Part two of this Q&A is available here.

Timothy Alouani-Roby: What are the possible futures for AI in design? What are you optimistic or worried about? 

Jet Geaghan: Society at large is scrambling to come to terms with AI, which – unregulated and unabashed – has arrived almost incomprehensibly fast. Dealing with the incoming changes will likely prove to be an abrasive experience for the design industry, but we will uncover unique opportunities, risks, and benefits to further progress from.  

When it comes to possible futures, I see two broad prospects for AI.

First, design-passive proliferation — a laissez-faire approach in which we simply let AI infiltrate design. In this scenario, the new tools will be thrust upon established disciplines by the market. This would see the integration of AI led by parties outside the design industry (for example, software developers) who would likely look to optimise them for regulatory requirements or profit. Because its focus would be defined by a binary sense of ‘do or don’t’, this direction would be in danger of coalescing with a compounded reliance on the status quo that would lack nuance or compassion, and probably result in outcomes that are ‘good enough’. The world does not need that. Designers would be forgoing key opportunities to ask why the status quo is what it is and what can be done to improve it.  

This prospect tends to be used as an argument against AI in the design industry. However, if designers try to fight the influence of AI, we will probably arrive at the same place – and perhaps with less credibility.

Second, design-active engagement — where designers take an active part in imagining the future. Already predisposed to think in terms of future outcomes, designers and architects are uniquely positioned to take part in the honing of AI tools in their disciplines. Seizing this opportunity with conviction will take some engagement with computer science and agility with regard to emerging trends. Even as we speak, large language models (LLMs) are breaking down the gulf between coding and human thought faster than we could have predicted.  

Overall, I am less concerned about some great redundancy for designers or architects than a widespread entrenchment of current trends or existing patterns. AI could allow designers to become agents in their development of design tools, augmenting their work rather than creating an artistic straitjacket. 

Related: AI editorial on human-AI collaboration

TAR: How can we understand AI in terms of augmented design? 

JG: AI should provide a seismic shift in the way that buildings, spaces, and places are conceived. But the first developments will likely be prosaic – a natural evolution of the human interface with technology. 

From the first orthographic drawings to CAD and CGI, the evolution of how design is produced is defined by large leaps. Although each jump has heralded key advances in productivity, whether they have always resulted in better design outcomes or working conditions is fundamentally questionable. How the tool has guided the designer’s hand throughout history can’t go unignored – there’s always been a give and take. 

When they work with AI, designers will have the ability to produce more, more quickly – an outlook that is at best uninspiring and at worst quite problematic in the context of a world that only wants ‘more’. Viewed more widely, however, the potential for generative AI in the hands of inventive designers is a chance for us to explore ideas in ways that were previously impossible – adding countless new strings to designers’ bows. 

One obvious new augmented AI ability is the use of natural language prompts to synthesise imagery, code or strategy that would otherwise take countless hours. While the time saved may be the most obvious benefit, far more intriguing is the emergence of co-creativity with non-human intelligence. Here, a human designer’s critical intelligence is buttressed by machine learning tools to process boundless sets of complex data. Together, we can systematically test countless design iterations, rapidly spit-ball ideas and aesthetics, and visualise dreams that might otherwise get lost in digital translation. 

Rather than being distracted by alarmism or moral panic, AI should be seen as a rapid amplification of the design arsenal. The potential to bounce ideas off or through a non-human agent will be an asset to designers — to help us step away from Photoshop or BIM for at least a moment, and to contemplate the meaning and influence of our work, explore the bigger picture and think outside the box.

The second part of Jet Geaghan’s Q&A on AI can be found here.

Woods Bagot
woodsbagot.com

Images
Courtesy of Jet Geaghan

We think you might also like this profile of Woods Bagot’s Domenic Alvaro.