PhD student, Tilburg Institute of Law and Technology, UdeM, Mila
To fully benefit from the potential of artificial intelligence, city officials need guidance. This new white paper aims to provide that.
Digital transformation is occurring at breakneck speed, and artificial intelligence (AI) is a big part of it. Cities and urban areas are fast becoming testing grounds for many new forms and applications of AI — from transportation, energy, and waste management to public safety, health care, and city governance.
To benefit cities and urban areas, AI needs to be integrated ethically and responsibly. That’s the central thesis behind a new white paper called AI & Cities: Risks, Applications and Governance. Published by UN-Habitat, the United Nations organization working on sustainable urbanization, and Mila — Quebec Artificial Intelligence Institute, this collaborative effort offers insights and guidance on how to harness AI to build socially and environmentally sustainable cities.
To benefit cities and urban areas, AI needs to be integrated ethically and responsibly.
How AI can be implemented to benefit society
“As a general-purpose system technology like electricity or the combustion engine, AI has both direct applications and the ability to complement other innovations,” says Shaz Jameson, PhD student (Tilburg Institute of Law and Technology, UdeM, Mila) and one of the report’s lead authors. “This means that there are many things that AI can do to support us in urban settings. For example, AI is very good as a predictive tool, particularly around electricity, water, and resources in the context of our changing climate landscape,” she says. In the area of energy, AI can be used in forecasting energy generation, optimizing the system, or predicting when maintenance might be required on energy infrastructure. “AI can also be a good predictive risk management tool in numerous public safety applications like fire or flood risks,” she says.
In transportation and mobility, AI can be applied to make public transportation faster and easier to use, or to predict road and track degradation on transportation infrastructure. It can be used to improve vehicle routes for waste collection or monitoring and optimizing water treatment processes. There are also many beneficial applications of AI in the health-care sector, from drug discovery to supporting case workers to allocating resources for patient care. And in areas of urban planning, AI can be used for data generation for mapping settlements or monitoring population growth.
Awareness of risks key to successful AI implementation
While AI and AI-enabled solutions offer myriad new opportunities for cities, they also pose potential for risks, such as bias and discrimination, amplification of existing inequalities, privacy concerns, and human rights violations through surveillance. “Another risk is that AI doesn’t include nuance,” says Jameson. “It cannot self-evaluate and doesn’t have an idea of good or bad. While it has an idea of its objective and whether it’s deviating from this objective, it’s up to human oversight to put that into context in relation to what we as a society consider to be desirable. That’s why Mila focuses on a responsible approach that considers the values we all care about in each of the AI design steps,” says Jameson.
The aim of the white paper is to provide local authorities with the tools to assess where and how AI adds value and contributes to the development of people-centred sustainable cities and human settlements and where there might be risks. “AI is not a simple plug and play, all-in-one solution, but part of a larger strategy that, when designed well and contextually integrated with input from people on the ground, can be an incredibly supportive tool with a tremendous amount of potential,” says Jameson.