Apple’s big event this week unveiled some noteworthy updates, including iPhone 17, iPhone Air, and AirPods Pro 3. But more than a year after Apple’s big 2024 WWDC, in which it trumpeted the new Apple Intelligence, the company’s AI play is still in limbo and conspicuously absent from this week’s updates.
In fairness, this week’s event was Apple’s annual focus on hardware. And some hardware updates came with a nod to Apple Intelligence, including AirPod Pro 3’s live-language translation. But aside from a few integrations to date, Apple Intelligence’s promise – including its role in Siri – has been unfulfilled.
All of this is further contextualized by last week’s revelation that Apple could be developing an AI alliance with Google. The thought is that it’s stumbling to get Apple Intelligence out the door as originally advertised. So it’s turning to Google to outsource AI and enlist Gemini to power Apple Intelligence.
Though Apple and Google already partner – such as Google’s default positioning in Safari and Siri on Apple devices – the news of this AI alliance was ironic to many. These are big-tech rivals after all. But they’re also incumbents fending off AI-pure play challengers. The enemy of my enemy is my friend.
All Roads Lead Back to the iPhone
But going beyond that initial knee-jerk reaction to the AI alliance, a bigger and deeper question we have is why Apple didn’t just do this all along. By ‘this,’ we mean to outsource the underlying functionality in Apple Intelligence. Why did it feel the need to not just offer AI but be at the center of its functionality?
One answer to that question is Apple’s classic vertical integration. It likes to own the entire stack so it can offer elegant sensor fusion and software/hardware cohesion. But at the end of the day, Apple makes most of its money selling you hardware, and it doesn’t need to go down an AI rabbit hole to do so.
Panning back for context, Apple operates several apps and software functions, such as Maps, Siri, Music, etc.. Though some are revenue centers (TV+, Music, etc.), the rest are loss leaders. Apple doesn’t make money from Maps et al, but rather dangles them as exclusive enticements for the iOS universe.
That’s a fancy way of saying it wants to sway you to buy an iPhone versus an Android, and that requires answering questions about what the device can do… and what you can do on the device. This gets back to Apple Intelligence: Just like all roads lead back to selling hardware, AI is meant to move iPhones.
Another Strikeout
The problem is that Apple’s eyes were bigger than its stomach. Once it went down the AI road, it realized that this is deep tech, and it ain’t easy. Moreover, it can’t excel. AI requires data, and Apple can’t compete with the likes of Google (just as it learned when it launched Maps), and its knowledge graph.
So that brings us back to the present, and the part where Apple realizes it has to outsource AI if it’s going to offer something up to its standards. In fact, there are a few products that haven’t met that standard in the past few decades. At the top of that list is Siri, where it can’t afford another strikeout.
Further in support of outsourcing AI to Google is the elephant in the room: privacy. Over the course of the privacy age of the past decade, Apple has explicitly positioned itself as the privacy-first player. And that has meaningfully helped it achieve its core priority noted above – to sell more iThings.
But Apple’s biggest strength in this privacy persona is that it’s believable. I believe Apple when it tells me that it’s intent on protecting my data, simply because it doesn’t need that data to fuel its core revenue engine. I don’t believe Meta, Google, and other ad-supported businesses when they sing the same tune.
Privacy Positioning
Synthesizing all the above, the question re-emerges about why Apple ever wanted to own and operate its AI engine. Its biggest advantage in its privacy positioning – its perceived motive – erodes if it becomes an AI company. At that moment, it gets lumped in with Google and Meta in terms of trust issues.
The endpoints of its data collection for AI would be first-party utilities as opposed to ads – so perhaps Apple is a bit more trustworthy in that scenario. But the point remains that Apple degrades its privacy position by going down the AI road. Perhaps it wanted to do it anyway, given AI’s gravitational pull.
But now, here we are, at a point when Apple is shifting gears in and considering an unlikely alliance. And perhaps that’s a good thing in that it can offer a better AI product, sidestep the CapEx-intensive AI arms race, and maintain clean hands when it comes to privacy, data collection, and public image.
In fact, that formula would align with Apple’s main objective, which once again is to sell more devices. Of course, this path isn’t fully optimal and has some downsides, including enriching Google with a gold mine of training data. But it’s the lesser of two evils and much cheaper than buying a high-ticket AI company.
Header Image Credit: Sumudu Mohottige on Unsplash
