Every creative project has a beginning, an end, and a middle. The beginning is full of possibility—sketches, fragments, early sparks. The end is where things harden into form—decisions made, tensions resolved, the work stepping into its final voice. But the middle has always been the part that determines whether something becomes merely competent or genuinely alive. It is the place where ideas deepen, where contradictions collide, where a project stops being an exploration and becomes a commitment. Writers like Elena Ferrante describe the middle as “the point at which the work argues back.” Filmmakers such as Denis Villeneuve call it “the negotiation phase,” when the film reveals what it wants but also what it refuses. Even choreographers like Pina Bausch spoke of the middle as “the struggle that gives the piece its spine.”

Generative AI has changed almost everything about the mechanics of creation, but it has not eliminated this middle. If anything, it has made the middle more fragile and more necessary than ever. Tools like Midjourney, Runway, Claude and Sora compress the early creative stage—what once required days of drafting, sketching or scouting can now unfold in minutes. The systems offer abundance before intention has fully formed. They provide clarity before direction has matured. And while this acceleration is extraordinary, it risks distorting the rhythm of creative development: it can make the beginning feel like the middle, long before the creator has done the thinking the middle requires.

This is where Copy Lab’s philosophy guides our approach. The sacred partnership between humans and GenAI depends on the human’s ability to hold space for the middle—to protect the period of interpretation, deliberation and emotional calibration that no model can replace. AI can generate a visual direction, but it cannot determine whether that direction carries the resonance a project needs. It can propose language, but it cannot sense whether the tone aligns with the underlying intuition. It can suggest structures, but it cannot feel the discomfort that signals a deeper idea is still waiting to emerge.

The middle of a project has historically been the least glamorous phase. It lacks the thrill of early ideation and the satisfaction of completion. It is characterized by self-doubt, re-evaluation, and the slow building of conviction. Writers like Zadie Smith describe it as “the place where you lose faith before you gain it again.” Musicians like Thom Yorke speak of the middle as the moment when a song “refuses to reveal its logic.” The process demands clarity of purpose, not because the project is finished, but because it is resisting finality until the creator understands what it needs to become.

AI does not resist. It generates. And that difference matters.

Generative tools remove the friction that once forced creators to interrogate their choices. When a designer had to manually redraw variations, each variation carried weight. When a director had to reshoot or re-edit scenes, every decision demanded justification. Today, the cost of producing options has effectively disappeared. Variations populate instantly. Adjustments happen in real time. The technology removes the labor but also removes the forcing mechanism that once compelled creators to understand the work.

This does not make GenAI a threat. It makes it a partner that demands more from its human counterpart. The middle becomes a deliberate choice rather than an unavoidable phase. The creator must slow themselves down, even when the technology speeds up. They must ask the questions the system cannot: Why this tone? Why this structure? Why this rhythm, this color, this pacing? What tension is the work trying to resolve? What feeling is the work trying to protect? What truth is it trying to uncover?

At Copy Lab, we see the middle not as a hurdle but as the heartbeat of a project. GenAI gives us the freedom to arrive there faster, but we never skip it. The middle is where intention becomes visible—where the human decides what the work should mean, what it should resist, what it should insist upon. The models can supply alternate paths, but they cannot make the choice of direction. That responsibility remains human, and appropriately so. It is in this space that taste, intuition, ethics, cultural awareness and personal experience converge.

The risk of the GenAI era is not that machines will take over creativity; it is that humans will forget to engage with the middle. But the opportunity is far greater: when used with awareness, AI frees creators from mechanical tasks so they can invest more fully in the interpretive, emotional, conceptual labor that defines real artistry.

The middle still matters. It is where the work becomes itself. And in the era of GenAI, it is the place where the sacred partnership between machine abundance and human intention is forged—quietly, thoughtfully, and with a depth no system can automate.

/Carl-Axel Wahlström, Creative Director Copy Lab, 2025