While you can already unlock your iPhone using your mug, it looks like Apple is getting more serious about facial recognition and AI.
Apple has quietly acquired Xnor.ai (opens in new tab), an edge-based AI startup, for $200 million. The move, first reported by GeekWire (opens in new tab), potentially positions the Cupertino company to take on Google in big tech's endeavors to move artificial learning away from the cloud.
Xnor.ai is a three-year-old startup that specializes in machine learning and facial recognition programs that run on consumer devices instead of the cloud. Its AI was recently used in Wyze's home security cameras, but has been removed following Xnor.ai's acquisition by Apple.
An increasing number of security cameras are gaining facial recognition; however, having the feature built into the camera, rather than relying on the cloud, is considered more secure, as the latter is more vulnerable to personal data exposure, as cloud-based AI requires transmitting private correspondence like Siri voice recordings.
The purchase was no frivolous move on Apple's part—the company has a developed a reputation for scooping innovative start-ups with sexy technology, like AR-glasses maker Akonia Holographics (opens in new tab). Xnor.ai joined Akonia's ranks as a valuable target when it made Forbes's list of most prom-sing AI companies this past September.
Xnor.ai's technology should give Apple edge computing capabilities that align with the company’s recent assertions about protecting user data — a controversy that convinced Apple to attend CES after a 28-year hiatus.
Edge-based AI: Apple vs. Google
Google is some ways ahead of Apple in its machine learning chips. Google Assistant uses chips on its Pixel 4 phones and Nest Mini smart speaker, rather sending voice recordings to the cloud and receiving improvements in response.
And while the iPhone's A12 Bionic chip supports machine learning to a degree, the Xnor.ai purchase suggests that Apple's future smartphones will be more capable of handling on-device AI.
In a report by Deloitte, analysts predicted that this year more than 750 million edge AI chips and computers will be sold, mostly in consumer electronics like smartphones. By 2024, that number is expected to rise to 1.5 billion.
Meanwhile Google has been developing Coral, its own edge-based AI initiative. Coral exited beta in October and looks to provide prototyping support and modules that can be used beyond consumer electronics.
But Apple is determined to command the AI field, leading all tech companies for AI-related acquisitions in 2019. CBInsights found Apple's AI spree has overtaken Google's growth in the same category.
So while there's reason to suspect Xnor’s AI-enabled image recognition tools will seep into future iPhones and Mac webcams, it's likely not the final play in Apple's bid to out-AI Google.