Remember when switching from iPhone to Android was simply a hassle? Brace yourself: With today’s launch of iPhone 16 and the proliferation of personalized AI clouds, it’s about to be virtually impossible.
Tech titans are wielding AI as if it were superglue, making it increasingly difficult for users to escape their ecosystems. Apple, Google, Samsung and other major tech companies are placing their bets on AI-as-a-service—not just to enhance your user experience, but to make sure you stay put in their walled gardens forever.
“Anytime you are introducing a new kind of tool, there is variation in the different implementations and approaches that will be big,” Nic Benders, Chief Technical Strategist at New Relic—a monitoring platform that uses AI to analyze huge amounts of web data in real time—told Decrypt. “So, even if all three of those companies bring out a feature that does the same general thing, at first they are going to do it fairly differently. That is going to increase the lock-in for their users.”
The tech behemoths are playing a long game, and your very personal data—the things you buy, food you consume, people you meet, your schedule, and all of the choices you make that define your day-to-day existence—is the ultimate prize.
“To train your AI models, you need data. Lots and lots of data,” said Benders. “Who knows the most about you? Who has your photos? Your emails? Your text messages? Your documents?”
That’s been “a huge advantage for companies like Google, Apple, and Facebook,” he added, and it tends to make the switching costs too high for most people. With the access that AI will increasingly have to the rest of your life, you’re going to want to fight rather than switch.
Users’ own data is ultimately more lucrative than publicly available data, because it provides a level of customization that cannot be achieved by a competitor with a generic training dataset. The platform that owns your most-personal data will own you as a lifetime customer.
Imagine a not so distant future when Apple rolls out its home robot helper—think “Sunny,” the, not coincidentally, Apple TV series. It makes sense that your new adorable robot, which would run on “Apple Intelligence,” is required to connect to your Personal Cloud once it’s unboxed.
Lock in aside, there are bigger potential dangers. The writer Yuval Noah Harari, in his new book, Nexus, cautions that AI is using this personalization to create digital intimacy, which he says is dangerous and will lead us further into a dystopia.
The iPhone 16, arriving in stores today, was built with AI at its core. And the more you rely on it, the harder it will become to leave. Apple Intelligence will be able to follow complicated voice interactions, tag your relatives, create movies, let users interact with its gallery, categorize emails, and of course super-charge Siri with ChatGPT—and that’s just the start.
Why would anyone switch to Android and lose access to the intimate knowledge Apple Intelligence has about you?
For the foreseeable future, for consumers, there is no data portability for AI models. These models are trained specifically with your data and continuously improve as they gather more information. It is impossible to train the same model twice with different foundational technologies, and the closest thing users can do is take all limited data such as photos, contacts, and interoperable documents from one platform, migrate to a second one, and start from scratch training a new AI-based system.
“Every company in every industry is looking at how they can use AI to increase differentiation for their product,” Benders said. However, he notes that features alone won’t be the differentiator: “The real moat in AI will be those ‘personal AI clouds,’ because that is where your data will be—and the capital, because AI is a very expensive business right now.”
This strategy isn’t unique to Apple, of course. Google’s AI-driven features in the Pixel phone series and Samsung’s Galaxy AI are following suit, each creating their own sticky ecosystem. A key difference between last year’s Samsung Galaxy S23 Ultra and the current S24 Ultra is the device’s readiness for AI.
The AI competition in China is even more heated, with consumer tech companies such as Huawei and Xiaomi powering widely diverse Internet of Things (IoT) ecosystems with proprietary AI and personal clouds. Huawei is even investing in AI-powered home robots.
During its recent developer conference, Huawei showcased how its PanGu Foundational AI model is capable of controlling its proprietary robot to not only recognize objects, but also perform daily chores like organizing household items or handling laundry.
This type of robot, built to interact seamlessly with the user’s ecosystem, is part of Huawei’s broader vision for tech in which AI agents don’t just live on your phone, but in your home—performing tasks autonomously. “It can even cook you lunch!” Huawei said in a short video introducing its model.
Huawei has already introduced an AI-controlled robotic dog that leverages Huawei’s cloud data to improve at different tasks such as ground recognition and overall movement. This cloud was described as “the metaverse for robots” in an official Huawei Linkedin post.
Meanwhile, Chinese giant Xiaomi’s IoT ensemble includes smartphones and watches, all the way to cars, home products, and even clothing. Every smart device linked to your Xiaomi account communicates with your phone, giving you granular levels of data, personalization, tasks, and insights that no other company is able to provide without relying on third-party data. (It also has an AI-powered robotic dog, specifically trained to mimic real pets and serve as a good companion.)
But in the West, “Apple is going the farthest to embed AI as just a thing powering new features that are interesting on their own merits,” Nathaniel Whittemore, an AI expert and CEO of AI skilling company Superintelligent, told Decrypt. “In other words, they’re not selling ‘AI’–they’re selling what AI can do for you.”
He added: “It’s definitely possible that AI makes ecosystem lock-in even worse, given how much personalized AI will rely on having access to personal data.”
So what choice do we have if we care about privacy and the freedom to switch to another platform?
“There’s no practical alternative at the moment,” Whittemore said. Users can either give up their data or dumb down their devices—which is what’s happening in Europe, where the European Union is compelling Apple to open up its operating system to third-party developers. If and when it rolls out Apple Intelligence there—it’ll be an iOS upgrade to iPhone 16 and iPhone 15 Pro in the U.S., expected in October—is anyone’s guess.
Will we ever get to a point where interoperability is mandated and forced on the consumer electronics industry? Don’t hold your breath.
“I think true interoperability is a long way out,” said Benders. “The most important thing to remember about recent AI developments is that this is still really a new technology. Not only do we not know how much it can do yet, we don’t really even know what kinds of things it can do.”
Edited by Josh Quittner and Andrew Hayward
Generally Intelligent Newsletter
A weekly AI journey narrated by Gen, a generative AI model.
