Deflation \ stacker news ~mostly

pull down to refresh

3862 sats \ 13 comments \ @elvismercury 22h mostly_harmless

@k00b recently posted this article which is worth a read. But as I said in a comment there, (and as he also basically said) the article hits the same point over and over. Instead of an article, it could have been a brief bulleted list. (How meta.) This isn't a new phenomenon, of course (may I introduce you to airport business books?)

It would have been heretical even a year ago, but I'm starting to change fundamentally in how I engage with information. I'm notable irl for how slow I am at stuff; I talk slow, I read glacially slow, I ruminate. Even for material that's not that important, my behavior was to suck all the juice out of it that I could.

But I finally get that that's dinosaur strategy The asteroid has hit. The game theory on attention has it that what you encounter, in meatspace or out of it, is mostly sponge cake. Funny that LLMs are the great equalizers in how they take outlines or malformed gibberish and inflate a well-written 500%, and now my job, our job, is to let the air out and re-extract the outline. The realization of an existing trend.

So I "read" this article flicking through it, glancing at the section headings (at least the author did a nice job of organizing them) and was left with a thoughtful paragraph to think about. That's not nothing, but it's a different style of being in the world that I've established in the course of my life.

This seems like the way, like it or not.

view all related items

306 sats \ 2 replies \ @Undisciplined 22h

I'll propose that you've simply adapted a new strategy for sucking all the juice out. A strategy that recognizes that some fruits don't have much juice to begin with and you need to get the pulp out of the way.

42 sats \ 1 reply \ @elvismercury OP 21h

I'm like some weird South American monkey mowing through the gross fibrous fruits.

151 sats \ 0 replies \ @Undisciplined 21h

201 sats \ 1 reply \ @Scoresby 22h

I suppose it's true that the paragraph words in articles are often there so the headings don't look naked. At least, "if it wasn't worth putting in the heading, why should I bother with it" is a plausible, if slightly unreasonable, subconscious path.

Certainly headings (and the increasing frequency at which they occur in writing) is something new(ish) with which both writers and readers are learning to contend.

I'm writing something just now. My routine is to write a paragraph or three and then return later to find some catchy summary for a heading. Is it the case that the paragraphs are mostly data and citation, while the headings contain the most digested analysis? Possibly.

Perhaps I should abandon headings altogether in favor of single sentence paragraphs a la Tractatus.

202 sats \ 0 replies \ @elvismercury OP 21h

It's interesting how much work a writer can do for the reader, and the trade offs that different tech brings to bear - if you're in dandelion world, where there's trillions of things to read and write, taking great care is generally a bad investment. If you get one shot, you polish it like a gem.

I hadn't considered that as both a reading and writing strategy. And LLMs are a new thing entirely.

I like the "headings as proof of work" idea.

101 sats \ 2 replies \ @antic 17h

I think about this every time I write a substack--if I were more of a manipulative entrepreneur who wanted to, I could extend this (intentionally brief) article: https://antic.substack.com/p/forget-universal-basic-income-embrace into a 200 page book called "Universal Basic Resources: Building a post-scarcity civilization, one free resource at a time." and flush it out with historical examples and references, and maybe I'd be doing podcast circuits as a thought leader on economics, but I feel like most things should just be casual zeitgeist seeding conversational shares and not published books.

Also reminds me of a buddy I had decades ago who was a writer, and we talked a lot about writing process, and it came up that what a writer doesn't say and leaves for the reader to figure out can be more powerful than just laying it all out there in bulk words.

0 sats \ 1 reply \ @elvismercury OP 11h

Under what circumstances would you think it worth your while to write the 200 page version? That's been on my mind a lot recently.

Have you heard of Hemingway's iceberg? If not, worth a google, you will dig it.

101 sats \ 0 replies \ @antic 6h

Part of the blocker is that I don't have any desire to publish a book like that so it doesn't seem worth my time to write it out in that much depth/detail. If I were going to spend that much time writing, I'd finish my book of short sci-fi stories that's still incomplete. I tend to write sub-stacks on topics that pop up frequently enough in conversations that I get sick of making the same arguments over and over and instead want a place to point someone to a place where I've already written my thoughts on it. Trolley Problem: check, UBI: check, Quantum attack on Bitcoin: check. I don't have to write a text bomb when someone brings it up, I can just send them a link and if they don't read it, at least the conversation is over :)

I've got a couple of things unpublished that are more generalized essays/thoughts but in the past I've deleted those from medium or other places because I don't see much value in sharing that much with the wider world...

101 sats \ 1 reply \ @k00b 17h

This is how I’ve always read online, unless I’m reading someone that’s incompressible.

I find myself reading this way more efficiently lately, trying to source relevant goings on for sharing here, annoyed by how most other sharers don’t bother to at least discover the ledes because they don’t care enough to read more than the headline and annoyed by writers and “writers” that debase the art.

0 sats \ 0 replies \ @k00b 17h

*debase the training data

217 sats \ 1 reply \ @seashell 18h

Used to read everything like gospel, now I just scan for signal and skip the noise. Most writing today’s just padding around one decent idea, LLMs broke everyone’s bullshit detector flooding everything with synthetic fluff. If an idea still makes sense after you compress it down that’s how you know it’s real. The best writing today feels like code comments

0 sats \ 0 replies \ @elvismercury OP 17h

Also interesting in that code is (to date) the most no-noise communication, with utterly concrete semantics: within its interpretive context, meaning is locked in. Comments on code would then anchor to this solid foundation - drifting upward from it, but still closely in orbit.

In "normal" writing what's the reality we tether to? Perhaps the haziness, the non-consensually of that has loosened writing. The falcon cannot hear the falconer, or will not.

0 sats \ 0 replies \ @Rothbardian_fanatic 13h

This is interesting! My writing habits for long form documents includes first writing an outline, using Word, then filling the outline out with the paragraphs of information, often using the outline phrase as the paragraph subject sentence. Since I write that way I seem to read that way, too. I become somewhat confused when I cannot discern the outline form of a document. Until now, I thought it was a deficiency in my reading of the article.