AI will Transform Cities. Can Cities Transform AI?
Jinhua Zhao, J. Phillip Thompson, Ross Gittell, Michael Leong, and Kevin F. Hsu
Introduction | What's in this paper? | Executive Summary
Introduction
Cities and new technologies have always been defined by a special kind of relationship. Urban density and complexity provide the means for new technologies to develop and scale rapidly, and these technologies, in turn, influence the economics, institutions, and social functions of cities. The rise of artificial intelligence (AI) is no exception to this rule, except that it is evolving faster than any technology in recent memory.
In this moment, municipal governments are faced with a complex set of overlapping pressures. On one hand, AI offers significant opportunities to improve public service delivery, support economic growth, and empower residents and businesses. On the other hand, city leaders are wary of ‘losing control’ of AI, and are on the front lines of protecting communities from its risks. These include threats to privacy and civil rights, weakened democratic oversight, widening income inequality.
In national and international AI policy discussions, however, cities are often overlooked or constrained by preemption, even though they bear many of the real-world consequences of these decisions. We believe that this is a serious oversight that must change. Instead, municipal and city government’s involvement with the practical challenges of implementation gives rise to unique insights and strategic leverage. As such, they have an active role to play in coordinating local stakeholders, building public trust, and guiding AI deployment towards beneficial applications that improve human welfare, strengthen democracy, and enhance the quality of life.
What's in this paper?
In this moment of rapid change, this paper outlines three key questions and corresponding actions for city governments:
Part 1. What does “success” look like for AI and cities? (jump to Part 1)
Does the pursuit of increased efficiency define success alone, or are there other angles such as public trust, human agency, and control of democratic processes? We propose a multi-modal definition of success that offers cities a ‘moral compass’ to evaluate the AI tools, policies, and strategies they pursue. We also articulate the key levers cities have, and the key capacities cities must build, to accomplish this vision.
Part 2. How do cities articulate “good” and “bad” uses of AI? (jump to Part 2)
Do cities have a clear process to understand what types of problems AI can solve, what types of AI tools are best suited to tasks, and how to assess their risks? Are they adopting AI applications that address real municipal problems, or purchasing “solutions in search of problems”? We propose a comprehensive 4-step process to articulate municipal challenges, select suitable AI technologies, thoroughly evaluate implementation risks, and assess consistency with a city’s normative and strategic objectives for AI.
Part 3. How do cities capture and share the value that AI promises to create? (jump to Part 3)
Can cities ensure that its own residents benefit from broader economic effects of AI, boosting human capabilities and human capital instead of simply automating away jobs or exacerbating social inequities? We believe that municipal governments have a unique role in developing the intellectual capacity, physical infrastructure, and civic ecosystems that enable prosperity to be shared among workers, residents, and businesses.
AI is already transforming the technological, social, economic, and political terrain in which cities operate.
With far-sighted leadership, creative collaboration across sectors, and a clear commitment to the public good, we believe that cities can in turn transform AI—and shape AI’s influence so that it strengthens, rather than destabilizes, the civic sphere.
This is Version 1.2 of our White Paper. We welcome your feedback on this product, which will continue to evolve.
Explore the sections of the paper below
AI will Transform Cities. Can Cities Transform AI?
Jinhua Zhao, J. Phillip Thompson, Ross Gittell, Michael Leong, and Kevin F. Hsu
1. What does “success” look like for AI and cities?
-
Three factors offer a ‘moral compass’ for AI adoption: Does AI improve efficiency and productivity? Does AI protect privacy and improve human agency? Does AI advance community interests and democratic governance?
-
Over-optimizing any one of these factors can lead to suboptimal end-states. Effective AI governance ensures AI improves all three factors harmoniously, even if none is fully optimized.
-
Cities need not be “passive” takers of solutions; they have four key levers to influence the direction of AI development: Expertise in data & administration, the power to convene, direct authority over essential domains, and procurement contracts.
-
Cities must develop internal capacity to join the conversation effectively and work to transform AI, including technical expertise, data governance and infrastructure, education, public trust, and strategic partnerships.
2. How do cities articulate “good” and “bad” uses of AI?
-
Step I: Identify the real problems facing public agencies that AI could solve. A framework of common data and information flows across public agencies to identify problems.
-
Step II: Determine an appropriate human-AI persona and AI technology. A framework to selecting an AI technology and its relationship to humans, and assessing if it solves a problem.
-
Step III: Evaluate the risks associated with deployment. A framework to articulate and assess possible risks of AI deployment given scope, reliability, and consequences of mistakes.
-
Step IV: Evaluate whether AI advances the city’s vision of success. Reconnects with the ‘moral compass’ to revisit if AI is genuinely the right tool for the task.
3. How do cities capture the value that AI promises to create?
-
Knowledge of AI: Cities need to develop AI literacy at four different levels — at a baseline public level, at a professional level, at a managerial level, and at a regulator level.
-
Physical & Digital Infrastructure: Cities need to develop capacity to store and manage data, securely deploy apps, enable near-universal access, and engage solution providers.
-
Public Trust & Civic Consciousness: Cities need to develop policies to address primary (individual), secondary (societal), and tertiary (environmental) concerns about AI, and should establish a civic AI board and partnerships with various stakeholders to achieve it.
-
Sustainable Source of Funding: Cities should find a source of revenue to fund programs that manage AI impacts and upgrade intellectual, physical, and civic capacity for the AI age.