There’s a pattern I keep running into in the rooms I find myself in. It usually shows up first as career advice, sometimes as a side-hustle pitch, and it goes like this: you pick a popular platform with a lot of activity, you scrape or paraphrase whatever’s getting attention there, and you put it somewhere you control with ads alongside, and a fraction of the attention that was flowing to the original ends up flowing to you. The wrapper changes depending on who’s selling the course (content arbitrage, passive income, the AI advantage), but the underlying move is constant.
For a long time the version of this that bothered me most was the website-shaped one. Personal blogs got strip-mined by content farms; forums got strip-mined by listicle sites; recipe blogs run by people who actually cooked got buried under aggregator domains that had stitched a hundred recipes together. The pattern is older than the web, but the leverage that ranking algorithms gave to a person willing to operate fast and cheaply turned it into a kind of cottage industry. That part has been written about at length, and I’m not going to add much to it.
The same move runs across more or less every medium where attention is mediated by an automated system, not just websites. The logic is the same, and so is the algorithmic blind spot that lets it work.
You can see it most clearly in the things aimed at children. Open Twitch and search for any popular game and you will find “livestreams” that are not livestreams; they are loops of pre-recorded gameplay set up to look live, often with a chat overlay promising free Robux or skins or coins to anyone who clicks the link in the description. The viewer count is real, the kids are real, and the attention being captured is real, but no one is streaming. The operator running the loop is not playing anything; they are sitting somewhere far away, having spent ten minutes setting up a script that funnels children towards account-credential phishing and survey-scraper sites under the cover of a fake person engaging with them. YouTube is full of the same idea in slightly different shapes: automated supercuts of popular streamers, AI narrators reading scraped Reddit threads over Minecraft footage, short-form ragebait so optimized for algorithmic surfacing that you feel a little stupider after three of them in a row.
The operator running the play, in any of these shapes, is not producing anything that ought to repay the attention they are extracting. They are building a structure that sits between an algorithm and a viewer, intercepting whatever the viewer was looking for, capturing whatever can be captured, and leaving. The work, where there is any, is finding a niche where the algorithm doesn’t notice and the audience can’t tell. That work is now nearly free, which is why we are seeing so much more of it.
It is not, to be clear, a moral problem in the narrow sense. Most of these operators are not stealing in any way you could prove in court, and most are not doing anything platforms have explicitly forbidden. They are doing something more general, which is taking our time and giving us back something worth less than what we paid. For every recipe-aggregator click, there is a person who actually cooked a recipe and didn’t get the click. For every ragebait view, there is a video someone actually thought about that didn’t get the view. The fake livestreams are worse: there is a child being told that someone is talking to them when nobody is, while a script harvests whatever it can from them in the background. Aggregated across a population of millions, the cost is enormous, and we are paying it out of our attention, which is the only thing online that is genuinely scarce, and a budget we do not usually think to defend.
I find this offensive in a way I don’t always know what to do with. The kids being deceived by a fake person while a script lifts what it can from them, the writers and creators whose work is being silently siphoned, the rest of us getting a worse internet because so many people have figured out that wasting our time is a viable business: none of those costs land on the operators running the play. They land on us, and we don’t get to vote.
The arrangement is bad, and the arrangement is also not inevitable. The reason this pattern works is that the systems mediating attention (search rankers, recommendation algorithms, livestream directories, every recommendation feed in your life) cannot reliably tell whether a thing is original, or whether a stream is live, or whether a video has any value at all. Those are answerable questions, technically; we have not built systems that answer them, in part because the platforms operating the rankers do not carry the cost of getting the answer wrong. The viewer carries that cost, and the viewer doesn’t have a vote.
Asking the operators to stop won’t work. You cannot tell people whose income comes from running this play that they should stop, and have it stick; the spreadsheet pencils out for them, and as long as it does, someone else will take the slot if they don’t. What’s needed is technical infrastructure that makes the answers to “is this original,” “is this live,” and “is this worth your time” legible to the systems that decide what gets surfaced, and to the people on the receiving end. That is a buildable thing, not a metaphysical problem; the reason it has not been built is that the people who would benefit are not the people who run the platforms.
All of which is a longer conversation than this post wants to have, and most of what I have been thinking about for a couple of years now. The pattern of intercepting attention without doing anything that earns it has spread well past the web of fifteen years ago, and we are all paying for it in time we did not choose to spend. The right response is not to shrug and treat it as the normal shape of the internet. It is a missing column on a spreadsheet that the market has decided someone else can pay for, and we could, if we cared to, build the column ourselves.