all the James Clear-like life advice writers with their bold text, catchy statements, and soundbites? Is it a coincidence that the most popular writers use bold text and write eminently-skimmable articles?

The tricky part is how to deliver the best value for the end-reader at the same time as fishing for curators.

Note: after I wrote this post, but before I published it, Dynomight wrote a longer and probably better post about formatting. I only skimmed it.

III: Applied Antimemetics

If, for some reason, we wanted an article not to go viral, ala Nadia Asparouhova's anti-memetic self-keeping secrets, we could deliberately emphasize the boring parts. The exact same content, with different emphasis, could make it so that curators who didn't read the whole thing would think it was a badly written, boring article. But readers who read the whole thing would realize that it wasn't, or even, that the parts that weren't in bold directly contradicted the parts that were in bold.

In the spirit of adventure, I could publish two versions of this essay. One of them would highlight the interesting parts, and attempt to optimize for the skimming curators and gatekeepers. The second version would only highlight the worst and boringest parts, and optimize for the readers who read the whole thing, and specifically try to prevent the curators from discovering it.

IV: Holes in the Sieve

Certain articles or books take a while to get rolling. Steinbeck comes to mind: an entire chapter about the weather. It seems these would be at risk of being passed over by a skimming curator.

And some types of material resist shallow engagement. They seem boring, or dumb, or empty when you blaze though them. But, sometimes, they’re not!

Once a week I go through my comparatively modest collection of ~151 RSS feeds (some of which are, in turn, curated collections of links), as well as a few aggregators, and select my week’s reading material. Like Caroline, “I’ve become very good over the years at allowing my eye to slide over everything, stopping when I see a headline or phrase that looks promising.” The first step of triage happens in my RSS reader. For noisy feeds, I only read the headline. Or, I skim the article. I open the articles that seem interesting, and I save the best to Pocket/Wallabag/Instapaper/Readeck.

Arguments for and against things, self-help, and hot takes are easy to skim, and get a decent sense of whether they will be interesting. These are what I will call “point based” writing. They have a point, and it’s either right or wrong, and it’s either relevant or irrelevant to my life, and it’s either well written or not.

But there are a few categories of things that can get triaged when they shouldn’t.

Buried nuggets. Often, 98% of the value I get from an essay is a single idea or piece of information, which I haven’t encountered before, buried somewhere obscure — often in a footnote. Real-life example from yesterday: a Dwarkesh Patel interview with Gwern Branwen crossed my screen. I skimmed it briefly, and triaged it as “modafinil mode” content (the hyper-intellectual cottonwool prose of the Rationalists, which takes a lot of effort to understand for an uncertain payoff). Because the software I wrote for my ereader was buggy, the articles I chose for the week did not show up on my ereader, and the offending 16,000-word Gwern essay downloaded itself, along with a bunch of Amazon book listings and other random crot. I was bored, so I read it, along with some of the other triaged essays. Some of it was modafinil mode, and I skimmed, but two little asides made it worth it. I learned that like me, Gwern is broke, even though he’s a household name. And I learned that he isn’t into psychedelics, which totally broke my model of Gwern as the reckless pharmacological self-experimenter. (His argument nearly matched an unpublished essay I wrote, Experimentation and Downside Risk.)

Emergent payloads. Another type of material that breaks triage workflows is when the whole is greater than the sum of the parts. The piece, every part of which is mundane, creates a vibe, and something about the gestalt of all the boring subpieces is educational or moving. The Russian classics come to mind.

Quiet depth is also triage-resistant. Some people, like Wendell Berry, write in a very plain tone about ordinary things. It might not jump out at you. They don’t use the flashy literary tricks, and the pace skews slow. But if you pay close attention, and let the seeds they plant in your mind bloom, something extraordinary happens. Often, these seeds appear to be platitudes. Alone, they would bounce off the modern consciousness. But the rest of the essay fertilizes the readers mind for the small, commonplace seed. If you rush through these essays, the seed will not germinate. It will just fall on dry soil, and get laughed at, because it’s such a tiny and ordinary seed.

V: Dumpster Diving

To deal with un-skimmable material, curators follow other curators that keep tabs on specific beat. If a sub-curator with a good track record surfaces something that seems boring on the surface, chances are it’s worth a deeper look.

However, even if the multi-leveled curatorial juggernaut worked perfectly, there's another concern. The curators are trying to find the best articles for their audience. The more influential the curator, generally, the more general the audience. Everything they pick has to be good, and it has to be evenly good: palatable enough they don't lose their audience.

But you and I are not an audience. We are individuals, and our needs cannot be optimized for by a system of this scale. Curators are a band-pass filter. They remove most of the bad stuff, and some of the good stuff that can be confused for the bad stuff.

But I have found some real treasures in dumpsters and at recycling centres. You can't always judge the best by the mean or the median. I have built electric bicycles out of windshield wiper motors, even though the entire car they were removed from was dead and smashed up. Curators don't tend to serve us smashed up cars with cool components that still work.

That's why, in addition to The Browser and Hacker Newsletter, I subscribe to high-noise-to-signal uncurated feeds like HN Blogs. Sometimes I want to go to Tiffany’s; sometimes I want to hold my nose and go dumpster diving. The average is worse, but you find things you won't find anywhere else.