New problems naturally cause new headaches. This increases the urgency and importance of finding new tools to deal with new challenges. Today, global governance must rise to the challenge of making cities more efficient and livable in an era of exponential increase in the use of digital instruments, including artificial intelligence. As urbanization continues to demand innovative and practicable solutions, the social implications of a rapidly advancing AI need to be navigated constructively rather than merely endured.
The widespread urbanization going on in China requires the application of all available skills in managing the organizational, logistical and energy and transport-related challenges that come with that. Europe’s development in this sense has been gradual, but the challenges are equally complex given the need to adapt lifestyles to EU sustainability goals; China, on the other hand, has been extraordinarily rapid in shrinking the gap with more advanced economies.
Moreover, China has literally millennia of experience in dealing with extreme, large-scale climate events, especially considering monsoon events and river-flow management. In response to current climate changes, major infrastructures must be made compatible with violent and intermittent precipitation. Some partly traditional and time-tested techniques are being used to make cities more resilient to extreme events, in particular, the concept of “sponge” cities in the case of heavy precipitation and/or flooding. It is therefore possible to apply a pragmatic attitude to developing a combination of traditional and highly innovative solutions.
Approaches such as the “15-minute city” strive to make services more efficient for the purpose of limiting travel and fostering a more rational use of urban spaces. Various digital technologies – ICT as well as the more advanced forms of AI – have a fundamental role in this context, which suggests the need for a holistic approach.
On the other hand, the idea of the “green city” already links multiple concepts, aiming for a sustainable form of resilience that fosters reduced human impact with a view to increasing capacity for prevention and adaptation to potentially damaging weather and climate phenomena.
In any case, even though they are regulated by central governments and coordinated – even internationally – for both practical and legal reasons, many efforts to mitigate environmental threats fall directly under local authority, even down to the mayoral level. The result is an imbalance between specific individual measures and the general situation. For instance: while electric bicycles can be an effective urban transportation solution, they often remain incompatible with traffic security concerns and with traditional public services. Thus, detailed analysis must be applied to reconciling environmental, efficiency and quality of life goals.
It is essential to foster shared citizen, community and national government responsibility; in particular, many sustainability measures underestimate short- to medium-term social impact and end up generating a disconnect between general objectives – above all, decarbonization – and the management of urban pollution and the deterioration of water services.
Some daily problems, such as those associated with water supply and control, are being better understood and perceived by citizens, which is where to begin to build broader consensus on sustainability interventions. The relationship between the economy and ecology must be considered with a view to preventing political dynamics from impeding change in the collective mindset; in this sense, the challenge is a cultural one. And the challenge is even more complex when considering the time factor: the acceleration of the macroscopic problem of climate change is forcing authorities to impose measures that may appear too costly but are rendered indispensable by the scarce attention paid to the problem in the past.
In order to incorporate AI into urban governance – i.e., “smart cities” – it is necessary to fully understand the basic features of this family of technologies. The AI systems currently available can no longer be considered “embryonic” but rather at the “teenage” development stage: peak learning capacity but with various issues yet to be addressed, indicating the need for a certain prudence in adopting them for use. In particular, the black box problem is still unresolved, since human programmers and users are not yet capable of truly understanding how AI makes some decisions.
Yet, standards can be developed with a view to protecting users through output control and the clear establishment of liability for the tangible effects of algorithm use. In a broader sense, it is critical to educate the public about the potential of these systems and how they work in order to reduce the risks of manipulation and incorrect interpretation. Ensuring a relatively equitable distribution of AI-based services will be equally important to effectively incorporating this technology into the social and economic fabric, limiting its more negative collateral effects, not least in terms of skills that are not going to be easy to acquire.
Another underlying uncertainty regards the real availability of sufficiently reliable and precise data to develop AI-skilled applications that depart solely from algorithm training systems. Neither is internet access capable on its own of ensuring that the quality and quantity of data are sufficient and adequate. Moreover, it is becoming clear that the true weakness of current systems is found at the level of deductive reasoning, which goes well beyond the inductive logic typical of data-driven algorithms. In any case, data sharing remains a controversial issue that calls for innovative international governance.
A large number of daily-use objects are already “smart”, at least in the sense that they are interlinked and/or connected to the digital network. These objects are therefore to some extent “active” in a manner that differs from that of the past, in the sense that they have computational capacity and can interact with users in a customized way. This has brought a potentially enormous change to the “smart cities” picture; the process already underway is unregulated and quasi-organic, but additional progress at the level of sustainability can only be achieved by means of cautious programming. AI use on a large scale poses a greater challenge to our societies than previous computerization phases given that it requires a series of autonomous actors, not merely additional digital filters in relations between human beings.
The technical complexity of AI systems is producing a paradox: regulatory mechanisms themselves tend to be at least partly developed by artificial intelligence. This is probably an inevitable development that refers back to the reliability of software and the issue of algorithm transparency – which naturally pits private manufacturers interested in protecting copyright against public authorities interested in protecting individual and collective citizen rights.
In the general scheme, one of the most promising areas of AI application to life quality is that of health and medicine, given the existence of extensive digitalized databases for the development of better diagnostic systems and more personalized therapies. The sector is also extremely important in economic terms, which nevertheless involves a delicate trade-off between collective objectives and private interests (including corporations, which are the essential actors in innovation).
Continuous cooperation between the EU and China on all these closely interlinked considerations will provide a tangible contribution to the management of innovative processes. Since there are no pre-constituted formulas for unprecedented issues, every political, economic and values-oriented system can offer a useful perspective.