The arrival of generative tools that can produce images and video has split the film industry into camps that talk past each other. One camp sees a tool that will liberate creators from the drudgery they have always complained about; the other sees a machine that will replace them with something cheaper and worse. Both camps are partially right, and the partial rightness of each is what makes the conversation so difficult. What the technology actually changes about filmmaking is narrower than the enthusiasts claim and more consequential than the skeptics fear, and the gap between the two is where the next few years of the medium will be argued out.

The things the tools are genuinely good at, today, are the things filmmaking has always disliked and always required. The shot that needs a thousand extras is expensive, and a generated crowd is cheap. The extension of a set that could not be built in time is a problem that once delayed releases; a generated extension is a few hours of computing. The removal of an unwanted object from a frame, the cleanup of a noisy plate, the generation of reference for a sequence that does not yet exist — these are the tasks the tools do well, and they are the tasks that consumed the budgets and the schedules of the films that came before. That the tools can do them is not a threat. It is a relief.

The line between drudgery and craft

The difficulty is that the line between the drudgery the tools relieve and the craft the tools threaten is not a line at all. It is a gradient, and the point on the gradient at which a task becomes craft rather than drudgery is the point at which reasonable people disagree. The compositor who spends a week cleaning a frame is doing drudgery, by one account, and craft, by another, and the tools that automate the one are the tools that devalue the other. The industry is discovering, awkwardly, that the work it once considered beneath its artists was, in retrospect, the work that employed them.

The more honest framing is not about replacement but about redistribution. The hours saved on the tasks the tools can do are hours that can be spent on the tasks they cannot — the judgment, the taste, the decisions that distinguish a film from a sequence of generated images. The studios that treat the tools this way, as accelerants of the parts they already valued, are the ones whose work is improving. The studios that treat them as substitutes for the parts they resented paying for are the ones whose work is deteriorating, and the audience can tell the difference even when the critics cannot.

The question the technology cannot settle

The deeper dispute is not about what the tools can do but about what they were trained on, and this is the part the debate tends to elide. A model that learned to generate images by studying the work of artists who did not consent to teach it is, whatever its capabilities, a model built on a taking, and the films that use it inherit the question. The industry that spent a century establishing who owned what image, and who was credited for what contribution, is being asked to accept a tool whose provenance is deliberately opaque. Some will. Many will not, and the audiences who care about the distinction are growing rather than shrinking.

The resolution will not come from the technology, which has no opinion on its own origins. It will come from the agreements the industry negotiates, the standards the guilds enforce, and the choices the audiences make about what they will watch and what they will reward. The tools will improve. The questions will not get easier, and the studios that pretend they will are the ones that will be caught, years from now, on the wrong side of a settlement they assumed would never arrive.

The craft that remains

What the tools do not change is the thing that has always made films worth watching: the decision about what to point the camera at, and why. The generated image, however convincing, does not choose itself. The story that holds across a hundred minutes is not a sequence of impressive frames. It is a structure of intention, and the intention comes from the people who made the decisions the tools executed. The filmmakers who keep their craft are the ones who treat the tools as instruments of their intention rather than replacements for it. The films that result will be faster to make, and some of them will be cheaper, and a few of them may even be better. The ones that are better will be the ones made by people who understood that the tool was never the point. The intention was always the point, and it remains, for now, irreducibly human.