MIT Researchers Publish Visual AI Book, Citing 331 Cameras and Privacy Risks
Updated
Updated · MIT News · Sep 24
MIT Researchers Publish Visual AI Book, Citing 331 Cameras and Privacy Risks
1 articles · Updated · MIT News · Sep 24
Summary
Routledge this month published “How AI Sees the City,” a book by MIT Senseable City Lab researchers arguing visual AI can scale urban observation far beyond traditional fieldwork.
One example used 331 New York City traffic cameras and machine learning to identify vehicle types and estimate emissions, showing how images can track traffic, safety, public-space use and greenery at citywide scale.
The authors say the same methods can extend indoors: a separate study of 400,000 Airbnb listings found interior design styles still vary significantly by geography rather than converging globally.
Privacy and bias remain central constraints. London has about 210 cameras per square mile, while Shanghai has more than 5,000, and the book warns constant monitoring can erode personal freedom and reinforce social biases in AI systems.
Framing AI as a new tool in a long urban-studies tradition, the researchers argue cities should use it cautiously and critically rather than treat it as neutral.