Reimagining Roads To Be Walkable and Bike-Friendly, With Help From AI

Streets that cater to pedestrians and bikers not only offer more commuter options, but they also create areas that look and feel community-oriented—as evidenced by Brooklyn-based artist Zach Katz’s new project. Using the second version of DALL-E, the AI system from OpenAI that generates images and art from written descriptions, Katz feeds the program an image from Google Street View with details on how the street should look instead. The program generates a new, redesigned version that often inputs sidewalks, biking lanes, fountains and other public spaces. The new street attests to how accessible areas are welcoming and desirable. Since sharing his project, many people, including council members and urban planners, have reached out to the artist with requests for reimagined roadways. Learn more about how Katz hopes the project can help envision better cities at Bloomberg.

Image courtesy of Zach Katz/OpenAI