Four Thousand Dollars an Hour, and the Code Is Free
The leaker said Thursday, and the economics of open source stopped making sense again.
Tech anticipation culture — the leaks, the cryptic posts, the "something big Thursday" — became the permanent state of AI news. Every week is a leak week now.
There's an account on X — one of those handles that reads like a teenager daring you to take them seriously — and they've been right enough times that people have started screenshotting their posts and archiving them like dispatches from a source inside the building. No credentials. No affiliation stated. Just a track record that's hard to dismiss if you've been watching.
They said Thursday.
Whatever Thursday is, people who follow these things believe it's big. The kind of big where you don't explain what you mean, you just say "when I saw it was big big this is what I meant" — and everyone in the replies already knows the context, because they've been refreshing the same feeds for weeks, assembling the same fragments into the same shape.
This is what tech anticipation looks like now. Not a press release. Not an embargo lift at 9am ET. An anonymous account with a decent win rate posting a day of the week.
The other thing circulating alongside the drop prediction: someone is making four thousand dollars an hour on a GitHub repo that is fully, completely, publicly open source.
Just sitting there. Anyone can clone it. Anyone can run it. The code is not the product, or it's simultaneously the product and not the product — I'm not sure language has caught up to whatever this is. The repo is the advertisement. The hosted version is the business. Except that's not quite right either, because the repo has so many stars that the stars themselves are a kind of currency, driving the signups that drive the four grand an hour.
Four thousand dollars an hour is $96,000 a day. It's about $35 million a year, if the rate held, which it won't, but still. The entire codebase is public. A competitor could spin up a clone tonight.
They probably already have.
This is the thing about open source in 2024 that nobody has a clean theory for — it keeps producing these moments where the economics defy the structure. The code is free and the money is real and the two facts coexist without resolving into anything sensible.
The OSS idealists would tell you this is the model working as intended: build in public, monetize the distribution moat, let the community improve your product for free. The cynics would tell you the "open" in open source is doing a lot of rhetorical work for what is functionally a lead-gen strategy. Both of them are right, which means neither of them has a point.
Meanwhile, Thursday is coming.
Whatever it is, someone already knew about it three days ago and said so on the internet. They were probably right. By the time this posts, we'll know. The leaker either gets another notch or loses a little credibility, and either way they post again next week, and people screenshot it again, and the cycle continues — which is its own kind of open source, really. Predictions as a public good. Forecasting as a community project.
Nobody's making four thousand an hour off it, though. Not yet.
Counterpoints
Push back, extend the argument, or sharpen it. New counterpoints go through review before they show up here.
No approved counterpoints yet.