The term “smart cities” refers to the use of technology to improve the quality and efficiency of urban public services. Hong Kong’s Smart City Blueprint 2.0 outlines a commitment to becoming a smart city, including in transport, health, education and the environment. With nearly US$200 billion expected to be spent on smart city platforms worldwide this year, it is appropriate to ask whether the idea is worth the price tag.

The first generation of “smartness” focused primarily on technology applied to government operations and infrastructure functionality. The purpose was efficiency and, in some cases, data collection. The second generation saw a blending of technology and lived experience. In Hong Kong, programs such as “iAM Smart” and Covid-19 tracing apps showed how smart city ideas can be brought directly to people’s daily lives.

In a recent survey, I explored Hong Kong residents’ perceptions about quality-of-life improvements that could be gained from smart cities. Looking across common urban service sectors, I found that trust in the technology is higher than trust in the related policy systems. I also found that respondents’ faith in smart cities was higher for rudimentary issues such as mobility and transport than for more complex issues such as governance transparency and participation.

These findings are common around the world. There is also rising scepticism in some countries about smart city technology itself, particularly where it is seen as an instrument of government control. These trends come at a turning point in technology history. We have now moved to a third generation of smart cities – urban governments using artificial intelligence.

As with most waves of technological advancement, this one did not appear overnight. Versions of AI have operated for years in spaces that are not as visible or useful to the general public. As AI has broadened its applicability, its purpose has evolved from the traditional objectives of engineering optimisation and data collection to improving people’s lives.

One example of AI is ChatGPT. The effects have been immediate, and – in the case of education, for example – universities are scrambling to establish protocols for what is acceptable use in research and student assignments.

Smart mobility is another direct public experience with AI. This month, the newly created Smart Mobility Technology Alliance held a forum focused on how mobility and transport are affected by AI. Panellists discussed the applicability of AI to autonomous vehicles – not only private cars but also buses, street cleaners, delivery trucks, construction equipment and port vehicles. According to one prediction, efficiency gains could be 30 per cent or more for certain tasks and industries, with vehicles working 20 or more hours per day.

Full realisation of smart mobility is reliant on C-V2X, a platform for vehicles to communicate with nearby infrastructure, data networks, other vehicles and people.

This all requires significant investment in public infrastructure and installed capacities. Roads must be retrofitted with sensors and network capabilities. Other enabling factors are low-latency LTE coverage and sufficient radio wave allocation.

Trials for autonomous vehicles in Hong Kong have shown promise, but seamless operation citywide requires infrastructure at scale. The technology is ready, but bottlenecks lurk in practical implementation. Beyond hard infrastructure, smart mobility requires adjustments to policies and regulatory standards in transport, technology and the environment.

Depending on how it is implemented, smart mobility also raises potential user concerns. For example, Tesla cars can monitor users’ driving behaviour in real time. Profiles that determine the price of insurance premiums are built using related data. This approach is not new, but the level of depth and detail raises some privacy concerns.

The pathway to integrated smartness on Hong Kong’s roads could be long and costly. Our public transit system is arguably among the best in the world. How much better do we need it to work? On roads, smart mobility might improve safety, but can it also reduce traffic delays? Overall, will the gains of smarter cities be marginal or transformative, and will they be enough to motivate behavioural change and justify substantial fiscal commitments?

As I said in my study, “the current wave of techno-fetishisation can lead governments to over-promise about the benefits of smart cities and double down on a narrow and technocratic approach to measuring and framing policy problems.” The Hong Kong government will need to craft a convincing narrative that operational efficiencies will translate into a better quality of life for residents.