Apple takes artificial intelligence (AI) and machine learning very seriously, and is already as deeply embedded in these areas as Google, Microsoft, Facebook and Amazon.
At least, that’s the big takeaway from a Backchannel profile by Steven Levy, who was allowed on the Apple campus, given exclusive access to key Apple staff members (mostly the same public faces at events) and given a thorough rundown of Apple’s efforts, challenges and ambitions.
Yes, by Apple. The world’s most secretive company.
That it should invite a reporter to insist to him how good it is at a thing rather brings to mind the Margaret Thatcher quote about how being powerful is like being a lady (“If you have to tell people you are, you aren’t”).
Or, perhaps Apple higher-ups realized that when no one knows what you’re doing, after a while they stop assuming you’re secretly doing the most amazing things ever—like your competitors are. And there becomes a need to crack open the door just a little.
“It is unusual for Apple to tell its story, other than in carefully choreographed events,” Ezra Gottheil, principal analyst with Technology Business Research, agrees.
“I think this is a deliberate attempt to reassure customers that Apple isn’t losing it, while Google and Microsoft are showing off their chops,” Gottheil told eWEEK. “Amazon, by the way, also has some real AI capabilities, but in many ways, it’s more private than Apple.”
Jan Dawson, chief analyst at Jackdaw Research, says Apple’s broad efforts aren’t obvious in its finished products, which has led to some criticism from the media and even competitors.
“Competitors like Google and Microsoft have very much talked up their capabilities in this area at their respective events, and so it’s natural for Apple to want to showcase its own advances here,” Dawson told eWEEK. “These skills were discussed at [Apple’s Worldwide Developers Conference] in June, but the press coverage didn’t pick up on it much and the narrative didn’t really change. So I think Apple just felt that, short of inviting someone like Steven Levy in and talking them through it in depth, it’s tough to address that perception and the lack of understanding of what Apple is doing here.”
So what did Apple lay out for Levy?
Siri is smarter and does more than one might think. (“You can ask [Siri] who it thinks is going to win the game and it will come back with an answer. I didn’t even know we were doing that!” Eddy Cue, Apple’s senior vice president for internet software and services, told Levy.)
Apple has more AI and machine learning talent than one might imagine. (“We hire people who are very smart in fundamental domains of mathematics, statistics, programming languages, cryptography,” Craig Federighi, Apple’s senior vice president of software engineering, told Levy. “It turns out a lot of these kinds of core talents translate beautifully to machine learning.”)
Apple has created its own uphill battle by prioritizing user privacy more than its rivals do, though it expects that eventually everyone will follow. (“It’s great that we would be known as uniquely respecting users’ privacy,” said Federighi. “But for the sake of users everywhere, we’d like to show the way for the rest of the industry to get on board here.”)
And finally, much of this will be on display in iOS 10, which will be introduced this fall. (Siri has been the focus of four areas of research and upgrades: speech recognition, natural language understanding, execution and response. That last one will come in the form of a more human-sounding Siri—a result of deep learning—which it’s thought will make users trust and use the software more.)
Apple Deeply Serious About AI and Machine Learning, Says Report
To Gottheil’s point, Levy notes that the “AI cognoscenti” don’t know what Apple is doing behind its closed doors, and quotes Jerry Kaplan, who teaches a course on the history of AI at Stanford University. Kaplan offers, “It’s not part of the community. … Apple is the NSA [U.S. National Security Agency] of AI.”
Part of that is due to the fact that Apple has discouraged (or maybe forbidden) its scientists from publishing academic papers on their breakthroughs, instead hiring folks who will be satisfied by putting their findings to work.
However, Apple scientists have made such a significant contribution to the field of differential privacy, which will be put to work in iOS 10, that Levy reports Apple is “authorizing the scientists working on the implementation to publish a paper on their work.”
Dawson notes, “Privacy protections make it harder to digest big data sets on a massive scale centrally, while [Apple’s] secrecy requirements mean that people that consider themselves leaders in the field are going to settle for a much lower profile at Apple.”
As competition amongst the top tech companies heightens, significant-but-not-obvious updates and behind-closed-door advancements may no longer be the best strategy.
“Apple’s products, especially its iOS products, continue to evolve, but the fact that the PC, the smartphone and the tablet are all mature means that the improvements are less noticeable than when they were new,” said Gottheil.
“My sense is that the Apple AI, as implemented, is as strong as the Google and Microsoft AIs and far less intrusive than either of the two,” he said.