Bridging or widening the divide? Exploring the role of AI technology in Smart City Development Efficiency and Spatial-temporal Divergence in China

Cities are complex organisms, constantly growing and changing, but as they expand, they often develop chronic ailments: traffic gridlock, polluted air, and strained resources. For decades, the solution has been the "smart city," a concept where technology like sensors and data networks helps manage these urban challenges more efficiently. The idea is that by connecting everything from traffic lights to waste management systems, a city can become a more livable, responsive place. However, the latest engine driving this transformation is artificial intelligence, or AI. While AI promises to make cities smarter, a new question has emerged: does this powerful technology help all cities grow together, or does it create a new kind of inequality where some places pull far ahead while others fall behind? This is the central mystery explored by researchers Zhong Fang, Ying Chen, and Peiqi Shi from Fujian Normal University, who set out to measure how AI is reshaping the landscape of urban development across China.

The researchers approached this question by looking at thirty provinces in China over a five-year period, from 2019 to 2023. They did not simply ask if a city was "smart" or not; instead, they built a detailed two-step evaluation system to understand the relationship between the technology itself and the city's overall performance. First, they measured the "AI foundation" of each region, looking at how much support the technology had, how well it was being used in local industries, and how much new innovation was being created. Then, they measured the "smart city" outcome, which included six specific areas: the technology infrastructure, the economy, the well-being of citizens, the quality of government services, the health of the environment, and the efficiency of transportation. By linking these two steps, the team could see if a strong AI foundation actually led to a better-performing city, or if the technology was failing to deliver on its promise in certain places.

The study revealed a clear and somewhat worrying pattern. On average, the overall efficiency of smart city development in these thirty provinces was quite high, sitting at 0.86 on a scale where 1.0 would be perfect. However, this number was slowly drifting downward, dropping from 0.89 in 2018 to 0.83 in 2023. The researchers found that the primary bottleneck was not the smart city systems themselves, which were actually performing quite well with an average score of 0.96, but rather the AI technology that was supposed to power them. The AI systems lagged behind with an average efficiency of only 0.77. In many cases, the lack of funding for research and a shortage of skilled talent were holding back the technology, which in turn dragged down the performance of the entire city. The data showed a direct link: when AI efficiency dropped, the performance of the smart city dropped with it. This suggests that without a solid, well-funded AI foundation, the promise of a smart city cannot be fully realized.

Perhaps the most significant finding was the emergence of what the authors call a "smart divide." This is a new type of gap where regions with strong AI technology pull further ahead, while regions with weaker technology fall further behind, creating a widening chasm in development. The study showed that this divide is not just about money; it is about the uneven spread of the technology itself. For instance, the eastern regions of China, which are already economically advanced, maintained high stability and efficiency, creating a positive cycle where good technology led to better city management. In contrast, the northeastern region faced a vicious cycle where weak foundations led to insufficient investment, which caused efficiency to stagnate. The researchers observed that the gap between the AI capabilities of different provinces was actually growing, and this technological inequality was spreading to the cities themselves, making the differences in quality of life and governance more pronounced.

To understand how this divide was moving across the map, the team tracked the "center of gravity" for both AI technology and smart city development. Imagine a balance point that shifts toward the areas with the most activity. The researchers found that the center of gravity for AI technology had been moving, showing a pattern of first gathering in specific hubs and then spreading out. By 2020, this center had shifted significantly toward the southwest, reflecting a rise in AI industries in central and western provinces, likely driven by new government policies and the relocation of data centers. However, the center of gravity for the overall smart city performance remained stubbornly fixed in the east, showing that while the technology was spreading, the actual benefits of smart cities were still concentrated in the wealthy coastal regions. In recent years, the AI center has started shifting back eastward, a sign that the gap between the rich and poor regions might be widening again.

The study concludes that while AI has the potential to bridge gaps, it currently risks widening them if left unchecked. The researchers found that the inefficiency in the northeastern provinces was largely due to the widening differences in AI capabilities within that region, while the central and western areas were held back by the lagging performance of their smart city systems. To fix this, the authors suggest that policymakers need to be very specific in their approach. They recommend focusing on boosting research and development funding and attracting talent in struggling provinces like Jilin, Zhejiang, and Hubei. For the northeastern region, they propose a dedicated revitalization fund to stop the "technological depression" caused by talent leaving and investment drying up. The goal is to ensure that the benefits of AI are not just for a few elite cities but are shared across the entire country, preventing the digital divide from becoming a permanent barrier to progress. Ultimately, the paper argues that a true smart city must be centered on the people living in it, using technology to serve everyone equally, rather than becoming a tool that only makes the already wealthy richer.