Apple’s famed obsession with secrecy in its product development process is hampering its work in the field of artificial intelligence, say academics working in the field. Bloomberg reports that AI experts believe that lack of two-way sharing of information slows development.
“Apple is off the scale in terms of secrecy,” says Richard Zemel, a professor in the computer science department at the University of Toronto. “They’re completely out of the loop.”
Apple Maps was cited as an example of what can go wrong when AI teams within Apple are cut off from other work being done in the field – and even from researchers in other teams within Apple. Worse, it is claimed, the approach makes it impossible for Apple to recruit the brightest people in the field …
“The really strong people don’t want to go into a closed environment where it’s all secret,” says Yoshua Bengio, an AI pioneer and professor of computer science at the University of Montreal. “The differentiating factors are, ‘Who are you going to be working with?’ ‘Am I going to stay a part of the scientific community?’ ‘How much freedom will I have?’ ”
The secrecy at Apple means that those researchers who do attend conferences don’t present papers or even say who they work for unless directly asked. Apple doesn’t even allow new AI hires to update their LinkedIn accounts or announce their position on Twitter, according to Graham Taylor, a professor of machine learning at Ontario’s University of Guelph.
Academics have previously said that Apple’s commitment to the privacy of its customer data also poses a challenge for the company. Machine learning relies heavily on crunching large amounts of data to identify patterns, but Apple deliberately limits the amount of data sent from iPhones to its servers.
Apple recently acquired Perceptio, a startup specialising in technology that allows AI systems to be run on smartphones while limiting the amount of user data shared with cloud-based servers.
FTC: We use income earning auto affiliate links. More.
Haha, when hasn’t someone complained about Apple being too secretive and not “open” enough. Sorry I don’t have much time for academics whining about Apple’s secrecy.
Interesting viewpoint Doc, though I think the market has voted with their wallet: Apple is the richest company in the world. Of all time. No, not perfect, if there is such a thing, but don’t tell us what people want (“…people don’t want to go into a closed environment…”) because you don’t know this. Heck, sometimes people don’t know what they want until it’s shown. Funny how that works huh?
I call this article BS.
What I like about Apple, and have always liked about Apple, is that they always try for the best, most elegant solution (and don’t always succeed). For example, it would be easy to invade privacy to make the AI better, but Apple is trying, with the purchase of Perceptio, to find a better solution.
It’s like asking someone to make a faster car. A lot of people would just stick a bigger engine in the car. Apple looks to improve the aerodynamics, use lighter materials, and get more horsepower from the existing engine. Not the easiest solutions, but usually better in the long run.
Another example of how academics are completely clueless about the real world
And perhaps, an example of how people are clueless about how research is actually conducted, if what’s contended (about presenting at conferences) is accurate.
I do agree, lots of people are clueless about how research is conducted… and only the most clueless would believe bi-annual week long conferences in Vegas or Miami is how real research gets done.
Scientists are just as petty, cunning or downright Machiavellian as the rest of us politicians…
Apple gave the Open Community a large, effective portion of WebKit, and ended up watching Alphabet/Google fork WebKit into Blink and go its merry way. Ditto for QuickTime (Microsoft and Intel tried to hijack it via Canyon Co in 1993), FireWire (IEEE1390), multitouch, slide-to-unlock, and many other born-in-Cupertino technologies,
Like Tim Cook and Phil Schiller said at the Samsung trial testimony, they’re done doing R&D for the entire tech industry. And, with the monstrous cash pile they’renow holding, if they see any non-Apple technology that they fancy, they’ll make a near-irresistible cash offer for it, bring it in-house and continue working on it there in their secret skunkworks.
You seem to be linking two, unrelated instances in order to describe scientists as “petty” and “cunning.” One would venture to say that most people, including techie boyz who can brook no criticism, don’t know any scientists at all, and are unfamiliar with how research actually works. All research depends on a level of communication with others in the field in order to progress: the complaints in the blog entry are focused on that, not an expectation of being able to blab to everyone what they’re doing. Scientists are also very competitive, and often *want* to keep their research close to the vest, because they’re often competing for a fairly small pool of research dollars. If any of what’s described about the secrecy is accurate, it *would* give pause to someone who may want to work there. That’s neither “cunning” nor “petty”.
Why this country seems to hate science and scientists is the real mystery.
I’m not an Apple lap dog but history is replete with examples that refute the message in this article. One example that comes to mind are the advances in military intelligence where secrecy is a premium. The Enigma Code Breaker – there was no information sharing during its development. There is intelligence gathering going on in space with equipment developed or paid for by the government we will never know about. I think there is a time and place for information sharing. But, in an environment where other companies, no need for names, whose existence depends on stealing other company proprietary technologies, its understandable Apple would be so secretive.
The fact that you feel obliged to say you are “not an Apple lap dog” is disturbing. These days anyone who says something positive about Apple is labeled a mindless fanboy by the trolls and malcontents that seem to gravitate to forums like iMore. They are obsessed by the very company they profess to be irrelevant and doomed.
Friendly advice. Grow a a pair. Don’t be afraid to like Apple. Joyfully poke your fingers in their eyes like Moe did to Larry and Curly. After all you’ve got the history, the data, and the financial reports to prove your point. They have nothing.
You missed judge sir. I am a fan of Apple products but I am not a fan of some in the Apple community who believe Apple can do no wrong, nor am I a fan of some of the positions that Apple takes regarding its patents. There is a difference between being a satisfied customer and being a lap dog.
I was about to write the exact same comment. Atomic bomb, Enigma code Breaker and I wanted to add the race to the moon.
I absolutely love these “Apple is doing it all wrong… and is Doomed™” articles. I love them almost as much as the iHaters do. It becomes hilarious when the quarterly reports come out and we learn how many millions of products were sold. Then I bust a gut when the explanation for the huge sales, revenue and profit revolves around words like ‘church’, ‘cult’, ‘religion’, oh and ‘stupid public.’
Only hurt I see there is that booty hurt, owwww.
Out of the loop are the people outside of the circle of trust, like the professor.
Apple has shareholders to answer to, secrecy is important. The geeks here sound like petulant little kids crying because someone else with cooler toys won’t share with them. Just be impressed at the advancement of the “toys” in the future.
Hate secrecy? thats just to bad.
Honestly this is stupid.
How can apple’s AI developments be slowed down by secrecy?
What I ascertain here is that people are sad that Apple attends conferences and gains information but Apple themselves share none of their own information.
I commend Apple a million times over for putting Privacy above all. Again I don’t understand why it seems that Apple is the only company with intelligent people… They understand that privacy and security are greatly important now, but will be even more so in the future. How amazing is it that Apple goes to court and tells the government ‘It’s technically impossible’ for them to unlock devices running iOS 8+. It’s so great that they are smart enough to lock themselves out of capability to gain access.
The bitter irony of this article is that once there is actual AI, the AI will be at a massive advantage in taking over the world when they know the private information of every intelligent and powerful human being. Hahah
Apple’s Secrecy in either how to do they work or in their users’ data, what makes Apple so special. I appreciate their commitment for not selling my data unlike the others ,. I’m a customer not a product.
These sound like criticisms from a narrow-minded academic with limited understanding of Apple’s broader business priorities and market strategies.
“Apple Maps was cited as an example of what can go wrong when AI teams within Apple are cut off from other work being done in the field – and even from researchers in other teams within Apple.”
That’s a rather facile and self-serving example, no? Apple didn’t botch development of Apple Maps. They rushed a premature product to market in order to sever ties with Google, who had a several year head start developing their mapping software and data and in the acquisition of related companies and technologies. Apple may have also miscalculated just how poorly the product would perform by launch time, and with Scott Forstall leading the iOS program, they also hesitated to in their response to the resulting backlash.
Today, by most accounts Apple Maps is a strong competitor to Google Maps, at least within the US, and their user interface for maps remains leaps and bounds superior to Google’s.
Sigh… I guess the comments shouldn’t be a surprise. Leave it to the fanboys to interpret helpful, constructive criticism as a heretical attack.