This paper explores how 15 Canadian local governments of various sizes and contexts are transitioning to a circular economy by analysing their roadmap currently in development. It provides qualitative insights into how physical, socioeconomic and institutional factors are influencing the content of roadmaps, along with their similarities and differences. Drawing from the literature on the geography of transitions, we show that while local physical and socioeconomic attributes typically shape the roadmaps by determining likely activities, their actual trajectory varies based on the roadmap instigators and the broader institutional contexts in which they operate. The findings suggest the importance of local governments supporting the roadmap instigators while also recognising that circular economy transition pathways can capitalise on policies and programs not only locally but beyond local boundaries.

This study explores the potential of visual generative artificial intelligence (visual GenAI) in augmenting city design workflows. Using customized DALL-E 3 interfaces, we facilitated engagement sessions with members of an academic planning community to assess their perceptions of AI-generated imagery before and after its use, with a focus on main street revitalization (n = 24 qualitative, n = 17 quantitative). Drawing on the Technology Acceptance Model, we assessed cognitive, operational, and participatory dimensions influencing user attitudes toward AI-assisted urban design. Perceived usefulness in cognitive and participatory tasks emerged as the strongest predictors of attitudes toward visual GenAI use, explaining up to 71% and 44% of the variance, respectively. While participants valued the ability to generate visuals and stimulate dialogue rapidly, challenges with prompt precision, output predictability, and interface usability limited broader accessibility. User expertise moderated perceptions, with higher proficiency participants generally expressing more positive attitudes toward its use. Our preliminary findings suggest that while visual GenAI may offer new opportunities to augment cognitive and co-design processes, its integration into city design workflows may also depend on diverse training datasets to address biases; human-centered design with clearer affordances and support for non-expert users; and, validation processes that maintain human oversight. This study contributes to the emerging research on human-AI work integration by providing initial empirical evidence on the opportunities and constraints of visual GenAI tools in city design contexts, while establishing a foundation for future research.

Welcome to this March 2024 Special Issue on Advancing Sustainable Development Goals with AI This issue features research on how AI and data technologies can be used to advance the United Nations (UN) sustainable development goals (SDGs) [1] .