From Authors to "Algorithm Curators": How AI is Reshaping the Designer’s Role


A few weeks ago, I tried using Midjourney for an interior design project I’m currently working on. I was looking for ideas for a system of wooden partition walls. Within seconds, the software spat out a series of images that were almost intimidating in their elegance: perfectly calibrated lighting, flawless wood grain, and an atmosphere straight out of a high-end magazine. Yet, after staring at those renderings for more than ten minutes, I started to feel a strange sense of rejection. There was something deeply wrong with them—or rather, something far too right.

The problem with these tools is that they operate on a statistical basis. This means the algorithm doesn't invent a form because it senses a need for it; it simply calculates which visual configuration has the highest probability of looking pleasant and familiar based on millions of earlier images. This generates a peculiar kind of visual boredom. When everything is optimized to please the average user, style converges toward an indistinct center. In commercial graphic design or web design, this standardization is already a visible reality: layouts all look alike, colors follow the exact same algorithmic seasonal trends, and corporate illustrations seem to come from the very same hand.

Philippe Starck’s A.I. chair for Kartell is often cited as the textbook case of this partnership: the designer asks how to save material, and the machine handles the structural calculations. But looking closely at that project, you start to wonder if it’s more of a brilliant marketing campaign than a genuine methodological revolution. The chair features biomorphic shapes that closely resemble bone structures—an aesthetic that generative design software (like Autodesk's) now replicates everywhere, whether you're engineering a motorcycle frame or designing a piece of furniture. The chair became an icon not so much for the intrinsic quality of its structural solution, but because it carried Starck’s signature and Kartell’s brand. In that case, the algorithm was used as an excellent generator of avant-garde stylistic tropes, rather than a problem solver.

This makes me reflect on something we often overlook: the time it takes to doubt. When I sketch by hand or model a piece of plaster, there is physical friction. I mess up a line, the eraser wipes it away but leaves a trace, the plaster breaks if I scrape it too thin. In that forced delay, in that mistake of the hand, critical thinking slips in. You pause and change direction. Artificial intelligence erases this latency period. It replaces the process with an instant result. The danger is that the image stops being a tool to describe a project and becomes the project itself—a flawless shell designed for the screen but disconnected from physical reality.

Deep down, the machine operates in total isolation from the world. I don't just mean the obvious fact that it lacks a body to feel the roughness of a material or test the ergonomics of a handle. It lacks historical and emotional memory. To an algorithm, Enzo Mari's design archive and a generic stock photo are mathematically equivalent: they are just strings of data. The machine cannot grasp the political value of a choice, nor can it decide to be deliberately inefficient or "wrong" to convey a disruptive message or a social critique.

I don't know what the outcome of this transition will be. I don't think it makes sense to reject these tools on principle; they are too powerful to be ignored, and they streamline the workflow extraordinarily well in many executive areas. However, I feel that the real difficulty won't be learning how to dialogue with the algorithm, but rather defending the outlier, the irony, and even the flaw. If a designer's value shifts entirely to the capacity to curate the machine's output, we risk turning from creators into mere editors of a continuous flow. And I'm not sure that's enough to justify our role.

#GenerativeAI #DesignTheory #Authorship #AestheticStandardization #HumanAICollaboration