AI Won't Fix Your Bad Story

April 16, 2026 · 10 min read · Opinion

Scorsese Saw It Coming

Martin Scorsese, one of the greatest living filmmakers, could not get Hollywood to finance The Irishman. The director of Taxi Driver, Goodfellas, Raging Bull, The Departed. Hollywood said no for eleven years.

Producer Gerald Chamales said: "The exhibitors, the people who own the movie theaters, said they couldn't make money on it. Since they didn't want it, we were dead in the water."

Netflix eventually funded it. The Irishman received ten Oscar nominations.

In 2021, Scorsese wrote that cinema is being "systematically devalued, sidelined, demeaned, and reduced to its lowest common denominator: content." He said films today are "market-researched, audience-tested, vetted, modified, revetted and remodified until they are ready for consumption."

He said this before ChatGPT. Before Midjourney. Before any filmmaker had generated a storyboard with AI. The problem was already there.

The Numbers

The 2025 domestic box office totaled $8.87 billion, down 22% from 2019's $11.4 billion.

Here is where some of the money went:

• Disney's live-action Snow White remake: around $250 million budget. Lost around $170 million.
• Mission: Impossible – The Final Reckoning: $400 million budget. Grossed $598 million worldwide, far short of the roughly $1 billion needed to break even theatrically.
• Tron: Ares: $220 million budget. Lost $133 million.
• Elio (Pixar): $150 million budget. Lost around $100 million.
• The Alto Knights (Robert De Niro): $45 to $50 million budget. Made $9.6 million.
• After the Hunt (Julia Roberts): $70 to $80 million budget. Made $9.4 million.

These are not small indie films that quietly disappeared. These are the biggest studios, the biggest stars, the biggest budgets. And they lost a combined total of well over half a billion dollars in a single year.

But here is the part that matters: some of these were sequels and remakes. Some were original films. Both failed. Being original did not save Alto Knights. Being a franchise did not save Snow White. The common thread was not whether the film was original or a sequel. The common thread was that none of them had a story the audience needed to see.

The Safe Bet Does Not Exist

For years, franchises and known brands were the foundation of the studio business model. That made sense: building a new brand from scratch is expensive and unpredictable. Some sequels are great films. The Godfather Part II. Aliens. The Dark Knight. The issue was not that studios invested in franchises. The issue was that they stopped investing in anything else.

Look at how the top of the box office changed. In 2000, the ten highest-grossing US films included Gladiator, Meet the Parents, What Lies Beneath, Dinosaur, and Scary Movie. Roughly half the list were original stories. In 2025, the top ten was almost entirely sequels, remakes, and franchise installments. The only original film to break through at that level was Sinners.

In 25 years, the industry went from a mix of originals and franchises at the top to almost no originals. That shift predates AI by two decades. It happened because studios followed the data, and the data said sequels were safer. Until they weren't.

CNBC reported in January 2026 that "the old movie sequel trick is falling flat." Marvel's Captain America: Brave New World and Thunderbolts both lost tens of millions. Even Wicked: For Good and Avatar: Fire and Ash underperformed expectations.

The formula stopped working. Not because audiences stopped going to the movies. Because audiences got tired of watching the same thing.

But going original did not automatically fix it either. Alto Knights had Robert De Niro. After the Hunt had Julia Roberts and a Venice Film Festival premiere. The Smashing Machine had Dwayne Johnson. Big names, original scripts, and they all flopped. A-list talent is not a safety net any more than a known franchise is.

There is no safe bet. There never was. The industry just convinced itself there was one.

None of this is to say stories are the only problem. The theatrical decline is also about streaming, COVID-changed habits, rising ticket prices, and an ecosystem that gives audiences more reasons to stay home than to go out. Those are real structural issues. But those are distribution problems. The stories problem is on top of those, and it is the only one filmmakers can actually control.

What Actually Works

In the middle of all this, some films did something different. They told stories that mattered to the people making them. They were not optimized for data or designed by committee. And audiences responded.

Of course, not every film made with conviction succeeds. Many do not. But conviction is the floor, not the ceiling.

Sinners (2025)

Ryan Coogler wrote and directed an original horror film set in 1932 Mississippi. Not a sequel. Not a remake. Not based on existing IP. Michael B. Jordan played twin brothers confronting a supernatural evil in the Jim Crow South.

Budget: $90 million. Box office: $370 million worldwide. It became the first original film to cross $200 million domestically since Coco in 2017. It received 16 Academy Award nominations, the most in Oscar history, and won four including Best Original Screenplay and Best Actor.

Fangoria called it "Hollywood's biggest original box office hit in years." The data did not predict this. A filmmaker with a vision did.

And Sinners was not alone in 2025. F1 made over $600 million worldwide. Paul Thomas Anderson's One Battle After Another crossed $200 million. Weapons performed strongly. Originals were not absent from the year. They were just the exception, not the rule.

Parasite (2019)

A Korean-language film about class inequality. The conventional wisdom was clear: American audiences do not watch subtitled movies. Bong Joon-ho made it anyway.

At the Golden Globes, he said: "Once you overcome the one-inch-tall barrier of subtitles, you will be introduced to so many more amazing films."

Budget: $11.4 million. Box office: $263 million worldwide. First non-English film to win Best Picture at the Oscars.

Everything Everywhere All At Once (2022)

A multiverse story set in a tax office, starring Michelle Yeoh as a middle-aged Chinese-American laundromat owner. No franchise. No existing IP.

Budget: $25 million. Box office: $143 million worldwide. Seven Oscars including Best Picture. First A24 film to cross $100 million. Its budget was less than the craft services on a Marvel production.

Anatomy of a Fall (2023)

A courtroom drama. In French and English. No stars American audiences would recognize. Budget: $6.6 million.

Palme d'Or at Cannes. Five Oscar nominations including Best Picture. Won Best Original Screenplay. Sold 1.9 million tickets in France.

The Zone of Interest (2023)

Jonathan Glazer filmed a Holocaust story with no crew on set, hidden cameras, no artificial lighting, minimal close-ups. He removed everything that makes filmmaking feel like filmmaking. Budget: $15 million.

Steven Spielberg called it the best film about the Holocaust since Schindler's List. Won the Oscar for Best International Feature Film.

The Worst Person in the World (2021)

Norwegian language. Director Joachim Trier cast Renate Reinsve in the lead role. At the time, she was about to quit acting and become a carpenter. Trier wrote the film specifically for her. Budget: around $5.7 million.

Reinsve won Best Actress at Cannes. The film was nominated for Best Original Screenplay at the Oscars. Their next collaboration, Sentimental Value, won the Grand Prix at Cannes 2025.

The Pattern

The films that failed had budgets between $50 million and $400 million. They had the biggest stars, the biggest studios, and the most recognizable brands.

The films that worked had budgets between $5.7 million and $90 million. They were in Korean, French, German, Norwegian, and English. Three of the five were not in English.

What separated them was not budget, not language, not whether they were original or based on something else. What separated them was that someone had a story they believed in enough to fight for. Bong Joon-ho told audiences to overcome the one-inch barrier of subtitles. Glazer filmed with hidden cameras and no visible crew on set. Trier called an actress who was about to give up. Coogler set a horror film in 1932 Mississippi.

These were not safe decisions. These were convictions.

AI is the Perfect Scapegoat

"AI is killing creativity." "AI will replace filmmakers." "AI is the end of cinema."

But cinema was already in trouble. The box office was already declining. The franchises were already failing. The originality was already gone. That happened because of spreadsheets, focus groups, and retention metrics. Not because of AI.

The real concern with AI is not that it kills creativity. It is that studios will use it to cut the human labor that creativity depends on. That is a legitimate fight. But it is a fight about working conditions, not about storytelling.

AI is a tool. It can help you plan shots, generate storyboards, visualize scenes before you shoot them. But it cannot decide that a story about a Korean family infiltrating a rich household is worth telling. It cannot decide that a tax office multiverse makes sense. It cannot call an actress who is about to quit and say "I wrote this for you."

Those are human decisions. And they are the only decisions that matter.

The Only Question

The question is always the same: do you have a story worth telling?

If yes, tell it. The budget does not matter. The language does not matter. The tools do not matter. Use AI, use CGI, use a pencil on a napkin. Whatever gets the story from your head to a screen.

If no, nothing will save it. Not $400 million. Not Tom Cruise. Not Robert De Niro. Not the most advanced AI on the planet. The audience can always tell when a film was assembled instead of created.

Bong Joon-ho knew his story was worth telling. Justine Triet knew. Jonathan Glazer knew. Ryan Coogler knew. Joachim Trier called an actress one day away from quitting because he knew.

That is what makes a great film. Not the tools. Not the budget. The courage to tell a story you believe in.

If you are making a film right now, do not ask what AI can do for your story. Ask what your story can do that nothing else can. If you cannot answer that in one sentence, no tool will save you.

Sources

• Screen Daily: North American Box Office Hits $8.87bn for 2025
• CNBC: Hollywood has a box-office problem
• World of Reel: Snow White Was 2025's Biggest Box Office Bomb
• The Direct: The 9 Major Box Office Flops of 2025
• CBR: 10 Biggest Box Office Flops of 2025
• Fangoria: Sinners Is Hollywood's Biggest Original Box Office Hit In Years
• Variety: Scorsese Says Cinema Is Being Devalued to "Content"
• Boston.com: How The Irishman Almost Never Happened
• Deadline: Bong Joon-ho on the "one-inch barrier of subtitles"
• Variety: Everything Everywhere All At Once Crosses $100 Million
• Screen Daily: Jonathan Glazer on removing the artifice of filmmaking
• SlashFilm: Spielberg on The Zone of Interest
• Nordisk Film: Joachim Trier and Renate Reinsve on The Worst Person in the World