“Google has been regarded as the leading force for the development of artificial intelligence (AI) technology thanks to its AlphaGo AI computer program,” Ashley Huang and Luke Lin report for Digitimes. “While Google still holds the advantage in the cloud-AI sector, Apple appears to have outraced Google in the development of on-device AI products with the launch of its iPhone X and iPhone 8 devices.”
“The built-in dual-core neural engine chip incorporated within the A11 processors, which power the iPhone X and iPhone 8 devices, integrates functionality such as machine learning, inference model and related algorithm to perform as an application-specific integrated circuit (ASIC) chip for image recognition and achieve hardware acceleration,” Huang and Lin report. “The Face ID application allows iPhone X users to actually experience on-device AI applications. Through the use of the TrueDepth camera, users can build up a depth map of face and the neural engine can identify and confirm the face and unlock the iPhone.”
Read more, including how Google is also hopelessly behind not only in processor speeds and capabilities, but in augmented reality, too, in the full article here.
MacDailyNews Take: On-device AI means user privacy – about which Google could care less and, in fact, hates, as Android’s very existence is to serve as a conduit for vacuuming user data to Google servers for ad delivery and only God knows what.
There is no such thing as a free lunch.
If it’s not an iPhone, you’re not the user; you’re the product.
SEE ALSO:
iPhone 8’s Apple A11 Bionic chip so destroys Android phones that Geekbench creator can’t even believe it – September 30, 2017
Apple’s A11 Bionic chip is by far the highest-performing system on the market; totally destroys Android phones – September 19, 2017
Apple’s A11 Bionic chip in iPhone 8, 8 Plus, and iPhone X leaves Android phones choking in the dust – September 18, 2017
The inside story of Apple’s amazing A11 Bionic chip – September 18, 2017
Apple’s A11 Bionic obliterates top chips from Qualcomm, Samsung and Huawei – September 18, 2017
Apple accelerates mobile processor dominance with A11 Bionic; benchmarks faster than 13-inch MacBook Pro – September 15, 2017
Apple’s A11 Bionic chip in iPhone X and iPhone 8/Plus on par with 2017 MacBook Pro – September 14, 2017
i have to disagree with mdn. i do not believe google could care less. i believe any selfless consideration for user privacy they might once have had was exhausted long ago. there’s little, if any, capacity left at google for honest, undisguised self-examination.
I think MDN/Solo mean, “Google couldn’t care less.”
yeah, i know. i was just being a snarky, contemptible little prick 😛
Google has a huge advantage using consumers’ data and they should be happy. Wall Street praises Google constantly for having everyone’s information at their fingertips. All the companies that are Wall Street’s #1 favorites list are harvesting user data and no one gives a fig about it. As long as there is no data breach, Google, Facebook and Amazon will continue to stay on top. Apple is basically fighting a battle with one hand tied behind their back or maybe both hands tied behind their back. Apple’s data privacy policy is going to continue cost Apple in terms of company value. Smart assistants must have as much information as possible if they’re going to help you. It’s having all that information that makes them smart. Siri is stupid and it shows.
Any consumer using Android OS is happily giving their information away for ‘free’ services so that’s quite a large number of consumers. We’re talking about 90% of the smartphone population. Apple serving only 10% of smartphone users with privacy isn’t going to change anything in their favor. Apple is never praised for keeping user privacy because hardly anyone sees value in it. That’s just Apple’s self-imposed cross to bear.
…more like albatross.
Exactly!
Apple has no choice if they want be relevant down the line..
They either have to compromise their mantra or find a way around it to amass huge amounts of data , develop the proper Analytics and Algorithms…. and try to CATCH UP BIG TIME!
The present difference in AI is not acceptable nor even an option! Imo
It’s a path to oblivion….
Google Smartphones (ala Android) are the Swiss cheese of Smartphones, “full of holes”, where Google AI actually works between the phone and Google AI servers, with only one aim …. to know ever more about you, so that this continually refined profile can allow advertisers to target YOU!!
Apple Smartphone AI resides wholly on the IPhone, is never provided to third parties because your privacy is of paramount importance to Apple Inc.
I know which experience I value and it’s not Google’s!!!
The built-in dual-core neural engine chip incorporated within the A11 processors, which power the iPhone X and iPhone 8 devices, integrates functionality such as machine learning, inference model and related algorithm to perform as an application-specific integrated circuit (ASIC) chip for image recognition and achieve hardware acceleration
That’s great! Now let’s see it perform something useful other than just face recognition. Or am I missing/forgetting some other function for the neural engine at this point? Does it provide any functionality to Siri? Bitching about Siri vs everyone else’s “A.I.” has been a core subject of the week.
Face recognition is natural for humans, and is a cornerstone of all kinship and extended social interactions across all cultures and many species. Machine learning mimics the processes that babies use to join the human race. Early in life we learn our place, who to trust and who not to trust, through this intricate form of pattern recognition. It’s far from a simple database lookup and definitely not a glorified google search. It is a form of artificial intelligence, just not the conscious rational thinking that you and most other people assume must constitute general intelligence, provable through the Turing test. In Daniel Kahneman’s language, it’s system 1 thinking, not system 2.
This is a big step. Even if no more steps happened, it would still be as marvelous as the joy of recognition flowing between you and your first-born as you learnt to know each other, a basic feeling that one is not alone, that there exists another, a Friend.
No actually. Newborns typically can’t see much of anything. They cue on sounds, smells, touch, warmth, food (therefore taste). Once their eyes are able to focus and make out detail, they do indeed then respond to faces. Their response is both functional and emotional. The basic need of course is security. But what’s wanted is everything to do with love.
Back on topic: Expert system face recognition has no needs of any kind. It has no intelligence whatsoever. It collects data, either on its own or via user input. It is asked questions. It looks up the answer within the data available. It provides and answer. There’s no ‘intelligence’ of any kind. It’s a machine. It’s a tool. Put a moving plastic baby face over it, add some baby burble sounds, and it’s still just a machine, just a tool.
But thank you for your reference to Daniel Kahneman.
BTW: I still think you’re greatness. I’ll also express that I’m having a particularly trying day due to the severest of Other-People’s-Problems being inflicted upon me over the past 48 hours. Therefore, I’m in vigilance and discipline mode. Therefore, never any offense intended toward you.
I was thinking the technology could be used for ‘reading lips’.
In public you could use silent commands to your phone or text by just holding the phone near your face. And of course … spy apps reading others conversations.. Stretch I know, but the facial recognition AI is going to be used for more than unlocking phones.
I was thinking the technology could be used for ‘reading lips’.
I’d imagine so. However, as with language recognition, there are many variations in human mouths and how we use them. There are also sounds made entirely internally, requiring no lip motion that would be missed by such a system. Beyond those two limitations that I know of, I defer to lip reading experts.
Clearly, 3D facial recognition (versus the worthless 2D facial recognition Samsung foists on victims) is a step forward in machine learning and comprehension of information available to them from the real world around them.
The limitation: The more data a machine collects or is fed, the more difficult it becomes to traverse/sort through that data for the answer to a query. How that difficulty is manifested, we’ll be learning as the elusive goal of real artificial intelligence is pursued.
Making AI better because of the collection of your “personalized” data is a big RED HERRING.
It doesn’t make it better. Un-personalised data is just as good.
Name one function that Google AI does better today because it knows something personalized about you. Does it for example tell you that it’s time for something only you do everyday? Is it capable of writing a letter to your girlfriend explaining how you being found in bed with someone else was merely a simple misunderstanding.
Can someone list some “obvious” AI differences between GAI and AAI that might impress in some small way. And NO, SIRI saying, “Heres what I found on the web, take a look”, verses GAI saying, “The answer is 47”, is not what I would consider a difference in “Inteligence”.
“MacDailyNews Take: On-device AI means user privacy – about which Google could care less”
Nuff said…
Ahh so Google DOES care, since they COULD care less. I thought MDN meant the opposite.
If AI across the board were anything other than at best a misunderstanding of computing or at worst a total farce, I’d actually give a poop. :/
To compare this topic with Apple Pay, once there is a security problem all others are affected. But not those who use Apple Pay. Apple will obviously come out on top since they care about security & privacy of customers. They are well set up for the future.