**1. Introduction**

### *What Is an Artificially Intelligent City?*

During the current Anthropocene era—the geological epoch which has had significant human impact on Earth's geology and ecosystems—we have developed technological capabilities that have enabled us to greedily use limited natural resources for economic profit [1,2]. This ruthless capitalist

practice not only brought about anthropogenic climate change, but also caused socioeconomic inequalities to soar globally [3].

In recent years, technology, as part of knowledge-based development efforts [4], has been viewed as the solution to severe environmental, economic, and social crises [5,6]. Consequently, the smart city concept has come to the forefront of discourses on urban planning and development [7]. Accordingly, many see emerging technologies, particularly artificial intelligence (AI), as a way to safeguard our civilization from the catastrophic consequences of climate change [8], biodiversity loss [9], natural disasters [10], unsustainable development [11], pandemics [12], and so on.

Simply, AI is defined as "machines or computers that mimic cognitive functions that humans associate with the human mind, such as learning and problem solving" [13]. AI-driven computational techniques are diverse and range from rule-based systems to deep learning systems. A popular AI knowledge map was created by Corea [14]. His conceptualization brings together the AI paradigms and problem domains (Figure 1).

The AI paradigm and its technology-enabled solutions—whether it is autonomous driving, home automation (so-called domotics), robotics, chatbots, or advanced data analytic tools—have opened up new opportunities for cities, where most of the world population resides, where most of the production and consumption activities take place, and also where most of the negative environmental externalities are generated [15,16]. While some scholars see AI as an opportunity to advance smart cities (or smartness of cities) [17–21], others see AI generating a whole new city brand, especially when the AI applications become mainstream in our cities [22]. In other words, in the near future, we will see a trend to build 'artificially intelligent cities' from scratch, or to retrofit traditional cities, converting them into artificially intelligent ones.

We define an artificially intelligent city as an urban locality functioning as a robust system of systems, and whose economic, societal, environmental, and governmental activities are based on sustainable practices driven by AI technologies, helping us achieve social good and other desired outcomes and futures for all humans and non-humans.

In the age of smart cities—where urban locations are starting to be wired with smart technologies including sensor networks—and given the highly sophisticated capabilities of AI, we foresee a potential dramatic change in our cities and societies [23]. There is, hence, an increasing need to investigate the urban and social implications of AI. This is an understudied area of research, and a gap in the literature on AI and city/society.

We also note that there are different levels of AI, including: (a) reactive machines (e.g., IBM's Deep Blue); (b) limited memory AI (e.g., chatbots, virtual assistants, self-driving vehicles); (c) theory of mind AI (a concept that is in progress at the moment); and (d) self-aware AI (only hypothetical at this stage) [24]. There is also another categorization of levels of AI, such as: (a) artificial narrow intelligence (represents all of the existing AI today); (b) artificial general intelligence (its main idea is that AI agents can learn, perceive, understand, and function completely like a human being); and (c) artificial superintelligence (an idea that AI replicates the multifaceted intelligence of human beings and becomes exceedingly better at everything it does) [25]. The disruption of each level of AI will be different in our cities and societies. Throughout this paper, we focus on the current level of AI: artificial narrow intelligence.

Against this backdrop, we prepared this viewpoint in order to help in bridging this gap along with promoting further research on the topic. In this paper, we introduce a provocative artificially intelligent city notion as the potential successor of the currently popular smart city concept, where city smartness today is increasingly depending on the use of viable technology solutions, including AI. The paper, by placing the AI literature, developments, trends, and applications under the microscope, provides a commentary on whether building artificially intelligent cities can safeguard humanity from natural disasters, pandemics, and other catastrophes.

### **2. Conceptual and Practical Background**

### *2.1. Has the Artificial Intelligence Era Already Begun?*

AI is one of the most disruptive technologies of our time and its capabilities have progressed rapidly [26]. The uptake of AI in organizations is on the rise. For instance, between 2018 and 2019, the number of organizations that deployed AI grew from 4% to 14%, and among the AI applications, conversational AI is at the top of corporate agendas spurred by the worldwide success of Amazon Alexa, Google Assistant, and Apple's Siri [27]. Gartner [28] provides insights into the hype cycle for AI applications, which reflects the growing popularity of machine learning, intelligent applications, and AI-as-a-Service (AIaaS) or AI-Platform-as-a-Service (AI-PaaS) (Figure 2).

**Figure 2.** Hype cycle of AI applications, derived from Gartner [28].

In recent years, governments around the world have started to see AI as a nation-defining and global economic competitiveness-increasing capability [29]. In recognition of the increasing importance of AI, as of February 2020, 50 countries have already developed specific national AI strategies—where these countries represent 90% of global gross domestic product (GDP). Figure 3 illustrates the location of these countries; a brief further info on each country's national strategy is provided in Appendix A (Figure A1).

AI-driven computational techniques are diverse and wide-ranging. For example, AI has been in use for quite some time in the tasks that are risky or cause harm to humans. This includes the use of automated robots for bomb detection or combat of unmanned aerial vehicles, and the use of autonomous trucks in the mining industry or mobile reconnaissance units for space exploration [30].

AI-enabled applications include robotic processes [31] for automating public sector tasks, and autonomous delivery bots [32] and chatbots [33] for enhancing business intelligence, stakeholder engagemen<sup>t</sup> experience, and customer service quality. Today, AI is rapidly changing the nature of jobs. Many of the services that have been offered by human workers are now being revolutionized by technology. For example, chatbots automate the work of information technology (IT) professionals [34] and human resource (HR) departments [35], so that they can focus on higher value tasks.

**Figure 3.** Countries with a national AI strategy, derived from Holon IQ [29].

Autonomous vehicles and driverless shuttle buses are being trialed worldwide. Driverless shuttle bus services are expected to start carrying fare-paying customers in Scotland later in 2020 [36]. Likewise, robot police services are planned to be launched in Houston, Texas to curb petty crime and free up law enforcement resources in 2020 [37].

AI-based systems are providing various solutions. These solutions facilitate the creation of new products and services in many different fields. Particularly, sensor networks are undergoing grea<sup>t</sup> expansion and development and the combination of both AI and sensor networks has now become a reality to change our lives and our cities. The integration of these two prominent technologies— including AIoT (AI-of-Things)—also benefits other areas such as Industry 4.0, Internet-of-Things (IoT), demotic systems, and so on [38,39].

AI is being employed to model the spread of COVID-19 to assist decision makers in understanding the future implications of the virus and the measures that should be taken to limit its spread [40]. For instance, in China, AI is being used to minimize the spread of COVID-19 by mobilizing robots that do cleaning and food preparation tasks [41]. Moreover, the European Union [42] launched the EU vs. Virus challenge via a Pan-European hackathon to find ways to tackle COVID-19 via AI and other applications.

AI also has the potential to help in addressing some of the planetary challenges (Table 1). The World Economic Forum [43] underlines the following eight AI applications as "game changers": (a) autonomous and connected electric vehicles; (b) distributed energy grids; (c) smart agriculture and food systems; (d) next-generation weather and climate prediction; (e) smart disaster response; (f) AIdesigned intelligent, connected, and livable cities; (g) a transparent digital earth; and (h) reinforcement learning for earth sciences breakthroughs.


**Table 1.** AI application areas for addressing planetary challenges, derived from World Economic Forum [43].

### *2.2. How Is Artificial Intelligence Being Utilized in Cities?*

In the previous section, we have provided some examples of the use of AI in cities. Here in this section, we share a few more examples to cover some of the other aspects of AI for cities. In particular, the AI solutions implemented in Australia have been taken as an example. Like many other advanced knowledge and innovation economies, AI is a rapidly growing field in Australia. Furthermore, the country has been an early adopter of smart technologies [44], particularly for targeting industrial and urban sustainability outcomes [45–47]. Some of the existing AI applications and experienced challenges in the country are discussed as follows:



Despite these issues, development of AI continues in a variety of fields in Australia, and has been investigated for its use in product/goods delivery [52], environmental and transport monitoring [53], disaster prediction [54], healthcare [55], infrastructure [56], data privacy [57], and agriculture [58]. Just to provide some examples, AI's contributions to healthcare practice are listed in Table 2. Additionally, AI applications have been used in big data analytics, such as its use in social media analytics to aid natural disaster management. Figure 4 is an example of the disaster severity map generated for the 2010–2011 Queensland Floods with the help of machine learning technology [59].

**Table 2.** AI applications and motivation for adoption in healthcare practice, derived from Park [60].


**Figure 4.** AI and big data analytics in natural disaster management, derived from Kankanamge et al. [59].

In terms of strategizing AI, there have been some promising developments in Australia. The most notable one is the AI roadmap, codeveloped by CSIRO's Data61 and the Australian Government Department of Industry, Innovation and Science. The roadmap identifies strategies to help develop a national AI capability to boost the productivity of Australian industry, create jobs and economic growth, and improve the quality of life for current and future generations. The roadmap emphasizes the need to concentrate on the three key domains: (a) natural resources and the environment; (b) health, aging, and disability; and (c) cities, towns, and infrastructure [61]. Table 3 below elaborates the objectives of these AI domains. Additionally, OECD's [62] AI policy observatory provides a useful repository of AI in Australia.

**Table 3.** Priority AI specialization domains and their objectives, derived from Data61 [61].

