The moment will be cultural, the experience epic, and the game, I expect, a masterpiece.

It will also be the last of its kind.

Not because Rockstar won’t make another great game (it will). Nor because AI will make today’s games faster and cheaper (it already has). GTA VI represents the culmination of a model of creativity that AI is about to upend.

For more than three decades, artistic ambition in narrative video games has been measured by scale. Bigger teams built bigger worlds. Larger budgets bought denser cities, longer scripts, more for players to do. The achievement was creative, technological and increasingly organizational. Rendering a virtual city is a technological challenge. Populating it with ten thousand hand-authored details, on schedule, across three continents, is a feat of creative coordination.

With GTA VI, Rockstar orchestrated thousands of artists, programmers, writers and designers, working for about a decade in pursuit of a single creative vision. While Take-Two has never disclosed what GTA VI cost to develop, public estimates range from $1 billion to $2 billion.

Creative work often requires deletion. Novelists cut chapters they love. Filmmakers leave their best scenes on the cutting-room floor. Late in the development of Red Dead Redemption 2, Rockstar cut roughly five hours of content from a game developed by some 1,600 people. Five hours is roughly the length of an entire Call of Duty campaign. That is the nature of the process. It has to be created before anyone can know it doesn’t belong.

The bigger the creative team, the harder it is to erase, rethink and begin again. In game development, every important creative decision sets hundreds, sometimes thousands, of writers, artists, engineers, animators and audio experts in motion. And when those pieces come together, the original idea may no longer work.

At that point, changing direction becomes a systems problem. Abandoning an idea can mean unwinding hundreds of interconnected decisions across disciplines, teams and production schedules. The later the change, the more expensive it becomes. Creative decisions become financial decisions.

AI lowers the cost of changing your mind.

Game developers have always prototyped, tested and discarded ideas. What AI changes is how far an idea can be developed before a studio commits the full weight of production behind it. If a small team can build something playable and expressive enough to assess earlier, the expensive decisions can come later, when there is a much better sense of what works. As the economics change, the creative process begins to invert: less about what can be made, more about what is worth keeping.

Painting faced a similar inflection point nearly two centuries ago. When photography arrived, many visual artists feared for their livelihoods, and some were right: the market for commissioned likenesses collapsed.

Yet others asked a different question: What will we create now? Freed from the need to represent reality faithfully, they created new, even radical, art forms: Impressionism, Cubism and abstraction.

Something similar is likely with games. Photography freed painters from the obligation to reproduce the visible world. AI can free game developers from much of the cost of changing course.

I suspect cultural historians will one day look back on GTA VI the way art historians look back on the last great realist painters: the fullest expression of one creative model just before the boundaries of the medium broke open to reveal entirely new forms.

The next masterpiece rarely looks like a bigger or cheaper version of the previous one. It looks like something no one thought to make before.

The opinions expressed in Fortune.com commentary pieces are solely the views of their authors and do not necessarily reflect the opinions and beliefs of Fortune.