1 Facebook x.com Reddit
Last updated 16 minutes ago
The industry says Apple is lagging and asks when the iPhone maker will ever release AI — but the industry has always underestimated Apple.
Today, by the end of 2024, Apple Intelligence still hasn't been deployed. What frame of reference allows us to postulate how important it is for “early adopters” to get to something before Apple?
It's all happened before
Analysts hired by Apple's competitors and journalists looking for both sides of every issue have been thrilled by the prospect of reporting that Apple is lagging and unrepresented in what is presumably the latest, greatest technology.
That jubilation is also shared by Samsung bigwigs and the bearded sheriffs praying for Google’s Pixel to become relevant. Both manufacturers, among other Android licensees, have unveiled multiple examples of generative AI on mobile devices this year, before Apple has even announced any generative AI ambitions.
Of course, as I wrote back in 2017, when Apple is two years behind you, clean up your act. In the article, I laid out how Apple had seemed to fall behind in some way every year for the past decade.
In 2007, it was MP3 players — unlike the iPod, Microsoft’s new Zune could share data wirelessly! That same year, Apple debuted the iPhone, far exceeding market expectations. In 2008, Apple seemed to be out of the mobile app game — just before it took over and led the category with the App Store, which has since remained the most important and commercially significant mobile platform and app marketplace.
In 2009, Apple was never expected to catch up to Blackberry in enterprise instant messaging. The reversal of fortune was so dramatic that a movie was made about it. It was Blackberry's epitaph, all but forgotten now that Apple has utterly dominated IM in the United States, where Blackberry was destined to be the “Crackberry” that Enterprise couldn't get enough of.
In 2010, Apple was lagging behind in Push Notification Services (PNS) behind webOS, which is also gone. Today, Apple clearly has the biggest PNS. In 2011, Google was pushing NFC and Google Wallet, which was “first” only in bragging rights, while Apple belatedly launched its own initiatives with Bluetooth 4, Continuity, and Apple Pay, a set of technologies far more advanced than touch but also providing working NFC taps for payment that people actually used. No one else even comes close.
In 2012, it was Samsung's plus-size phones, eclipsed by the fusion-powered iPhone 6, which was not only big but also fast at 64-bit; in 2013, it was Android's “Modern UI” Halo, arriving just before iOS 7; in 2014, it was Samsung's new and “first” wearables, eventually crushed by the Apple Watch; and in 2015, it was “detachable” PC tablets like the Microsoft Surface Pro, crushed by the iPad Pro. I've kept reading if you want to read the entire article covering every year of the last decade.
Surely, there's no way Apple could catch up to Microsoft's sweet (package?) deal with OpenAI that began back in 2019 today, right? Bueller?
This year, Microsoft beat Apple to the punch with Copilot PC, offering generative AI as tightly integrated into Windows PCs as Bob and Clippy. Note that both of those are currently only in museums, literally right next to the Zune. They could add Windows Phone and Surface.
The wonderful Seattle Museum of History and Industry
In mobile phones, both Samsung and Google, among other Android makers, have rushed to market phones with generative AI before Apple. Who can guess where this will lead?
The Year of Generative AI Without Apple?
In early 2023, Microsoft committed another $10 billion to OpenAI, a year that has been widely dubbed in retrospect as “the year of generative AI.”
In the spring of that year, OpenAI released GPT-4. Content generation with this tool (and its ilk) quickly became both widespread and alarming, as many students replaced their writing degrees with simpler suggestion engineering; as Hollywood executives threatened to use AI to replace writers and even actors with AI-generated words and faces; and as social media exchanged active discussions of reality with passive likes of plausible fictions.
Where was Apple in 2023? Experts noted that it was definitely not among the leaders in generative AI. Apple’s generative AI apparently didn’t exist, except for the extensive use of machine learning (AI), which Apple has been incorporating into its iOS and macOS for years for a variety of purposes, from object recognition to computational photography, body tracking, and fitness apps.
Of course, Apple released its A11 Bionic chip in 2017 with a new Neural Engine. This neural network processor equipped both the iPhone 8 and Apple’s next-generation iPhone X with the hardware ability to perform accelerated machine learning on-device. It was used most dramatically to power Face ID and Animoji, but Apple’s use of machine learning has been broader and deeper.
In addition to the Neural Engine, Apple Silicon also ran machine learning in its GPU and image signal processor, the engine that handles complex computational photography. That same year, Apple also released a new developer API called Core ML that allowed third parties to run a variety of AI/ML tasks using any available ML-accelerated hardware features.
Apple wasn’t doing the type of generative AI that tools like ChatGPT and DALL-E made famous in 2023, but it was deeply involved in delivering AI-powered features that were far less controversial.
Apple has been heavily involved in AI/ML development since 2017! It even laid the groundwork for Edge AI—running AI on-device rather than offloading all tasks to the cloud—with increasingly sophisticated and powerful versions of the Neural Engine. Microsoft and Qualcomm are belatedly racing to catch up to Apple Silicon’s built-in machine learning power.
Instead of depicting the Pope in a poofy white Balenciaga he never wore, or cranking out a writing project so you didn’t have to think about writing or form any new important connections in your brain, Apple’s ML used the same underlying technologies to deliver core features that were valuable to users and developers.
Even before opening up CoreML to developers, Apple had been using ML internally for years before, before performing facial recognition in your photos, as well as tracking faces and facial landmarks in Face ID and Camera. Apple also used ML in its graphics, making Siri’s voices “learn” to sound more natural, and to feed relevant information into the new Siri WatchFace, which debuted in watchOS 4, along with a host of other tasks, from text prediction to fitness tracking.
Additionally, Apple's senior vice president of software engineering Craig Federighi detailed how Apple will use machine learning to address privacy, noting that “Safari uses machine learning to identify trackers, separate cross-site scripting data, and store it, so now your privacy and browsing history are yours.”
Apple has been talking about privacy and machine intelligence for years
Apple has been a bold leader in useful AI/ML applications in many ways, simply because iOS has been the undisputed leader in premium mobile phones. So why did Apple come under fire last year from a crowd of pundits who were actively releasing and deploying ML hardware and software six years ago?
It wasn't because generative AI couldn't be implemented on the Mac or iPhone. Despite Microsoft's billions, or perhaps “because of” them, OpenAI released its own iOS client app last summer, simply because iOS was the undisputed leader in premium mobile phones.
OpenAI's press release for the iOS app at the time stated, “P.S. Android users, you're next! ChatGPT is coming to your devices soon.”
Where was Android?
Why wasn't Android criticized for lagging Apple's platform in leading generative AI app adoption throughout the Year of AI? Perhaps “generative AI” was greatly inflated as an imaginary problem for Apple simply by the very nature of the fact that iOS was the undisputed leader among premium mobile phones.
By the end of the Year of AI, there was no difference in the availability of generative AI tools across Apple's platforms. But this year, the Samsung Galaxy S24 and Google Pixel 8 both narrowly became the first to introduce integrated mobile generative AI features.
Were users excited? It doesn't seem to have gained much traction in the market. Their rollout of generative AI has brought with it not only compelling upgrade features, but also some useful features that could just as easily have been included in an app.
Features like Samsung’s Live Translation are impressive and useful, but they’re not a reason to buy a new Samsung phone. “Circle to Search” lets you identify an item and ask Google what it is. That’s not even generative AI, it’s an app feature that uses object recognition technology that Apple debuted back at WWDC 2017.
AI photo editors like Google’s “Magic Editor” and Samsung’s “Generative Edit” erase unwanted objects in photos and create believable backgrounds. Apps have been doing this for a while on iOS. Neither of these things are significant OS-level AI/ML integrations that provide any reason to upgrade or move to a new phone, right?
And given that the vast majority of Androids are carrier-friendly, good-enough devices (in Samsung’s nomenclature), Android makers have no reason to go beyond AI, generative or otherwise, because their largely anemic processors can’t do anything transformative anyway. They’re focused on affordability.
High-end phones are a very thin slice of the Android pie, typically serving as showpieces with bragging rights for being “first” at something, from the curved Edge screen to the Folds and Flips, which only exist because Samsung makes displays and needs to show off what it can do. They don’t drive sales. They just drive hype.
The downside of being first is getting arrows in the back. Google's generative Gemini has raised more eyebrows than excitement as it becomes clear that the company is desperately trying to deal with hallucinations by beta-hitting the fences with a high virtue-to-noise ratio.
If anything, it looks like — once again — “being first” is really just preparing the market for Apple's introduction.
Apple Intelligence Is More Than Just a Chatbot
Remember how Meta has been pushing Oculus VR non-stop for a decade, paving the way for Apple's Vision Pro to really impress the elite with free money to fund the development of the future of television?
Where Was Microsoft?
You can give mass-market phone makers a break for not delivering super-useful new technology while focusing on serving the low-end of the market. But what about Microsoft, which was lauded as a genius corporation for pouring so many billions into OpenAI to embrace and expand on the latest technology so that it became just another facet of its desktop monopoly diamond?
I'm not here to rub Windows Phone in Microsoft's face, but while sticking with desktop Windows, it's useful to note that the once-vaunted monopoly that Microsoft once ruled with an iron fist has been vastly outsized by iOS. Redmond's ability to “adopt, expand, and destroy” has waned, as is clear from its utter failure to launch Windows Phone using its desktop. Oops, I did that.
Back to AI/ML, as I predicted, it turns out that Microsoft's Copilot PC wasn't much of the glorious golden nugget praised by loyalists, but was instead just crap with featureless corn kernel pulp.
This week, Salesforce CEO Marc Benioff complained in a tweet: “When you look at how Copilot has been delivered to customers, it's disappointing. It just doesn't work and doesn't provide any level of accuracy.”
Benioff continued, “Gartner says it’s spilling data everywhere and customers are left to clean up the mess. To add insult to injury, customers are then told to create their own custom LLMs. I have yet to find anyone who has had a transformative experience with Microsoft Copilot or is committed to training and retraining custom LLMs. Copilot is more like Clippy 2.0.”
Note that Benioff is not one to be easily timid about AI. In the podcast, he exclaimed, “I’ve never been so excited,” when asked about the future of AI. When asked specifically about Microsoft’s Copilot, Benioff, unsurprisingly, touted SalesForce’s AgentForce as a tool for creating and customizing AI agents.
“I think Microsoft has done a huge disservice not only to our entire industry, but to all the AI research that's been done… And customers are so confused based on this Microsoft story… I don't think Copilot is going to exist. I don't think customers are going to use it, and I think we're going to see a transformation of the enterprise with agents, and Agentforce is going to be the number one vendor.”
That's hardly an endorsement of Microsoft's efforts, and certainly not from a competitor, so think about it that way. But then again, the market has spoken, and it's not exactly buzzing about Copilot PCs.
Microsoft's Copilot computers are like the Google Pixel: an expensive attempt to deliver a high-quality, premium product that matches Apple's quality from a company that's really only known for achieving mass-market sales with mediocre, cheap hardware running on widely licensed, generic software that makes a lot of compromises to work everywhere.
Apple, on the other hand, is pretty well known for delivering deeply integrated hardware and software that actually excels at the tasks it was designed for.
I think it's too late to make any bets on who will drive mass adoption of AI/ML tools.
Follow AppleInsider on Google News
