Our founder and ECD James Hilditch joined a panel at the APA’s Future of Advertising event for a session titled “AI - We’re On It,” alongside Tim from Knucklehead/Airhead, Ramy from Common People Studio/Future Creatives, and Morgan from The Sweet Shop/The Gardening Club. The session was chaired by Luiza Cruz-Flade, Executive Producer at Territory Studio.
The conversation covered what’s actually changing in the briefs landing on production company desks, the practical realities of AI in production, and the uncomfortable question of how studios build sustainable business models when budgets are shrinking.
The briefs are genuinely different now....
Six months ago, many of the briefs coming through our doors simply wouldn’t have existed. The panel identified three distinct categories emerging across the industry.
The first is scale. Brands are asking for hundreds of versions of a single video, localised and adapted across territories and languages. The sheer volume of deliverables required would have been unthinkable without AI in the pipeline.
The second is creative unlock. A client has an ambitious idea but the budget wouldn’t stretch to a traditional live-action approach. AI means that idea can come to life in a way that’s achievable without compromising the creative vision. Our Escape Convention project is a good example: achieving that kind of animal performance on set would have required huge expense and significant unpredictability. AI made it possible. As one panellist put it, it’s about solving problems that were “literally intractable” just two years ago.
The third, and perhaps the most revealing, is the rescue brief. Agencies or in-house teams have tried to produce AI content themselves, hit a wall, and need experts to come in and finish the job. This is happening more and more, and it says a lot about where real expertise sits.
Not every brief that arrives labelled “AI” should be an AI project though. Several panellists made the same point: nobody ever said “this is a Flame project” or “this is an After Effects project.” You describe what you want to make and let the experts decide how to approach it. AI should be treated the same way. It’s a tool in the kit, not a category of work.
Every panellist agreed on this without hesitation. There is no AI project that doesn’t involve traditional craft. Every production is a blend: live action, VFX, motion design, compositing, colour grading, editorial. AI is woven in where it adds value, but the idea that you press a button and get a finished piece of content is still a long way from reality.
Ramy from Common People put it well. You have to approach every project with direction, emotion, and deep creative oversight. The human expertise is what brings the extra quality that makes the work actually land. Without that, it falls apart.
The tools are evolving at an extraordinary pace. New models and platforms launch weekly. You can start a production using one tool and find something better drops halfway through, or discover that your chosen platform can’t do what you need it to. Producers are constantly navigating that landscape, making judgment calls about when to switch, when to blend, and when to fall back on traditional methods.
This is not a frictionless process. It requires experience, taste, and a deep understanding of what each tool can and can’t do.
One theme that ran through the entire discussion, even if it wasn’t named explicitly, is what we at BearJam call the 85% problem. AI gets you tantalisingly close to a finished output. But that last stretch, the final quality that makes something feel right, look polished, and hold up to client scrutiny, still requires human craft.
Image generation can replace huge parts of the traditional CG pipeline. But it takes hundreds of iterations. It takes days, sometimes weeks, per shot. And even then, there’s almost always something that needs fixing: a hand, a lighting inconsistency, a movement that doesn’t feel natural. That’s where compositors, colourists, VFX artists, and editors step in.
The craft hasn’t disappeared. It’s relocated. The skillset required has shifted, but the need for experienced people making considered creative decisions is as strong as ever. Maybe stronger, because the margin for error is different when you’re working with generated material.
The session closed with a point that deserves more airtime. At a previous APA event, a concept called Craft Intelligence was introduced, and it resonated deeply with the panellists.
Craft Intelligence is the proprietary know-how a studio uses to combine multiple tools, techniques, and specialists into a single, coherent production system. It includes workflow design, tool orchestration, QA gates, handoff standards, and the creative judgment required to choose the right tool at the right moment.
This matters because the uncomfortable business reality is real. Projects that once commanded £six figures are now being quoted at less than half. If studios apply the same markup structures to smaller budgets, the maths doesn’t work. The industry risks devaluing itself.
Morgan from The Sweet Shop made a sharp point about where the opportunity sits. A huge slice of marketing budgets has been consumed by versioning and adaptation: reformats, translations, territory edits. AI can genuinely streamline that part of the process. If studios use AI to reduce the cost of versioning, it frees up more of the budget for the craft and production that actually drives brand impact. The goal isn’t cheaper work across the board. It’s spending less on the mechanical stuff so there’s more resource for the creative stuff.
The answer isn’t to accept less. It’s to be clear about what clients are actually paying for. The pipeline design, the creative orchestration, the ability to adapt when tools change mid-project: that’s where the value sits. It should be recognised, respected, and charged for accordingly.
As James noted during the panel, this is one area of the budget that shouldn’t be stripped out. Getting the production pipeline right at the beginning is what determines whether the work holds together or falls apart. That takes creativity, craft, and experience. It takes Craft Intelligence.
The tools are new. The creative possibilities are bigger. The workflows look different from even a year ago. But the fundamentals haven’t moved. Great ideas, strong production discipline, human judgment, and craft. That’s what makes the work good. AI is the most exciting addition to the production toolkit in a generation. But it’s an addition to the toolkit. Not a replacement for the people holding it.