Toldy Construct
Toldy Construct — Journal

The boundaries of creativity: how generative AI is changing design

Dr. Toldy Gábor - Toldy Construct · · English translation · Updated:
Illustration: The boundaries of creativity: how generative AI is changing design

Editorial illustration

Architecture and design have always been the meeting point of creativity, human intuition and engineering rationality. However, in recent years, a new player has entered the design stage: generative artificial intelligence. The question, which is being formulated more and more sharply in the profession, sounds like this: is AI an inspiring partner or a controlling actor that is slowly taking over and undermining the designer's autonomy?

A new source of inspiration

Generative algorithms are undoubtedly revolutionary tools. They are able to generate hundreds of alternative design options in a few minutes that a human designer might never have thought through. This enormous possibility of variation is particularly important in areas where sustainability and energy efficiency play a central role.

A good example of this is the British office Foster + Partners, which is known worldwide for its iconic buildings. The company has incorporated generative design tools in several of its projects in order to find energy-efficient and environmentally optimized solutions. Rather than taking over the creative process, AI acted as a catalyst: designers were able to more quickly evaluate and compare different variations, leaving the decision ultimately in human hands. It is a synergy where technology expands the horizons of human thought.

The designer as operator - the danger of autonomy

But not all stories are so positive. One of the world's largest design companies, the American Gensler, has used generative AI in many office design projects to increase efficiency. The software was able to optimize work processes, lighting conditions, desk placement and movement routes. The end result is truly effective: the spaces work functionally, the offices are optimally utilized.

But is that enough? The criticism begins right here. The spaces created in this way often became the same - standardized, mechanical solutions that lost the specifics of the spirit of the place. The human touch, the cultural and emotional layers that make real architecture special have been pushed into the background. Some designers fear that if AI dictates the "most optimal" solution, the specialist will become a mere "operator" who approves the decisions generated by the machine.

Public AI: whose legacy is it learning from?

A more serious, less visible question also arises here: what does this AI learn from? Most generative artificial intelligence - such as public design models - is based on huge databases created by the work of human designers and architectural firms. But who contributed to this? Who gave permission for several decades of professional experience, design style and way of thinking to become a learning material for a machine?

A publicly accessible AI system is not necessarily open source. Designers should separately assess its terms of use, data handling and the verifiability of its outputs. Neither access model automatically resolves questions of originality or professional responsibility.

This process can be particularly dangerous in countries where design fees, copyright protection and the powers of professional chambers are weak. The end result: creative work is devalued, and the prestige of the profession is devalued.

A paradigm shift in design

So the debate is not about whether there will be AI in design - because there definitely will be. The real question is: who remains the decision-maker, and who owns the knowledge from which AI learns?

If designers retain their autonomy and treat algorithms as consciously integrated tools, AI can become an extension of human creativity. If, on the other hand, the profession passively gives in to technology, we can easily lose diversity, authorial value and, ultimately, professional identity.

In order for AI to truly serve the profession, transparent sources, legal protection and professional control are needed. The algorithm cannot be an independent creator - only a tool that serves human creativity.

Sources

Foster + Partners: official project presentations and press releases

Gensler Research Institute - AI in Workplace Design reports

ArchDaily, Dezeen - interviews with architects on the use of generative AI

OECD & RIBA: Ethical AI in Architecture and Design (2023)

EU Artificial Intelligence Act - Draft Regulation, 2024.

Related articles

The next chapter

Let’s discuss your next exceptional building.

Start a conversation