Answer: Voting preferences.
Turns out, if your neighborhood has more sedans than pickups parked on the street, there’s a good chance the overall voting preference is Democratic. An 88 percent chance, to be specific. And on the flip side, if there are more pickups than sedans, the likelihood that you’re surrounded by Republicans is 82 percent.
A deep learning neural network developed and trained at Stanford University figured this out by looking at cars parked on neighborhood streets through Street View in Google Maps. The goal of the team behind this artificial intelligence (AI) was to demonstrate the potential of these deep learning AIs. The algorithm was programmed to make connections between the make, model and year (of every car produced after 1990), and education, income, race and voter preferences based on ground census data and more than 50 million Google Street View images.
“We wanted to show that useful insight can be gained from images, the same way people do this for social networks or other textual-based data,” lead researcher Timnit Gebru told Digital Trends. “Some of the car-politics or car-race associations were intuitive, but still surprising that we could capture from our data.”
A paper on the research findings can be found here.