It's not often you hear about layoffs at Apple.
So it came as a surprise Wednesday when CNBC learned that Apple was removing 200 employees from its self-driving car unit. Apple confirmed the staffing change, but reading between the lines of a spokesperson's statement, it sounds like the move is the latest in the company's broader goal to improve its artificial intelligence and machine learning capabilities as it faces increased competition from rivals Google and Amazon.
"As the team focuses their work on several key areas for 2019, some groups are being moved to projects in other parts of the company, where they will support machine learning and other initiatives, across all of Apple," the company spokesperson said in a statement to CNBC Wednesday.
Self-driving car technology may still be an important initiative at Apple. But reading between the lines, it looks like it's taking a back seat a Apple beefs up its general AI staff.
"I think they're making the decision that, at least in the near term, it's better to have these people doing AI in other projects," said Gene Munster, a venture capitalist and analyst at Loup Ventures.
Apple's self-driving car project, called Titan internally, started out with the desire to create an Apple-branded electric car, The Wall Street Journal reported in 2015. But over the last few years, Apple has scaled back its ambition and lost leaders and other employees in the process. Today, the unit mostly explores the underlying technology that makes self-driving cars possible. CEO Tim Cook has repeatedly called self-driving "the mother of all AI projects."
Since Apple started its self-driving division, the consumer AI space has exploded through the rise of digital assistants like Amazon Alexa and Google Assistant and devices like Amazon's Echo. Apple's own digital Assistant, Siri, had a head start when it launched on the iPhone 4S back in 2011, but has not kept up with the competition.
To address the shortfall, Applehired Google's head of AI John Giannandrea away from the search giant in April. Within a few months, Apple had reorganized its entire AI and machine learning teams under Giannandrea, the company announced to TechCrunch. And just last month, Giannandrea was promoted to Apple's executive team as vice president of machine learning and artificial intelligence strategy.
Self-driving may still be an important piece to Apple's AI research. The company said in its statement Wednesday: "We continue to believe there is a huge opportunity with autonomous systems, that Apple has unique capabilities to contribute, and that this is the most ambitious machine learning project ever."
But as far as products go, competition in self-driving and electric vehicles has grown dramatically in the last four years.
Waymo, Alphabet's self-driving car company, recently opened up its self-driving car service to the public in Phoenix, Ariz., and is widely considered to be the leaders in self driving. Legacy car companies like GM working on self-driving technology. And it's not just Tesla making electric cars. Porsche, Audi, Mercedes and other legacy car companies have all announced electric vehicles. It's hard to imagine how Apple would stand out.
"The sense that I had is they're not as far along as I had hoped," Munster said of Apple's decision to remove the 200 employees out of its car division. "But they still have initiative there."
Giannandrea's rapid rise at Apple is the biggest signal yet that Apple intends to invest a lot of time, money and talent in improving AI. Plus, according to leaked comments from Cook at a recent company all-hands meeting reported by Bloomberg, Apple plans to continue hiring in AI "at a strong pace" even as it slows down hiring in other divisions.
In short, we're seeing Apple eliminate jobs in self-driving and increase the number of people working more broadly on AI.
It may already be paying off. Late last year, a study from Munster's company, Loup Ventures, showed Siri vastly improved its ability to correctly answer a series of 800 questions. The Loup study said Siri answered 74.6 percent of the questions correctly, up 22 percentage points from just nine months earlier. By comparison, Google Assistant answered 87.9 percent of the questions correctly. Alexa got 72.5 percent of the questions right.
It's not just about getting questions right, though. The messier problem for Apple is training its AI while convincing users that it's keeping their data secure.
Google trains its AI systems in part using the massive amounts of public data available on YouTube and the Google search engine. (It's also started using a program that strives to protect users' data.)
But Apple has taken a hard stance against unfettered data collection, and promotes its concern over user privacy as a reason to buy its products. In a speech in Brussels last year, Cook called the privacy practices of companies like Google "surveillance," for example. It also put up a giant ad about its privacy stance in in Las Vegas during CES earlier this month.
So Apple will have to continue to improve its AI while sticking to its goal of keeping people's personal information private.
"I think it's very clear Apple is a believer in AI and most of the products will be very subtle about how AI is used," Munster said.