thirteen years
Grand Theft Auto VI made $3.79 billion on its first day, by Rockstar’s count. It came out thirteen years after the last one, and was delayed twice on the way. The same week, developers were telling CNBC that agents write more code than anyone reads. Those two stories are about the same thing: what happens to the price of time when speed gets cheap.
Thirteen years.
That’s the gap between Grand Theft Auto V, released on 17 September 2013, and Grand Theft Auto VI, released on Thursday. On Friday Take-Two said GTA VI had made $3.79 billion in first-day retail sales and sold 32.5 million copies before release. It called that “the biggest entertainment launch in history.” Those are the company’s own estimates, pre-orders included, of what players spent, not what Take-Two keeps. Sensor Tower’s outside estimate was lower, about $2.36 billion on 24.5 million pre-orders. Either way it’s several times GTA V’s first day, which CNBC puts at roughly $800 million. CNBC also reports an estimated $1 billion or more in production costs.
It was late, and Rockstar said so twice. The original window was 2025. Last year it slipped twice, first to 26 May 2026 and then, in November, to 19 November 2026. “We are sorry for adding additional time to what we realize has been a long wait,” the studio said, “but these extra months will allow us to finish the game with the level of polish you have come to expect and deserve.” Bloomberg’s Jason Schreier reported after the first delay that nobody he talked to at Rockstar had thought 2025 was a real date, and that management appeared to want to avoid brutal crunch.
Remember that last part. Crunch is what a studio does when it wants the calendar to move faster than the work: the same months, more hours, more people. Rockstar did the opposite. It kept the hours and bought more months.
the sugar has to melt
Fred Brooks said it about software in 1975, in The Mythical Man-Month: “The bearing of a child takes nine months, no matter how many women are assigned.” Some work is sequential. Adding effort doesn’t shorten it.
Henri Bergson made the same point nearly seventy years earlier, and his version is better. In Creative Evolution (1907), he wrote: “If I want to mix a glass of sugar and water, I must, willy-nilly, wait until the sugar melts. This little fact is big with meaning.” The waiting, he says, “coincides with my impatience.” It’s lived time, not clock time, and you can’t “protract or contract” it as you like. Some things are only finished when they’ve had the time.
The same week GTA VI shipped, CNBC ran a piece on developers who miss the struggle. A Stack Overflow survey has 83 percent of developers using at least one AI tool. James Broadhead, the CTO of Taktile, told CNBC the agents are “spitting out so much code that [engineers] haven’t read it all anymore.” That’s what the other end of the curve looks like. Writing code, which used to be the slow part, mostly isn’t anymore.
When output gets cheap, the price moves to whatever can’t be hurried. I think that’s what Thursday was. I don’t know how Rockstar spent the extra year, and nobody outside does yet. But a game like this isn’t one big asset. It’s thousands of small ones that each needed real elapsed time: a city looked at by enough people for long enough that the corners hold up, a world tuned past the point where any schedule would have stopped. Money can buy more people working on that. It can’t buy the months, which is exactly Brooks’s point.
The 32.5 million pre-orders say the same thing from the other side. People paid $80 or more for a game they hadn’t played, because the studio had shipped GTA V, which has now sold more than 230 million copies according to Take-Two. That money wasn’t buying a product. It was cashing in thirteen years of a reputation for not shipping until the thing was done. You can’t speed reputation up either. It only comes from shipping things that were worth the wait, over and over.
My hypothesis is that this is where value is heading in general, not just in games. As generation gets cheaper, anything that can be produced quickly gets copied quickly, and its price falls to what the quick version costs. What holds its price is whatever took elapsed time: a world, a codebase that’s survived contact with real users for years, a name people will pre-pay. Rockstar had an unusual luxury, too. GTA V kept selling the whole time it waited. Most of us can’t afford thirteen years. But the market just showed, at a scale nobody can wave away, what waiting is worth.
the part that took time
Here’s where it lands for you. Your review this year probably measured some kind of velocity: tickets closed, PRs merged, features shipped, time to first draft. Those numbers were always a proxy. They’re getting worse as one, because the fast part of your work is exactly what an agent can now do as fast as you, or faster.
So it’s worth taking an honest inventory of the part of your work that couldn’t have been hurried. The system you’ve run long enough to know where it lies. The domain you understand because you sat through three years of customer calls. The people who trust your estimate because you’ve been right before. None of that shows up as velocity, and all of it is the sugar. If everything you produce could have been produced in a week, that’s the problem, not the speed.
And if you can, keep one thing on your plate that runs on a longer clock than the sprint: a tool you’re building properly, a migration you’re doing right, a piece of writing you’re not rushing. It won’t look productive in any given quarter. That’s how you know it’s the right kind of work.
Rockstar apologized twice for the wait. Then it sold $3.79 billion in a day. The apology was for the delay, and the money was for what the delay made.
Sources: Take-Two: $3.79 billion, 32.5 million pre-release, “biggest entertainment launch” ↗ · AP via ABC News: company estimates, $80, GTA V’s 230 million ↗ · CNBC: Sensor Tower, GTA V’s $800 million, $1 billion production cost ↗ · Engadget: the 2025 window and the first delay to May 2026 ↗ · VGC: Rockstar’s delay statement, Schreier on crunch ↗ · CNBC: some developers miss the struggle ↗ · Brooks, The Mythical Man-Month, ch. 2 ↗ · Bergson, Creative Evolution, trans. Arthur Mitchell (1911) ↗