You ever have one of those ideas? A flash of brilliance for a short film, a music video, or even just a killer scene that you can see perfectly in your head. The sweeping camera shot, the specific mood, the exact look on an actor’s face. Then reality hits. You don’t have a million-dollar budget, a Hollywood crew, or two years to bring it to life.
I’ve been there. So many of my best ideas have just sat in a notebook, gathering dust because the gap between vision and reality was just too massive. Well, it feels like that gap just got vaporized overnight.
There’s a short film making the rounds that looks like a high-budget thriller pulled straight from tomorrow’s headlines. It opens with a US stealth bomber streaking across a dark sky toward Iran. Then it cuts to Tehran, where a woman is quietly feeding stray cats in the rubble left by airstrikes. The footage is polished, cinematic, and incredibly tense.
Except none of it is real. The location doesn’t exist. The bomber is a digital ghost. The woman is a figment of code. The entire thing was generated by AI.
This is the insane new world of filmmaking we’ve just stepped into, and it’s a total game-changer.
🎬 Cinematic News at the Speed of Culture
The directors behind this, Samir Mallal and Bouha Kazmi, aren’t just making cool clips. They’re pioneering a whole new genre they call “cinematic news.” Their 12-minute film, Midnight Drop, was born from a single line in a news article about the real US-Iran tensions: a detail about a woman who walked the streets feeding cats.
Think about that. They took a fragment of a news story and turned it into a cinematic sequence that feels like it was shot by a top-tier director. Before this, they made another film called Spiders in the Sky, a recreation of a Ukrainian drone attack on Russian bombers. What would have normally taken at least two years and millions of dollars to produce, Mallal pulled off in just two weeks on his own.
This is nuts. We’re talking about creating high-level, emotionally resonant films about current events, delivered almost in real-time. Hollywood moves at a glacial pace. Culture moves at the speed of a tweet. AI is finally letting creators keep up.
⚙️ The Tech Behind the Magic
So, what’s powering this revolution? It’s not one single magic button, but a cocktail of incredibly powerful new AI tools. The big one here is Google’s Veo3, an AI video generation model that is absolutely rocking the creative world. It’s the engine inside Google’s new filmmaking tool, Flow, which can also generate speech and sound effects.
But they’re not just using one tool. The filmmakers are mixing and matching the best of the best:
- Google’s Veo3: For generating high-quality, controllable video sequences.
- OpenAI’s Sora: Another heavyweight in the AI video space, known for its stunning realism.
- Midjourney: Still a king for creating initial concept art and still images that can inform the video generation.
- ChatGPT: They even used this to help edit and structure the narrative, taking a lengthy interview with a real drone operator and sharpening it into a compelling script.
It’s a hybrid workflow where the human director is the conductor, and these AI tools are the orchestra. You still need the vision, the taste, and the storytelling skills. The AI just gives you the power to execute that vision at a speed and scale that was previously unimaginable.
✍️ Welcome to ‘Prompt Craft’: The New Directing Skill
Mallal calls his process “prompt craft,” and this is the key takeaway for any creator right now. Your ability to translate your directorial vision into text prompts is becoming the new essential skill. It’s not about coding; it’s about communication.
I love this quote from him: “The creative process is all about making bad stuff to get to the good stuff. We have the best bad ideas faster.”
That’s it! AI accelerates the iteration process. You can try a dozen camera angles, lighting setups, or character designs in the time it used to take to set up a single shot.
Here’s how you can start thinking like a prompt-craft director:
- 💡 Be Specific & Cinematic: Don’t just say “a sad woman.” Describe the shot. “Low angle shot, a woman in her 50s, face etched with worry, stands in the ruins of a building at dawn. Soft, hazy light filters through smoke and dust. The mood is somber but resilient. 35mm film look.”
- 🚀 Layer Your Instructions: Start with a simple idea and build on it. Get the basic scene down, then refine it with follow-up prompts. “Make the lighting more dramatic, like the golden hour.”
“Add more debris in the foreground.”
“Change her expression to one of quiet defiance.” - ✅ Mix AI with Human Touch: The best results right now come from a hybrid approach. Generate the visuals with AI, but bring in a real human voiceover. Use AI to help brainstorm script ideas, but do the final writing yourself. Edit the AI-generated clips together with your own creative timing and pacing.
- ✨ Embrace Rapid Prototyping: Got a wild idea for a scene? Don’t just write it down: make it. Generate a rough cut in minutes. See if it works. If it doesn’t, who cares? You spent a few minutes, not a few thousand dollars. This freedom to experiment and fail quickly is a creative superpower.
🌐 The Bigger Picture: Ads, Hollywood, and… Slop
This isn’t just for indie filmmakers. The entire entertainment industry is on the verge of a seismic shift. TV producer Richard Osman said that an era of entertainment history has ended and a new one has begun. He predicts that by 2027, the majority of TikToks, ads, and movie trailers will be AI-assisted.
The advertising world is already bracing for impact. The CEO of Brandtech Group, David Jones, says we’re going from less than 1% of brand content being made with generative AI to nearly 100% being fully or partly created with it.
Of course, with any powerful new tool, there’s a lot of junk. We’re seeing a flood of what people are calling “slop”: low-effort, nonsensical AI content. But amid the weird diving dogs and bizarre content farms, masterpieces like Midnight Drop are showing us the true potential.
⚖️ The Elephant in the Room: Paying the Creators
We can’t talk about this without addressing the massive issue of copyright. These AI models learn by analyzing unfathomable amounts of data, including art, photos, and videos from across the internet. This has the creative industries justifiably furious.
How do we make sure that the original artists, photographers, and filmmakers whose work forms the foundation of these tools are compensated? Cross-bench peer Beeban Kidron puts it perfectly:
“Creators need equity in the new system or we lose something precious.”
Even Mallal, a champion of the tech, agrees. He wants to see a system where artists get paid for their contributions. This is a problem we have to solve to ensure a healthy and sustainable creative ecosystem for everyone.
This is just the beginning. We’re standing at the foot of a mountain of new creative possibilities. The tools are here, they’re getting better every single day, and they’re giving a voice to anyone with an idea. The human director, the storyteller with a vision, is more important than ever. The AI is just the ultimate tool to bring that vision to life.
The question is no longer if you’ll be able to make that film you’ve been dreaming of. The question is: what will you create?
- The U.S. Copyright Office has clarified that for a work to be copyrightable, it must be the product of “human authorship.” This means content created with significant creative input from a human using AI as a tool may be protected, but work generated almost entirely by an AI is not eligible for copyright.
- In response to the use of AI for creating digital replicas, actors and industry unions like SAG-AFTRA are pushing for stronger contractual protections. These negotiations focus on consent, control, and compensation for the use of an actor’s likeness, ensuring they are not used to train AI models or generate performances without permission.
- Beyond visual effects, AI is heavily influencing pre-production and marketing. Studios use AI algorithms to analyze screenplays, forecasting their potential box office success based on data from past films. Similarly, AI can generate and test multiple versions of movie trailers, optimizing them for different audience segments.
- A key concern among creatives is the risk of stylistic homogenization. While AI can efficiently generate content, critics worry it may rely on existing data and trends, potentially leading to films that feel formulaic and lack the unique, personal vision that a human director or artist provides.