Investors weren’t interested in Tosh Velaga’s idea when he started pitching it this summer. Facial recognition software for retail stores to catch shoplifters made sense in theory, but people were skeptical it would work. So Velaga and his co-founder Igor Nefedov did what entrepreneurs do best: they pivoted.
“We were like, ‘what if we repurposed this for something stupid, like AI that could detect the faces of investors?’” Velaga remembers.
This month, the resulting project finally made it onto the Play Store. It’s called AngelFace, an Android app that lets users identify whether someone is a venture capitalist by capturing a quick photo of their face. According to Velaga, “You just hold your phone up to someone’s face for a second, tap a button and their profile will pop up.”
Velaga and Nefedov scraped photos of investors from Signal, a directory of venture capitalists in different industries, as well as Google Images. They declined to specify how many photos they have, though they said it is over 1,000.
Velaga believes the app can solve a common issue for new entrepreneurs: meeting and talking to the people who can fund their ideas. “Part of it is like if you see somebody walking down the street in Allbirds and a puffy vest, you might be like who is this? VCs are not the most sympathetic crowd and it’s hard to just go up and talk to them. Now you at least know who they are,” he said.
When we tried out AngelFace at Vox’s office, the results were not impressive. The app didn’t recognize Casey Newton (not surprising) or Benchmark’s Bill Gurley (quite surprising). But whether it works is almost beside the point: peer-to-peer facial recognition software is still in its early stages and it’s likely to get more prevalent as time goes on. As Velaga said, “Anybody can make this technology — the tech we use, someone could figure it out watching YouTube videos.”
Today, AngelFace is focused on venture capitalists based in the Bay Area, though Velaga hasn’t ruled out the possibility of expanding to other cities. He’s marketing the app cautiously because, as he says, “it’s a slippery slope, this technology. We’re not even sure if it’s legal.”
In California, it likely is. Facebook, Google, and Apple already use facial recognition to detect who’s in your photos, Snapchat uses it to match your face with selfie filters.
But facial recognition in the public sector is more contested. While cities like New York and San Diego continue to employ this technology in their efforts to police cities, San Francisco and Oakland banned it from being used by law enforcement this past year. Illinois has required affirmative user consent under the Biometric Information Privacy Act (BIPA), which limits the use of facial recognition and biometrics, since 2008.
Despite these regulatory leaps, Americans say they prefer the use of facial recognition by law enforcement than by advertisers or corporations. A recent Pew survey found that while 56 percent trust facial recognition in the hands of police, only 36 percent trust it in the hands of tech companies.
Adam Schwartz, senior staff attorney at the Electronic Frontier Foundation, says he sees both these uses as dangerous. “People expect they should be able to walk around without being tracked by other people,” he said. “We can’t change our face, it’s the ultimate tracking tool.”