The Challenge to ‘Creative Cities’

Geographers have long observed that creative economic activities tend to cluster in space. Despite technological advances and books such as The World Is Flat by Thomas Friedman, the need for workers to gather in person has persisted, especially for activities requiring creative thinking and input, such as finance and software innovations and film production.

But the advent of AI has put a new question mark over clustering. Professor Patrick Adler in the Department of Geography, who specialises in the creative economy, has started to grapple with the issue.

The digital age and the emergence of diffusion technologies, such as the Internet, personal computers and mobile devices, were predicted to replace the need to cluster. This did not happen because revenue and new technologies continued to be generated from clusters, and teams still needed to meet face-to-face to function properly.

“With AI, though, something transformative is happening because it lowers the number of people you need on a team. You don’t need to be centred anymore. Maybe,” he said, “this time will be different.”

He pointed to the prediction, first mooted two years ago by OpenAI’s CEO Sam Altman, that a one-person, US$1 billion company is on the horizon – something that would be built by an individual using AI with no other humans explicitly involved.

“It’s very instructive that OpenAI is very small, that the DeepSeek team is very small. You would have thought they would need thousands of people, but they don’t. If there are smaller teams, there will be less need to concentrate physically. So it appears AI could become a decentring force,” he said.

Lessons from the recent past

On the other hand, there is the traditional importance of teams and collaboration to creative output, especially when it comes to complex products such as films and new phone technology. Being in close proximity not only makes workers more productive, it makes them perform better, he said. They establish common standards, which is important for maintaining and increasing quality. Knowing those standards and not having to explain them to each other can also speed up production. Clustering also fosters trust and encourages innovation.

“If you’re living in the same physical community as your buyers or the people you’re buying from, you’re going to be able to take more risks because you trust more. And there’s the idea that you will treat each other fairly today because you’ll have to encounter each other again tomorrow. When you’re less physically proximate, that’s less the case.”

The recent past has also shown that, if anything, technology might make proximity even more important. Despite the advent of email and instant messaging, the most important decisions still get made face-to-face.

“We never would have predicted that – the idea would have been that more work would happen at a distance because of the technology. But somehow, this increased the value of being close to the centre. People who are near head office do better than people who are far away. Another countervailing point is that the AI companies, like Nvidia and OpenAI, are all coming from the Bay Area, and they’re even more concentrated than in previous technological waves,” he said.

Hong Kong also has a concentration of people involved in what Professor Adler calls ‘creative inputs’ – using creativity and cognitive powers to generate value in finance, consulting and the like. The Greater Bay Area is starting to see clustering similar to Silicon Valley, while Korea has become a global hub of creative products.

Transitions matter

Still, AI threatens to disrupt things. To avert that, he believes the transition to the technology needs to be handled carefully, especially given the current unevenness of its adoption. In areas such as coding, it is replacing core work, while in fields like film production, adoption is patchy – visual effects specialists and film editors are embracing it, but screenwriters are shunning it. AI adoption may also require additional time spent interacting with it and monitoring its output, at least initially, which could reduce productivity. These transitional costs should not be left to chance, though.

“Transitions matter. It’s a mistake to think that we will just eventually adopt these technologies, because while that may be true, the ones we adopt really matter. The wrong technology could predominate,” he said, citing the example of Betamax, which was technologically superior to VHS [Video Home System] but stumbled at the market.

Creative industries therefore need to recognise that AI is here to stay and act on that. “The question is, is there something transformative about AI where people don’t need to be centred anymore? You don’t want to be historically naïve, but these technologies are qualitatively different and they are disruptive. We’ve been here before, so it’s worth thinking through what happened in the recent past,” he said.