The future is here, and it's human

When photography was invented in 1839, painter Paul Delaroche declared, "From today, painting is dead." 

In 2024, when I first saw AI films of a decent level, I felt the same way. It's over, I thought. Filmmaking as we know it is dead.

It's over, I thought. Filmmaking as we know it is dead.

Many filmmakers shared my sense of shock. We’d forged our careers on the assumption that live-action films would always be made by travelling to a location with a crew, pointing a camera at some humans and pressing record. Now, it seemed, a 'filmmaker' was someone who stared into a computer and entered prompts. 

Need I say, I didn't want to be that filmmaker?

Above: The invention of photography caused many painters to sound the death knell for their own art form.

I felt grief. I’d spent my entire career developing myself as a director and writer, and now it seemed a machine was about to do it better. This wasn't just a threat to how I made a living, it was a threat to my whole identity. If I wasn't someone who wrote and directed films, then who was I? I waited anxiously for the next advances in AI-filmmaking. And in 2025, they came.

Live-action advertising has long been a place where technical innovations are trialled before they reach Hollywood: Steadicam, slow-motion and macro photography all found their footing here before movies. And so it's natural that the first big swings at using AI in films were taken by brands. Throughout 2025, high profile ads by Coca-Cola, Google and McDonalds were released and… they bombed. They faced intense online backlash, and were swiftly pulled.

Filmmaking isn't just about the image you make, it's also about the audience who watch it.

Call me perverse, but this is when I felt the first shoots of optimism. Here’s what I think these AI failures taught us: filmmaking isn't just about the image you make, it's also about the audience who watch it. Cinema, and its offspring, live-action advertising, is a contract between the filmmakers and the audience, and that contract is built on trust and grounded in reality. 

Nobody who makes animated films pretends their characters are real people. But when you make a live-action AI film, you present pixel-people and hope the audience will feel for them as if they were human. Perhaps audiences aren't as willing to play that game as Silicon Valley hopes.

Above: The spectacle of David Lean's Lawrence of Arabia is all the more spectacular for the scale of its human endeavour. 

Humans are social animals. We’re interested in what other humans do. Movies give us a way of observing other humans, humans who find themselves in scenarios we might never encounter in our own lives. Films are powerful when they allow us to viscerally experience the life of someone else. It’s always been this way. 

When 1920s film goers saw a building collapse around Buster Keaton, the thrill they felt was because Keaton was really risking his life. When in the 1960s they watched David Lean's Lawrence of Arabia, they felt awe when they witnessed the charge of hundreds of real horsemen, camel riders and footsoldiers. Even a shot that is as simple as Timothee Chalamet staring into the fire and weeping at the end of 2017's Call Me By Your Name garners its strength because you feel a real person, living a real emotion. 

Films are powerful when they allow us to viscerally experience the life of someone else. It’s always been this way. 

That is the magic of cinema. Live-action advertising, as a child of cinema, follows the same rules. When Jonathan Glazer showed us horses jumping on the crest of a wave for Guinness Surfer in 1999, the sense of awe came, in large part, from wondering how on earth such a shot was made.

What happens when each of those shots can be faked using technology? They lose power. Tom Cruise knows this. He knows that although stunts can be faked, the thrill the audience feels is heightened immeasurably if he does them for real. And as for horses jumping out of waves? If you wanted to do that now, someone would suggest AI. If you used it, would anyone be talking about Guinness Surfer even a month after it was made, let alone twenty-seven years later?

Above: Guinness Surfer shows us the magic of filmmaking. 

I think the palette of elements that can go into making a film are like stocks. The more that the market is flooded with a stock, the less it is worth. AI is flooding the market with quick edits, montages, flashy transitions, beauty shots, varied locations, crazy camera angles, stunts, special effects, epic scale and weird surrealism. These elements are plummeting in emotional value: audiences will care for them less and less. In February 2026, an AI video of Tom Cruise and Brad Pitt fighting on a rooftop emerged. The Napoleon and Gladiator 2 screenwriter David Scarpa had this to say about it; 

When spectacle is easy to produce, it means nothing to watch.

"The real significance of the Brad Pitt vs. Tom Cruise video is that we are fast approaching the point where the cost of producing empty spectacle is zero. Visual imagery that once cost $250,000 per shot and thousands of man-hours to produce in CGI will be as plentiful as air, or water. We will soon be swimming in it, and its cultural and economic value will decline accordingly."

Scarpa's point cuts to the heart of it: when spectacle is easy to produce, it means nothing to watch. What rises in its place are the things AI cannot do well (yet?... ever?). Long shots, dialogue, subtle facial expressions, the nuance of people interacting, shots that capture a sense of real life, well written scripts. In other words, the work that is the most unmistakably human.