Cutting Edge: Apple’s purchase of Emotient fuels artificial intelligence boom in Silicon Valley
The arms race in Silicon Valley is on for artificial intelligence.
Facebook is working on a virtual personal assistant that can read people’s faces and decide whether or not to let them in your home.
Google is investing in the technology to power self-driving cars, identify people on its photo service and build a better messaging app.
Now Apple is adding to its artificial intelligence arsenal. The iPhone maker purchased Emotient, a San Diego maker of facial expression recognition software that can detect emotions to assist advertisers, retailers, doctors and many other professions.
Apple confirmed the purchase this month, but did not give details of the deal. Emotient could not be reached for comment.
The purchase comes during fevered interest in artificial intelligence, or A.I., which relies on algorithms powered by troves of data to bridge the gap between computers and natural human thinking.
This technology just gets smarter and smarter. There’s been a massive amount of change.
— Dan Kara, research director for robotics, ABI Research
Image recognition technology similar to Emotient’s is helping law enforcement identify terrorists. It’s allowing banks to authenticate customers. And it’s telling marketers which TV shows are worth advertising with.
The technology is set to take off now that the nation’s biggest tech giants are involved, and in some cases giving away their software free to encourage developers to help innovate the programs.
“This technology just gets smarter and smarter,” said Dan Kara, research director for robotics at ABI Research. “There’s been a massive amount of change. If you look at the funding in A.I. as recently as 2010, it wasn’t nearly what it is today.”
Facebook founder and Chief Executive Mark Zuckerberg said this month that he was attempting to build an A.I. system to control his home. Google, Microsoft, Amazon and entrepreneur Elon Musk have all recently made high-profile forays into research and development of A.I.
Apple already has Siri, a digital assistant that responds to users’ spoken questions. And it has also purchased A.I.-focused companies such as Perceptio, VocalIQ and Faceshift.
Several past projects from Emotient could easily be viewed as beneficial to Apple. The San Diego company built a prototype sentiment analysis application for Google Glass. It also worked with Innerscope Research to analyze the reaction of focus groups to ads shown during the Super Bowl, something that could have been interesting during its famous 1984 commercial.
Although Apple declined to say what it plans to do with the start-ups it has acquired, analysts think they could be incorporated into improving Siri or developing augmented or virtual reality products.
Emotient was founded by six scientists from UC San Diego in 2012. It uses software that recognizes facial patterns as they come into view of a camera. It had previously raised $8 million from investors, including Intel Capital.
Emotient’s software can detect emotions including joy, disgust, anger and surprise — sometimes more accurately than humans can. It mainly sold its product as a way to give real-time feedback to advertisers and retailers.
“We think that will be increasingly important in a world where how walk-up customers are being treated in general has never been more important,” Ken Denman, chief executive of Emotient, told the San Diego Union-Tribune in March 2014. “Given the trends these days in retail, the worst thing that can happen is for someone to walk out of your store and give a tweet about a bad experience.”
Emotient’s technology is largely based on research by Javier R. Movellan and Marian Bartlett, researchers at UC San Diego.
“We’re delighted at the success of Emotient getting purchased by Apple,” said Paul Roben, associate vice chancellor for innovation at the university. “This is how the system should work. We brought all of our innovation resources together, spun out a company, which developed a technology.”
UC San Diego has a sizable group of A.I. specialists, and it is planning to add more as part of its long-term growth in robotics. The campus has tried to jump into the big time by hiring star faculty, but has so far failed to seal some key deals.
The university made an unsuccessful attempt to recruit Gil Pratt, a former Defense Advance Research Projects Agency roboticist who recently joined Toyota. A short time later, Toyota announced that Pratt would oversee a $1-billion A.I. research program in Palo Alto, near Stanford University.
Despite a few setbacks, the university viewed the sale as a confirmation of what it was doing.
“The purchase of Emotient also highlights how relevant artificial-intelligence and machine-vision expertise is right now. And these are areas where UC San Diego is actively investing yet more research resources,” said Albert Pisano, dean of the Jacobs School of Engineering at UC San Diego.
The business use of the technology could be Web searches that use a camera to track facial expressions during a purchase, said Heather Honea, a marketing professor at San Diego State University.
She noted technology already used by major Internet companies customizes advertisements to users based on past purchases, image recognition and their online behavior. If facial recognition was used it would be a richer set of data — even if turning on your webcam for an iTunes purchase sounds odd now.
Honea said history shows that people get very concerned about a new technology at first, see how it works and then change their minds.
“I suspect at first everyone would say, ‘There’s no way. I’m not doing that,’” she said. “Then, the first time someone finds exactly the pair of shoes they were looking for with minimal effort, then they’ll click on ‘OK, I’ll allow it’ the next time.”
Molnar and Robbins write for the San Diego Union-Tribune.