SF Chronicle Maps Paint Colors of 115,000 Homes Using AI

A custom artificial intelligence tool built by the San Francisco Chronicle scanned the exterior paint colors of roughly 115,000 San Francisco homes, producing an interactive map that reveals which blocks and neighborhoods carry the city's most distinctive palettes.

The project, published March 7, combined data analysis with visual storytelling to quantify something residents have long debated informally: just how colorful is San Francisco, and where does that color concentrate? The Chronicle's tool processed exterior imagery at scale, cataloguing hues across the city's varied housing stock to generate what amounts to a chromatic census of San Francisco's built environment.

The resulting interactive allows readers to explore the data neighborhood by neighborhood, surfacing patterns in how paint choices cluster geographically. San Francisco's Victorian and Edwardian rowhouses, particularly those in the Castro, Haight-Ashbury, and the Painted Ladies corridor along Alamo Square, have long drawn attention for their elaborate multicolor schemes. Whether the data confirms those corridors as statistical outliers, or surfaces unexpected color-dense pockets elsewhere in the city, the map offers a data-driven answer to a question that has previously relied on reputation and anecdote.

The project reflects a broader trend in local journalism toward deploying machine learning tools on civic and cultural datasets that would otherwise require prohibitive manual labor. Cataloguing the exterior colors of 115,000 structures by hand would be logistically impossible; the Chronicle's AI approach compressed that work into a scalable analysis.

For a city whose architectural character is inseparable from its color choices, the project also raises questions that extend beyond aesthetics. Paint decisions are tied to neighborhood change, property investment, and the economic pressures reshaping San Francisco's housing. A block-level map of color density is, in effect, also a partial record of where resources flow and where they don't.