How Apple Just Undercut the AI Hype Cycle

Unlike too many of its rivals - Google, Facebook, Microsoft et al - Apple has taken the time to assess the market and gleaned that consumers and corporate customers are still wary about AI resulting in lower than expected demand . 

Rather than leading with breathless hype, Apple is focusing on what its already huge customer base might actually want and use, rather than making promises about features no one is sure they really want. And the result will both bolster its sales while undermining quite a few AI startups and their backers. JL

Derek Robertson reports in Politico:

Usually when Apple makes a big announcement about new technology, it comes with the fanfare (and price tag) befitting the company that gave the world the iPhone. The “Apple Intelligence” platform is more low-key, with Apple announcing improvements to its already AI-powered Siri, AI-generated emojis and the integration of ChatGPT into iPhone software. Those aren’t the radical changes the iPhone has delivered in the past. “Public demand doesn't match Silicon Valley's AI sugar rush. “Rather than lose themselves to the AI hunger of investors and VCs, [Apple is] focusing on what works, what consumers might actually use, while trying to game out the tangible value of these innovations.”

Apple unveiled its long-awaited “Apple Intelligence” suite of AI tools yesterday, but it didn’t garner quite the level of excitement to which the hardware giant is accustomed.

Usually when Apple makes a big announcement about a world-shaping new technology, it comes with the fanfare (and price tag) befitting the company that gave the world the Macintosh and the iPhone.

The “Apple Intelligence” platform is a bit more low-key, with Apple announcing some improvements to its already AI-powered Siri voice assistant, AI-generated custom emojis and the integration of ChatGPT into iPhone software.

Those features aren’t nothing. One VC openly pondered on X “How many startups did Apple kill with Apple Intelligence?”, positing that by rolling AI functions like translation, composition, and image generation into Apple’s ubiquitous hardware a whole generation of intermediary companies could get skipped over.

But even those (if they happen) aren’t the radical kinds of changes that the iPhone has delivered to people in the past — think ridesharing, dating apps, instant payment transfers and QR codes. Those weren’t just cool apps or features that made your phone itself more useful; they’ve had fast, jarring impacts on the physical world around them. Using ChatGPT to help compose a text saying you can’t go to a friend of a friend of a friend’s birthday party is unlikely to have the same effect as Uber had on the taxi industry. Analysts have accordingly predicted a slower rollout from the phone-buying public.

For the wider tech landscape, this isn’t just a business story. Apple’s AI non-splash is an object lesson in how wildly the hype around new technologies can diverge from their actual application.

Apple is different from other AI newsmakers: It’s not a startup seeking capital or hoping to hitch its wagon to a corporate behemoth, or an intellectual entrepreneur looking for a next gig or a flagging industrial concern in need of subsidy. It’s one of the world’s biggest publicly traded consumer product companies, facing the challenge to innovate while keeping the ship steady for shareholders — meaning its partnership with the would-be world-changers at OpenAI might give us that rarest of things, a pragmatic window on how people actually use AI tools.

“Public demand simply doesn't match Silicon Valley's AI sugar rush,” said Matt Mittelsteadt, an AI and cyber policy fellow at George Mason University’s Mercatus Center. “Rather than lose themselves to the AI hunger of investors and VCs, [Apple is] focusing on what works, what consumers might actually use, while trying to game out the tangible value of these innovations.”

Mittelsteadt is dead on about at least one thing: The public is still wary of AI, as polling featured in this newsletter from the AI Policy Institute has repeatedly shown, as well as other polls from the Pew Research Center and Elon University. While conversations in Silicon Valley center around whether AI could (quite literally) blow up the world, or whether the Chinese Communist Party might use it to take the world over, the average person is more concerned about whether AI tools will harm their reputation or even just let them down when they’re on deadline.

Policy concerns outside Silicon Valley, too, tend to be more quotidian if no less significant. Apple stressed in its presentation yesterday that as many of its AI offerings as possible would run natively on users’ devices, so as not to share data with outside parties, and that when networked computing was required, it would be done through something called “private cloud compute.”

Tech news site Ars Technica noted that while the presentation was light on details, Apple made a promise that the server code would be open to outside researchers who could probe and verify its privacy and security.

Of course, there is one other way this could be disruptive, at least to Apple.

Antitrust enforcers have been circling the AI industry. The Biden administration is deeply concerned about the growing web of connections between AI companies like Microsoft, OpenAI and Nvidia, and is currently divvying up jurisdiction between agencies.

This seemingly modest dip of the toe into the AI waters means that iPhone users will now be able to ask ChatGPT questions through Siri — something likely to grab antitrust enforcers’ attention. There’s already a Department of Justice case against Apple for how it uses its ecosystem to control the app business, as well as an ongoing lawsuit from the Department of Justice over payments from Google to Apple for its status as the iPhone’s default search engine.