tl;dr “As short as possible” often doesn’t mean actually short. If you want to run a marathon, you can’t do it in less than 26.2 miles; many goals simply cannot be achieved without a substantial expenditure of time, and the default assumption that there must obviously be some way to get the goods without the cost often strikes me as misguided and more than a little bit entitled.
One of my recent posts was 21,000 words. That’s pretty long! It’s over an hour of reading, for most people.
(Already, my brain hiccups. I pause to ask myself: is two hours of reading actually a long time? It certainly hasn’t always seemed so, and I’m not actually sure that past-me was wrong about that. There have been years of my life where I read long-form content for more than three hours every single day, and those were good years.)
One of the places I posted it was LessWrong. People on LessWrong care about being efficient. So it was not surprising to me to see a comment whose entire substantive content was ~“I kept not reading this because it was so long, so I put it into an LLM and here’s what the LLM says it says: [bulleted summary].”
I deleted the comment.
On January 29th, 2023, I posted the following to Facebook:
As time goes on, I have less and less patience for the tl;dr crowd. Like, I dunno if Twitter is cause or effect or both (probably both) but it's emblematic of the endumbening and oversimplifying of thought, along with the implicit “prove to me that I should read this” entitlement mentality—
(I have no interest in proving it to you; fuck off if you have the attention span of a goldfish; you are wrong to consider yourself a part of my intended audience)
—and that whole deal is bad.
People are more than welcome to choose to prioritize their time. People are more than welcome to not find something interesting, for any reason whatsoever, including its length.
But like. When I write down a complicated thought, every single time I get people demanding that I effortfully make it non-complicated for them, or else ~“I'm taking my business elsewhere,” and there's something weird about this dynamic in that it happens WAY more often than e.g. people demanding that there be more math in the poetry or more poetry in the math or more sports in the political discourse or more non-violence in the boxing match.
Like, length/complexity in particular seems to be a place where people think that it should be correct for everything to be twisted around to cater to their particular preferences and my soul’s answer is “fuck off”
(I don't go on Twitter and demand that people make their thoughts nuanced and complex.)
Being a tiny bit more charitable: it’s obviously not always entitlement. I’m sympathetic to people wanting to know a little bit more about a Thing before they commit to engaging with it—
(although Truth or Dare did actually start out with a summary, and that didn’t stop me from having to delete a different comment saying “this should start with a summary”)
—and it seems fine to ask, in an open, curious sort of way, “hey, does there happen to be a faster way to get there?” That’s fine. It makes sense to treat one’s time and attention as a precious resource, and it makes sense to want things to take no more time than they have to—
(Again my brain hiccups. Does it? Do I actually want the song Launch to be less than seven minutes and fifty-one seconds? Do I think it makes sense for someone else to want that?)
—but in practice, the open curious asks seem to me to be substantially in the minority.
In my experience, the aggressive tl;dr crowd is frequently wrong. They’re wrong about what’s possible, they’re wrong about what’s good for them, and they’re very often wrong about how to achieve their own goals.
[Insert essay on fabricated options]
[Insert essay on blindspots color blindness]
[Insert essay on irreducible complexity and the misleading nature of many short statements]
I see two broad categories of wrongheaded tl;dr’ing:
1. Situations where the activity is the point.
There are things we spend time on not because we have to, in order for some process to run its course, but simply because it’s actually worth spending time on them. Experiences that are valuable in and of themselves, moments worth savoring. Art. Connection. The slow scrape of a fingernail down your back. A piece of fine chocolate dissolving on your tongue.
(I don’t want my hugs at 3x speed!)
Most of what we treasure in a sunset is missing from a summary, and sometimes, the stuff I write is that sort of stuff. I think it’s good, actually, for people to just kind of be inside of my thoughts from time to time, and I bristle at the implication that further justification is required.
(I’m not begging people to be here. If you have somewhere better to be, go be there!)
2. Processes that require time.
There are plenty of metaphors here. “There’s no shortcut to building muscle; you have to actually put in the reps.” “Baking a cake at 3x the temperature for ⅓ the time will not work.” “The tl;dr on learning Mandarin is a list of the hundred most commonly used phrases, but that list will do you no good unless you spend weeks or months actually working with it.”
Some kinds of cognitive tasks can in fact take place approximately instantaneously; revelation and epiphany are real things.
But most of the changes a person can make to their mental landscape are much more like digging a trench than like breaking a dam.
(“But I did tell you that I love you! I told you way back at the beginning, on like our sixth date!”)
I used to be the curriculum director for the Center for Applied Rationality, which was an organization where at-least-some-of-us were trying to figure out Ways To Actually Help People Think Better. We’d put on intensive, four-and-a-half-day sleepover workshops in which we taught the participants a bunch of mental techniques and had them apply those techniques to their problems and goals, and as a result, some people’s lives actually changed for the better, at least a little bit, sometimes even in ways that would stick around for years afterwards.
(A small number of people. Four and a half days just isn’t a lot of time, to wear new grooves into a mind.)
I also frequently found myself in a position to give one-day crash courses, or hour-long summary talks. These were unfailingly well-received. People loved them.
(There’s one where I’m completely naked, if you’re into that sort of thing.)
Once in a great long while (though noticeably less often than with the workshops), I’d hear from someone that their life was better, downstream of the abridged version. Rad! But when I double-clicked, it was invariably the case that the growth actually occurred after my brief talk, not during it. Those talks were like handing out maps—the only people who ended up meaningfully benefitting were the ones who went out and actually walked the trails.
(Sometimes, I would hear from people who never attended a workshop at all, but who benefitted from diligently working their way through the handbook. Nobody ever messaged me to say they got value out of reading the table of contents.)
In other words: I deleted the AI summary of my Truth or Dare post because it was a trap. There are posts where the delivery mechanism is conceptual; the core idea is the content, and if someone can manage to deliver that core idea more efficiently, they’re doing us all a favor.
But the delivery mechanism of a post like Truth or Dare (or In Defense of Punch Bug, or Scott Alexander’s Meditations on Moloch) is osmosis. There is no way to get the value without putting in the time. Your options are “steep yourself in it,” or “don’t read it, and preserve your time and attention for other pursuits.”
Both of those are good! There’s nothing wrong with not reading an essay…
…just like how there’s nothing wrong with choosing not to listen to a symphony, or choosing not to do the-number-of-pushups-it-would-take to reshape your arms and chest.
Make the choice that makes sense to you, and try not to resent the world when it doesn’t offer up a shortcut.
Footnote: Two instances of the word “less” in the above made me twitch, but on reflection I (reluctantly) think they’re both correct. Even though miles, minutes, and seconds are discrete, length and time are continuous; it’s less [distance] than 26.2 miles and less [time] than seven minutes and fifty-one seconds.
I think the steelman of the tl;dr is an abstract. It has the problem, the solution, and the implication that if you really want to understand you need to read the whole thing.
There's a piece you didn't really touch on here - I'm often impatient with long-form content because it (understandably) is for a general audience, and doesn't know or care about which parts will be novel or interesting to me personally.
I'm not really sure how to solve this though - reading stuff meant for a smaller audience helps (but has echo chamber issues), and maybe an AI with a lot of context on me personally could do it, but for now it seems like just the nature of the thing.