Most editors recognise this pattern, even if it is rarely called out directly: the footage arrives, the first assembly begins, and somewhere along the early cuts, it becomes clear that something is not working in the way it was expected to.
The pacing feels slightly off, the tone doesn’t fully land, and a scene that looked convincing on paper struggles to hold together on screen.
Nothing is obviously broken, but the piece as a whole lacks cohesion.
At that point, the work shifts into adjustment. Tightening, restructuring, removing, and reworking, all in an effort to shape something coherent out of material that was never fully aligned to begin with. For a long time, this has been accepted as part of the editing craft. In reality, it is not a creative problem, it is a process problem. To be more specific, it is a timing problem.
Editing has been happening too late
In most production workflows, alignment does not happen in a single, unified moment. Instead it happens in stages, often across different teams and at different times, with each group responding to a slightly different version of the same idea.
Scripts are approved before visuals are properly understood, tone is discussed, but not experienced, and decisions are made sequentially, without a shared reference point that allows everyone to respond to the same thing in the same way.
By the time footage reaches post-production, many of those decisions are already fixed, which limits how much can realistically be changed. This is why editors often find themselves (frustratingly) solving problems that did not originate in the edit.
Prompting changes where the work happens
AI does not solve this by producing perfect outputs, and that is not where its real value lies. What it does is make thinking visible much earlier, giving teams something concrete to respond to before anything is locked in.
A prompt produces an output that, while not final, is tangible enough to interrogate. It allows teams to see how an idea plays out, understand how it feels, and identify where it holds together and where it starts to break down.
This immediately exposes gaps, removes ambiguity, and surfaces weak decisions before they become embedded in the production process. Crucially, all of this happens before production begins.
This is part of a broader shift across AI-driven workflows, where correction is no longer something reserved for post-production, but something that is pulled forward into the earliest stages of development.
Editors are no longer waiting at the end
When editors are brought into this earlier stage, their role changes in a meaningful and practical way. They are no longer responding to material that has already been created, but are instead shaping it while it is still fluid and open to change.
The core instincts remain exactly the same. Understanding structure, rhythm, pacing, and narrative clarity. knowing what to keep, what to remove, what is missing, and what does not belong. What changes is when those instincts are applied.
Instead of working around decisions, editors influence them. Instead of refining intent after the fact, they help define it from the outset. And because this happens earlier in the process, the result is not only faster, but significantly cleaner, with fewer compromises and far less need for rework later on.
Clarity, not speed, defines the story
AI undeniably accelerates the process. Iterations that once took days can now be explored within hours, and multiple creative directions can be tested without committing to a shoot.
However, speed is not the most important shift taking place. Clarity is.
When ideas are made visible early and tested before execution, fewer assumptions carry through to production, and far fewer surprises emerge in post. The edit becomes what it was always intended to be, a stage for refinement and strengthening, rather than a point of rescue.
The role doesn’t disappear. It sharpens
There is a growing perception that prompting may reduce the role of the editor. In practice, the opposite is happening.
As AI introduces more options and variation, the need for judgment becomes more critical, not less. The ability to determine what works, what resonates, and what ultimately feels right remains a human responsibility. AI can generate endlessly, but it can’t decide. It doesn’t t understand pacing in a human sense, and it can’t recognise when something lands with the intended impact.
That responsibility still sits with the editor. The difference is that it now comes in earlier in the process, where it has a completely different influence on what we see at the end of the process.
What’s actually changing
Essentially, editing is simply being repositioned. For years, it has operated at the end of the production process, shaping what has already been created and working within the constraints of earlier decisions. Now, it is moving closer to the beginning, where it can shape what gets created in the first place.
Prompting hasn’t removed the editor, it has made their role visible sooner. And, that’s where the beauty lies: it’s made them more central to the process than they have ever been.
Written by Garon Campbell, Founder and Director, Breadbin Productions