Every conversation about AI in filmmaking eventually arrives at the same anxious question: can a machine really make us feel something? Can it build a character we believe in — one whose grief lands in our chest, whose joy makes us lean forward, whose silence says more than dialogue ever could? It's a fair question. But here's the thing about that question: it has always been the question. Long before a single frame was generated by an algorithm, filmmakers were getting it wrong.
The emotional core of a story — the irreducible truth at its center — has never been easy to find, and it has never been a technology problem. It is a human problem. A problem of courage, of clarity, of knowing what a story is actually about beneath the plot mechanics and the genre trappings. AI is simply the newest arrival to one of cinema's oldest and most honest struggles.
The Myth of the Technical Solution
When sound arrived in 1927, studios panicked. Silent film had developed a profound visual grammar for emotion — the close-up, the cut, the held gaze. Suddenly there was dialogue, and with it came a wave of flat, stagey pictures in which characters stood still and talked at each other. The technology was thrilling. The humanity was missing. It took years for filmmakers to rediscover what emotion looked like in a world with sound.
The transition to sound produced some of the most emotionally inert films in Hollywood history. Applause (1929) and many of its contemporaries were technical marvels — and dramatically hollow. The craft of feeling had to be rebuilt from scratch inside a new medium.
The same story played out with color, with widescreen, with CGI. Each technological leap brought a period of spectacle divorced from soul. Filmmakers would reach for the new tool and, for a time, mistake the tool for the story. The audiences noticed, even when they couldn't name what was missing. Something wasn't true. Something didn't land.
"The technology changes the canvas. It does not change what it means to be a person standing in front of it."
Hollywood's Long History of Missing the Point
Consider how many big-budget productions — backed by the finest human writers, directors, and actors — have produced characters who feel like furniture. Films where plot happens to the protagonist rather than through them. Where the audience understands, intellectually, that a character has suffered, but does not feel it. Where the mechanics of story are dutifully assembled and the emotional core is simply… absent.
This isn't a rare failure. It is an endemic one. For every film that genuinely reaches inside a viewer and rearranges something, there are dozens that approximate the shape of emotion without ever igniting it. Franchises have been built on characters whose inner lives remain permanently opaque. Prestige pictures have won technical awards while leaving audiences unmoved. Beloved source material has been faithfully adapted — scene for scene, line for line — and emerged somehow emptied of the feeling that made the original matter.
This is not a criticism reserved for commercial cinema. Art films, independent films, international films — every tradition has produced its share of works that gesture toward emotional depth without achieving it. Finding the core is genuinely hard. The tools have never been the barrier.
The culprit, when critics and filmmakers do their honest post-mortems, is almost never the camera or the editing software or the visual effects pipeline. It is something quieter and more difficult: a failure to know what the story is really about, or a failure of nerve in committing to it fully.
What AI Is Actually Struggling With
When people criticize AI-generated film content — and there is plenty to criticize — the complaint usually circles back to a specific feeling: something is off. The faces are sometimes uncanny, the dialogue occasionally reaches for profundity and arrives at cliché, the characters move through situations without the sense that anything is genuinely at stake for them. There is production, but not presence.
This is a real limitation. But it is worth naming it precisely, because naming it correctly reveals how old the problem is. What AI struggles with is not unique to AI. It is the challenge of interiority — of making a viewer believe that a character has a continuous inner life that exists beyond the frame, that they want something badly, that they are afraid of something specifically, that the outcome of this story will cost them something irreplaceable.
"Interiority cannot be rendered. It can only be earned — through the accumulation of specific, chosen, true detail."
A character does not feel real because they are described as complex. They feel real because of the particular way they hold their coffee cup when they're hiding something, or the specific word they reach for when they're trying not to cry. Specificity is the technology of emotional truth, and it has been available to every filmmaker in every era. Many have still failed to use it.
The Shared Goal Has Never Changed
What's striking, when you set aside the novelty of the tools, is how consistent the aspiration has always been. Every filmmaker — from the silent-era pioneers hand-cranking their cameras to the teams now prompting diffusion models — is trying to do the same thing. They want to capture something true about what it feels like to be alive. They want an audience to sit in the dark and recognize themselves.
Ingmar Bergman said that film is a dream, and that it speaks directly to our emotions, bypassing our intellect. He was describing the goal. He was not describing something automatically achieved by the act of pointing a camera. His own films are filled with the evidence of hard-won choices — choices about what to show and what to withhold, about silence and the specific texture of a face in grief. Every one of those choices was in service of the emotional core he'd identified. He knew what his films were about, and he had the courage to stay inside that truth even when it was uncomfortable.
That's what the great films have always done. And that is precisely what the promising AI-generated films of the future will also need to do. The tools will improve — they always do, and faster now than ever. The resolution will sharpen, the motion will become more fluid, the dialogue will grow less stilted. But none of those improvements will locate the emotional core of a story. That work belongs to whoever is making the story. It always has.
A More Honest Conversation
If we're going to talk honestly about what AI needs to learn in filmmaking, we should also be honest about what traditional filmmaking has never quite mastered. The bar is not perfection. The bar is not even consistency. It is the occasional, hard-won, genuinely moving moment — the kind that makes an audience forget they are watching a screen.
AI will produce some films that miss the emotional core entirely. So does Hollywood, regularly, with enormous budgets and decades of accumulated craft. AI will also — inevitably, as the tools and the human intentions behind them develop — produce work that does land. Work that finds the thing underneath the plot, the feeling beneath the dialogue, the person inside the spectacle.
When that happens, it will not be because the algorithm got smarter in some abstract sense. It will be because someone, somewhere in the process, knew what the story was actually about — and had the clarity and the courage to build everything around that truth.
That is the only goal. It has never been anything else. Welcome to the oldest problem in cinema.
The tools change. The question doesn't.