Only the “chosen ones” will have AI first: is this the new digital divide or just FOMO


Only a small circle of “chosen ones” will have access to the latest AI features, such as Apple intelligence and Copilot+. In a world that’s already divided digitally, should we be worried about this?

Even if you own last year’s greatest iPhone, the 14 Pro Max, you might soon feel the first pangs of digital hunger because it won’t support the latest Apple Intelligence features announced at the Apple Developers Conference, WWDC.

Maybe you bought a Windows laptop this year or earlier? Then no Copilot+ for you, as only Windows “AI PCs with an NPU (Neural Processing Unit) capable of 40 TOPS (trillion operations per second) or higher” are compatible.

You may need a grand or more to buy a device capable of top-notch AI-powered productivity, creativity, learning, or entertainment features.

Some experts believe that this will only widen the digital divide. Heck, many people in the US, and more so everywhere else, are still waiting on reliable broadband internet.

A lot could happen if you’re 2-3 years late to accessing the latest AI capabilities, and no one knows what the consequences might be. Some believe that it won’t be a big deal, as AI is already overhyped and will be accessible even on cheaper devices. Others see two classes of “haves” and “have-nots” forming. However, many will be affected by the fear of missing out (FOMO), and it will likely contribute to sales.

The Digital Poverty Alliance describes digital poverty as “the inability to interact with the online world fully, when, where, and how an individual needs to.”

Another barrier for marginalized communities

“By restricting AI capabilities to only its top-line, most expensive devices, Apple is creating additional barriers to technology for people from more marginalized communities, whether they live in poverty or in regions where tech upgrades are harder to find,” said Gemma Galdón-Clavell, an advisor to the United Nations and EU on applied ethics and Responsible AI, and Founder & CEO of the Eticas Foundation (Eticas.ai).

And this threat compounds with other AI-related broader threats, she believes.

“For example, the biggest threat to digital inclusion and equity today is the use of AI systems that routinely discriminate minority groups, that are only accessible to those who possess certain skills or are only available in regions where Internet access is guaranteed,” Galdón-Clavell explains.

She sees many exclusions at play in the dynamic AI space, and the latest Apple move is another drop of water in an ocean of challenges.

“These issues must be urgently tackled if we want to protect some very key features of our democratic systems,” Galdón-Clavell warns.

Haves and have-nots

Tyler Cooper, editor-in-chief of BroadbandNow, which studies the social, economic, and political issues contributing to the digital divide, reminds us that “for one reason or another,” millions of Americans still do not have access to a broadband connection.

“All the while, we are racing to develop internet-dependent technologies that are poised to fundamentally reshape our economy and our society as a whole. What happens when we get there and that train leaves the station?” he asks.

He doesn’t doubt that Apple Intelligence is becoming the first AI solution to deeply integrate itself into a user’s personal context, bringing new opportunities and use cases. In time, that may change the way we interact with technology in meaningful and profound ways.

“By limiting the vast majority of the intelligence features to the latest models, a new class of haves and have-nots will likely be created,” Cooper noted. “But I’m not as pessimistic that it will lead to folks being left behind completely by this particular product.”

Apple is not alone in working on comprehensive AI solutions – all major tech companies are rushing to introduce their counterparts.

“Sure, this one feels practical and personal, but others like it will come. We can look to Google’s efforts with Gemini here as an example, as well as platform-agnostic solutions like OpenAI’s upcoming ChatGPT 4o voice model,” Cooper said.

He sees the market with all the services and technologies building towards a “unified something,” even if we do not yet know what final form that will take.

“So, while I’m not overly concerned that Apple will create a new digital divide on its own with Apple Intelligence, I am concerned that these features as a whole will widen the existing gaps we have in our broadband market today in the US,” Cooper concluded.

Cooper

Or maybe it’s just FOMO

Some data scientists do not see Apple’s AI proposition as that exciting. One of them is Ilia Badeev, Head of Data Science at Trevolution Group.

“I will express an unpopular opinion, but I am disappointed with Apple Intelligence. For two years (since the first hype of ChatGPT), the whole world has been waiting for a new Siri “on steroids” from Apple. But in the end, we get (at least as we see from the preview) a slightly more “understanding” Siri and just usability improvements in many basic applications,” Badeev said.

He noted that in the initial stages, Apple intelligence will only be available on top devices with the basic language set to English (US).

“So, actually, the circle of “chosen ones” is even smaller. But in my opinion, your questions reflect the fear of missing out – FOMO. Firstly, if you look in detail at Apple Intelligence, it is nothing more than an improvement in the usability of basic functions. You will be able to do what you did before faster, easier, and more efficiently. Does this give you a real advantage over other people? I doubt it,” Badeev said.

He also believes that AI will be rapidly incorporated into smartphones. Very soon, AI will be available on all devices, even the cheapest ones. Even now, while waiting for new Apple features, you can download apps for Claude, ChatGPT, or other models.

He advises users not to rush out and buy new devices, succumbing to FOMO.

“First, Google has long announced the integration of their Gemini into the Pixel 8 Pro, and in many ways, Gemini will perform similar functions (in the foreseeable future). And secondly, most likely AI will very quickly appear on Android in general. This means that AI in phones will soon become a standard, not a feature. Just as touchscreens and cameras once became standard in almost all phones.”

He believes that advanced AI expertise will become a significant state-level advantage, and governments should not interfere with or over-regulate the industry.

“Governments (and even public universities) cannot move at the same speed in AI development as private companies. So regardless of who “should” do it, private companies will be the ones to do it.”

Badeev

Why might AI not be that exciting yet?

While AI gradually makes phones smarter and easier to interact with, Badeev says that truly groundbreaking features are missing.

“I attribute this to the fact that generative AI tends to make mistakes and hallucinate. And when we talk about “native” integration of functions into applications and the OS, hallucinations are unacceptable,” Badeev explains.

Another reason is the issue of energy efficiency. On-device, Apple Intelligence will run a small three-billion parameter model with reduced functions, and it’s only capable of performing relatively basic tasks.

“AI requires a lot of resources, a lot of energy. Therefore, it is obvious that the first integrations are made only on “top” hardware due to their lower sensitivity to power consumption,” Badeev said.

The first generation of AI devices will have fewer capabilities, but only at first. Lightning-fast technology development can solve the energy efficiency problem.

It’s not that “Apple and other companies want to make AI “only for the rich” – this is nothing more than a current issue of model power consumption,” Badeev concluded.