Why Agency Scriptwriters Should Experiment with AI Filmmaking
Generative AI is not just changing how films are made. It is changing how stories are written—and for scriptwriters, that shift may be the most consequential creative development since the script itself became the industry’s dominant unit of meaning.
For simplicity, when I say “scriptwriter,” I’m talking about anyone who writes for moving images—advertising, branded content, short films, social video, or cinema. The problem is the same. Only the running time changes.
Over the past few months, I’ve found myself working on generative AI film and video projects, both as experiments, and as paid client projects, not as a technologist chasing novelty, but being asked to think simultaneously like a scriptwriter, a casting director, a costume designer, a cinematographer, a production designer, even a lighting director—roles that have traditionally been separated by process, budget, and hierarchy, now collapsing into a single act of imaginative intent. What’s striking is that in every one of these projects, regardless of how many creative decisions are being explored in parallel, the work still begins in the same place.
With the script.
Because before there is casting, before there is wardrobe, before there is light, space, texture, or movement, there is the problem the industry has largely learned to live with rather than solve: how do you get what is in your head onto the page in a way that survives contact with someone else’s imagination? How do you ensure that tone, rhythm, and emotional intent arrive intact when your work is read, interpreted, circulated, and judged—often by people who will never see the finished film, and long before the work ever has the chance to be seen at all?
For decades, screenwriting has lived with this gap—between imagination and interpretation—as a cost of doing business. Writers learn to compress sight, sound, movement, and emotion into text, hoping that the reader will successfully reconstruct the film in their own mind. Sometimes that happens. Often it doesn’t. Ideas don’t fail because they lack originality; they fail because they lose energy, texture, and conviction somewhere between the page and the reader.
Generative AI doesn’t eliminate the gap between intention and understanding. But for the first time, it gives writers a way to shorten it.
What’s changed is not that writers can suddenly “make films” on their own. That framing misses the point entirely. What’s changed is that writers can now externalise intention earlier, faster, and with far less friction than ever before. They can move ideas out of the purely internal, imagined space and into something that can be seen, felt, interrogated, and refined—without waiting for production, permission, or a cascade of downstream decisions.
This matters because scriptwriting has always been a form of sensory compression. Writers take sight, sound, movement, pacing, and emotion and collapse them into text, trusting that the reader will successfully rehydrate the signal. When that reader is generous, experienced, and aligned, the system works. When they’re rushed, distracted, or simply imagining a different movie altogether, it doesn’t. The script hasn’t failed on craft. It has failed on transmission.
Generative AI changes that dynamic by giving writers a way to test transmission itself.
When a writer experiments with AI scene development—even at a rough, imperfect level—they are no longer writing in the abstract. They are writing with consequences. Pacing stops being theoretical. Tone stops being aspirational. Emotional beats either land or they don’t, and the feedback is immediate. You can see when a moment drags. You can feel when a silence is doing real work—or when it’s just empty space masquerading as mood.
This is not about polish. In fact, the more unfinished the output, the better it serves the writing process. What matters is not visual fidelity, but emotional fidelity. Does this scene feel like what you intended? Does the transition carry weight? Does the rhythm support the idea—or undermine it?
For writers, this represents a fundamental shift in how ideas are developed. Instead of writing the entire script and hoping the meaning survives multiple layers of interpretation, writers can now work scene by scene, moment by moment, allowing intention to be tested, adjusted, and sharpened before it calcifies into a document that has to defend itself in meetings and coverage notes.
This doesn’t turn writers into directors, cinematographers, or production designers in any traditional sense. What it does is collapse consideration, not responsibility. Writers are not taking over other roles; they are engaging more deeply with the narrative implications of those roles. They are asking better questions earlier. What does this space feel like? What does this character’s presence do to the room? What happens emotionally if the camera holds instead of moves?
These are not technical decisions. They are narrative ones. They always have been.
The industry has long felt that these considerations emerge “later,” during production, as if meaning is something added downstream rather than embedded at the point of conception. In reality, those decisions are already being made—implicitly—by the writer, whether they acknowledge them or not. Generative AI simply makes those decisions visible.
This is where the opportunity becomes especially relevant for advertising and branded content, where ideas are judged long before they are produced, often by people who will never see the final work. In these environments, scripts don’t just need to be good; they need to carry conviction. They need to signal intent clearly enough to survive translation across strategy decks, internal reviews, client stakeholders, and procurement constraints.
A script that reads “fine” is rarely enough. A script that can be felt, even provisionally, has a different kind of authority.
This is why thinking about AI filmmaking purely as a production shortcut misses the strategic point. Its real value is upstream. It sits in the messy, generative phase where ideas are still elastic, where meaning can still be shaped rather than defended. Used this way, AI doesn’t accelerate output; it deepens authorship.
There is, understandably, anxiety around this shift. Some worry that writers who engage with visualisation are overstepping, blurring boundaries that exist for good reason. But boundaries exist to manage logistics, not imagination. And imagination has never respected org charts.
What generative AI offers scriptwriters is not control, but coherence—a way to ensure that what they meant is closer to what is understood, and to reduce the number of creative concessions made not because of budget or feasibility, but because meaning was lost early.
None of this diminishes the importance of collaboration. If anything, it strengthens it. When writers arrive with clearer intent—tested, explored, and stress-tested at the level of scenes and moments—everyone else in the process can do better work. Directors direct with sharper purpose. Designers design with clearer constraints. Producers make smarter trade-offs. At its best, this is not about replacing roles. It is about respecting them enough to arrive prepared.
Ultimately, this is not a story about technology at all. It’s a story about craft catching up with reality. The tools have finally begun to acknowledge what writers have always known: that stories are not made of words alone. They are made of rhythm, silence, movement, tension, and release—things that are difficult to describe, but instantly recognisable when they are present.
Generative AI gives scriptwriters a way to work in that space earlier, more honestly, and with less guesswork. Not to finish the film. Not to bypass collaboration. But to ensure that when the script leaves their hands, it carries more of itself with it.
And in an industry where ideas so often die not from lack of originality but from loss in translation, that may be the most important shift of all.
THE END.
If your team is grappling with how generative AI changes the way scripts are written, ideas are evaluated, and films are imagined long before production, I’m happy to compare notes. I’m spending more time working at this intersection—script, story, and AI-enabled filmmaking—and helping teams figure out what actually matters.