Inside R/GA’s 4-Day Sprint to Build 1,000 Brands With Help From AI

We no longer simply make things. We build systems that make things

R/GA has never lacked early adopters; tech-geekery is practically a badge of honor here. But even in a place like this, most people rarely get to experiment outside their discipline. Strategists don’t usually vibe-code and 3D animators aren’t writing strategic briefs.

That matters, because when access to other disciplines opens up, the walls between teams start to come down. You don’t become an expert overnight, but you do build empathy. You see how a decision in strategy ripples through design, production and engineering; creativity flows more freely.

So we ran a four-day, company-wide workshop with one goal: Everyone would build a brand from scratch using Gemini and Freepik’s multi-model creative suite—tools already on every R/GA laptop. The exercise wasn’t about crowning a prompting champion or generating the glossiest image. It was a stress test: What happens when AI stops being an occasional assistant and becomes embedded in how the entire organization works?

The goal: Build 1,000 brands—with names, iconography and unique personalities—in less than 100 hours.

We opened day one with a warning: this would be an emotional rollercoaster with moments of amazement followed by head-to-desk frustration. Working with generative AI demands the same mix of optimism and cynicism. Each day focused on a different stage of modern brand building.

Play

Day 1: Strategy With AI

Teams interrogated markets, pressure-tested positioning and shaped propositions, deliberately compressing work that might usually stretch across days. The amount of ground covered was remarkable. But the most important moments weren’t when the machine produced a polished answer. Friction became the point. Everyone was encouraged to challenge outputs, refine prompts, rework gem instructions and interrogate assumptions. You learn that you can’t take your hands off the steering wheel. Models will provide an answer, but it’s your responsibility to deliver sharper and strategically sound results.

Day 2: Brand Design With AI

Everyone took their strategies from the previous session and translated them into visual worlds. Client leads, strategists and producers found themselves working alongside seasoned designers. And although taste and technical depth still differentiated the groups, it was striking how far design-newbies could go. Designing in partnership with a machine develops your logic. Rather than crafting static logos, teams began thinking about generative frameworks–a nod to the designer’s evolving role from maker to orchestrator.

Day 3: Gen Video Brings Brands to Life

Scripts, storyboards and motion explorations that might once have lingered in decks for weeks came to life in hours. The ceiling lifted dramatically, but so did the temptation to overproduce. The strongest amongst us resisted that impulse, editing with discipline.

Day 4: Functional Prototypes Through Vibe Coding

The speed was undeniable. And while the surge in output was impressive, it was not uniformly excellent. Tools need taste. In a way, speed proved to be the least interesting takeaway. Over the past year, much of the industry’s attention has fixated on acceleration: faster decks, edits, copy and code. Speed is undeniably useful, particularly when margins remain tight. But an exclusive focus on velocity risks reducing innovation to the story of the faster horse.

The Takeaway: AI Raises the Stakes on Judgement

AI can produce endless outputs with remarkable efficiency. But it can’t decide which direction is worth pursuing. That still requires judgment, taste and beautifully imperfect human thinking.

The strongest outcomes didn’t come from those who generated the most material, but from those who edited most rigorously and asserted a clear point of view. Embedding AI across strategy, design, film and code in a compressed sprint revealed that we are no longer simply making things. We are building systems that make things. That is a very exciting future.

The real lesson of the workshop, then, is that AI raises the stakes on judgment. Agencies that succeed won’t be those that talk most elegantly about change, but those willing to rewire themselves while keeping imperfect human thinking in the loop.

Because if we don’t direct machines with taste, instinct and a little glorious irrationality, we won’t get a faster horse. We’ll get a stable full of lookalikes.