This article originally appeared in INDESIGN #90, the ‘City Futures’ issue – get your copy here.

There have been many cities of the future. Appearing at critical junctures in history where new technologies and existing crises converge, proposals about how cities might and should develop have often taken radical, utopian and science fiction-like forms. Today, we stand at one such historical juncture as artificial intelligence continues its rapid encroachment into everyday work and life. The climate emergency threatens crises in the form of uninhabitable urban zones, mass refugee populations and disaster events, while contemporary metropolises even at the best of times are dealing with housing shortages and inequities. Amidst this maelstrom, how will the rise of AI shape our future cities?

There are echoes of what the AI revolution (if that’s the right term) might entail in the story of a city such as Tokyo. The Japanese metropolis endured a number of catastrophes in the twentieth century and, combined with problems such as population growth, found itself at the forefront of some of the most famously techno-utopian urban thinking in history. Metabolism was the name given to an architectural movement based on principles of change, growth and impermanence at a city and even national scale, inspired by biological organisms, traditional Japanese building and modern construction technologies. If the future is now with AI, the future was also then.

Much of the radicalism of Metabolism lay in the conception of the city as a dispersed, decentred network of nodes. Rather than the traditional European idea of radially organised urban spaces with a building such as a cathedral at the centre, architects such as Kenzō Tange were imagining cities of the information age running along linear axes and featuring a core infrastructural grid into which capsules or pods such as individual dwellings could be inserted, removed, disassembled and reassembled elsewhere. In our age of digital technologies, this perhaps no longer seems so novel; a map of the networked city might look something like our idea of a map of the internet. 

New technologies can act to disrupt traditional forms and it seems likely that AI will mirror this kind of networked, dispersed, decentred thinking on a number of levels. First, there is the sheer complexity: AI in its current form is perhaps most notable in its data-sifting capacities. Like a supercharged search engine, it can analyse vast quantities of data with a power that far surpasses human abilities. In infrastructural and informational terms, this opens enormous new possibilities for urban space such as in developing and maintaining complex logistics, transport and communications systems. Ideas like Tange’s 1960 Tokyo Plan, which involved creating enormous new areas of artificial ground in Tokyo Bay and connecting the metropolis infrastructurally across a much wider area, suddenly seem more realistic. 

Related: Kristen Whittle in profile from INDESIGN #90

On a more prosaic note, however, it is the accumulation of small effects on the ground that will also add up to change. At the scale of the single architecture practice, AI is likely to have a great impact on the structure of work. That is not to say that jobs are simply under threat, but that the proportional distribution of types of work will change.

For example, AI will speed up the iterative design process by producing new variations at a much faster rate (just as the move from hand-drawing to CAD did). More of the work in architecture and design — and the more highly valued work — will become the stages at either end of the process: coming up with the initial ideas, and then refining and curating them to completion. As such, resources are freed up, more work is automated and, across almost all industries, there will be more scope for people to work from home. This flexibility lends itself to cities that spread out and have a consistent infrastructural grid with multiple centres rather than the singular CBD model. 

If the age of AI will be defined by one thing, it’s probably speed. While this presents possibilities for designers, it also embodies great dangers. Good, humane design will still need to be founded on a considered, thoughtful and reflective process — using it as a simple shortcut will only produce formulaic, dull architecture and inappropriate planning. New technologies, however powerful, do not arrive in a vacuum; much depends on how AI will be regulated and whether designers can learn to work with it rather than against or under it. Architecture certainly doesn’t operate with autonomy in the current world, so AI isn’t about to shatter some pre-existing idyll. It might even act as a force to disrupt the kinds of financial imperatives that dictate so much of our cities today. 

Tokyo’s era of radical ideas about the city’s future was founded on a mid-century material base of state power and planning. As we stand on the precipice of a new age of informational technology, the questions are open: how will AI be incorporated into the design process and will it act to reinforce or disrupt the status quo?

Find out more about INDESIGN #90 and susbcribe here!

We think you might also like this Q&A with Woods Bagot’s Jet Geaghan on AI.