AI and the end of linear creativity

New AI-enabled workflows are exposing our frameworks for what they always were: a coping mechanism. Creativity doesn’t happen in structured, sequential phases. It happens in messy, simultaneous, rapid loops.

For the first time in my career, new tools are fundamentally changing how we work. Not what we prioritise or the vocabulary we use, but how we get there. The shift looks small from the outside. It isn’t.

Linear processes are fiction

Most established creative frameworks, like the double diamond, tell a compelling story of separate divergent and convergent stages, where you efficiently move a project from problem to solution through distinct phases. It makes creativity feel manageable. But if you’ve been close to an actual creative process, you know it’s fiction.

When was the last time a breakthrough idea waited politely for phase two? Or the initial problem definition wasn’t challenged later on? These models imply you can finish diverging before converging. But in reality, creativity is about constantly doing both. The best insights and ideas come from the tension between diverging and converging, from rapid switching, not staged progression.

Breaking the bottleneck

The problem with simultaneous exploration and validation is that it’s very hard to do. It’s like drawing a map while you’re navigating the terrain. Separating it into structured phases is a workaround for human cognitive limits, not an accurate model of how ideation actually happens. But AI changes that. The cognitive load that once required sequential phases and large teams now fits into a tight, multidisciplinary team’s day-to-day work.

A creative team developing a campaign concept can explore ten visual and narrative directions in parallel, test each against brand strategy and audience response signals, kill weak directions immediately and evolve promising ones with rapid variations. The team that once had to commit to two or three directions before testing now validates dozens before narrowing. All in one fluid workflow with humans orchestrating and AI executing.

Multipliers vs replacements

Instead of using AI for efficiency or to replace team members, AI as a thinking partner makes small teams vastly more capable. The team brings curiosity about which directions matter, taste about what is compelling and judgment about when exploration hits diminishing returns. The AI brings tireless exploration, pattern recognition across vast possibility spaces and the ability to pressure-test every assumption simultaneously.

Neither works without the other. A team without AI is limited by cognitive bandwidth, but AI without humans lacks direction and meaning.

Humans steer and decide what matters and why. Agents act and execute within these guardrails. This has consequences for the shape and size of teams, for client relationships and for the outcome and format of deliverables. Three things that change immediately:

Smaller teams, zero handoffs

The traditional model organised teams by phase. Researchers hand off to strategists, who hand off to creatives, who hand off to production. Each transition loses intent and momentum. AI-powered workflows collapse these handoffs and preserve ideas from being diluted along the way.

Multidisciplinary small teams now work as one unit from day one. A strategist tests hypotheses while designers explore visual territories and writers develop narrative threads, all feeding into each other in real-time. AI agents act as connective tissue, translating insights across disciplines and maintaining coherence as multiple threads advance in parallel.

Clients as partners, not approvers

Processes like the double diamond positioned clients as approvers at phase transitions: artificial checkpoints that felt like tests rather than partnerships. With AI, a different type of client relationship becomes possible.

Clients are increasingly asking to see thinking develop rather than only reviewing finished work. AI-enabled workflows enable true partnership. Clients see thinking evolve and contribute domain expertise when it’s most valuable. Instead of arriving at a final presentation to approve or reject, they shape the work while it is still moving. The brief gets smarter. The work gets braver.

This requires organisational courage and comfort with ambiguity. And the clients who are open to it, unlock fundamentally different work.

Continuous proof beats big-bang revelation

Traditional deliverables reflected the structure of the framework: comprehensive strategies after phase one, complete territories after phase two, finished campaigns at the end. Big reveals that are either right or wrong, with no way to know until you’ve spent the budget.

AI-enabled delivery de-risks through continuous validation. You develop strategic hypotheses and test them across rapid cycles. You explore dozens of directions in parallel, killing weak ones quickly and evolving promising ones. You still have milestones and presentations. But what you’re presenting are validated directions backed by evidence, not untested recommendations backed by expertise. Big-bang revelations become continuous proof, making it easier to predict hits and avoid misses.

Embrace the mess

Instead of a double diamond, the creative process is evolving into something more like a double helix, where agentic systems and human judgment intertwine.

AI is creating conditions where insight can emerge from rapid cycles of expansion and validation. Teams can simultaneously explore broadly and focus sharply while clients become active participants in discovery. This is the work we should be doing, because it’s how creativity actually works.

The future of creative work doesn’t have to be perfect process. It never was. The agencies that embrace that reality now will define what comes next.

All opinions expressed throughout this article are the author’s own and do not necessarily represent those of AKQA or its affiliates.