In conceptualizing and exploring the city we rely a range of smaller areas—neighbourhoods, boroughs, wards and districts—in order to make urban space intelligible. While we can readily discuss how neighbourhoods are shaped by physical geography (topography, adjacency to lakes or rivers, etc.), ordinance (zoning, access to public transit) and economics (real estate prices, average resident income), machine learning does not really spring to mind when we are considering how we might define 'a neighbourhood'. Livehoods is a new project hatched within the School of Computer Science at Carnegie Mellon University that leverages 18 million Foursquare check-ins to draft up new urban 'activity zones' based on the patterns of frequent visitors. The venture essentially asks how does a location-based service reflect our sense of place within the city?