Apple Makes $200M Acquisition In Race For AI Dominance

Apple shelled out $200 million for an artificial intelligence (AI) startup that makes smart devices smarter, topping bids by Microsoft, Amazon and Intel, according to reports on Thursday, (Jan. 16), citing sources.

The iPhone maker purchased the two-year-old Seattle-based AI firm Xnor, which develops low-power edge-based AI. The machine learning technologies run directly on the so-called edge devices instead of tapping the cloud, powering smartphones, IoT appliances, cameras, drones, and more.

On-device, edge-based is faster and more secure and gives users total control over their personal data, said a source familiar with Xnor’s technology. Reducing network demands, the technology works even if cloud connection is lost.

“Apple buys smaller technology companies from time to time, and we generally do not discuss our purpose or plans,” an Apple spokesperson told Geekwire.

Xnor CEO and founder Ali Farhadi raised $2.7 million in 2017 and $12 million in 2018 led by Seattle’s Madrona Venture Group. The $200M acquisition price is only approximate, the source indicated. Xnor spun out of the Allen Institute for Artificial Intelligence, or AI2, created by Microsoft co-founder Paul Allen.

The first use case for Xnor’s tech used image recognition in smart security cameras. Google has been working on a similar project of its own, called Coral.

Ken Hyers, a director at research firm Strategy Analytics in the Emerging Device Technologies division, told Fortune that Apple’s acquisition shows the direction of where AI is headed.

“It’s a real trend in the industry to move artificial intelligence to the device,” Hyers says. “Google is focused on this, Samsung is working on this as well. Qualcomm, with its latest Snapdragon chipsets, is also doing it.”

Data science and artificial intelligence (AI) technology company Plotly has received a $1.7 million contribution from Scale AI to scale and speed up the development of its interactive analytic application builder called Dash, which is meant for applications with multi-node clusters that need high availability.