Apple, an early champion of AI and machine learning, is without a doubt lagging behind almost everyone else when it comes to generative AI – which is why I can almost believe the rumor that Apple is ready to get into quick bed with often rival and constant search partner Google to license the latter’s powerful Gemini models for iPhone-based AI operations.
I can almost buy it – but there are many reasons why I ultimately don’t.
However, let’s first remember how Apple first took off and then ended up in a less-than-mediocre position in the AI market.
If you think it started with Siri, you’re wrong. When I emailed him about the rumor, Apple analyst and said Creative strategies Chairman Tim Bajarin reminded me Knowledge navigator, a software agent-based search engine concept from 1987, not entirely unlike an AI chatbot, that predicted digital voice assistants like Siri, which Apple launched almost 25 years later. Apple has thought about this for a long time.
Siri was perhaps the first digital voice assistant on the phone, and for a brief second it was quite impressive. However, despite countless brain transplants and Apple’s knack for using AI, neural networks and machine learning for everything from battery management to autofill and photo processing, Siri quickly surrendered pole position to Amazon Alexa and Google Assistant.
By the time we reached the age of generative AI, Apple was inexplicably absent, offering nothing worthy of generative AI at WWDC 2023 and instead making a big fuss about an expensive mixed-reality headset. Granted, the Apple Vision Pro is an excellent distraction, but due to its price and the fact that you wear it on your face, it’s unlikely to have the same impact as a major AI update.
Does this AI partnership make sense?
Since the summer we’ve heard rumors (including from Bloomberg’s Mark Gurman) that Apple was hard at work on Project Ajax, which could include something called Apple GPT (domestically developed, prompt-based generative AI) and perhaps some ways to access the (large language model) LLM of your choice from various partners.
I think it’s also accepted wisdom that this year’s WWDC will be 90% AI. Apple cannot afford to disappoint.
Gurman’s latest rumor, that Apple might simply license Google’s powerful, multi-modal Gemini AI for various onboard iPhone operations, doesn’t quite add up — at least to me.
First of all, Gemini Nano is the only part of Gemini that meets Apple’s strict privacy requirements. It can be run locally and, like John Giannandrea, Apple’s head of AI, made this clear a few years ago, running any other way is “technically wrong.” I don’t suspect Giannandrea suddenly changed his mind and would allow more powerful Gemini models on the iPhone when they all require cloud access to work.
However, Gemini Nano is so limited (text suggestions, grammar checking, summaries) that it won’t provide anything for the generational leap that Apple and its next iPhone need.
DIY AI
I’m not suggesting that Apple would never work with Google. Your iPhone search defaults to Google because the search giant pays Apple billions annually. It’s a fantastic deal for Apple, but I think the company would like to have its own search engine (there have been rumors for years that it’s trying to build one).
There’s a reason Apple Maps, which also uses Apple’s own AI, exists. I think it’s a point of pride for Apple that after a disastrous launch in 2012, the company updated, iterated, and improved Apple Maps until its embarrassing origins were a distant memory.
It’s the same reason Chrome isn’t your iPhone’s default web browser. Apple likes to build things itself, mainly because of the complete control it gives it.
AI is perhaps the most important innovation in our lifetimes, and I think Apple wants its place on that plaque. If it licenses Gemini widely, its generative AI efforts will never recover. Apple’s massive user base would also boost Gemini past OpenAI’s GPT, Microsoft’s CoPilot and Met’s LLaMA.
However, Bajarin sees it differently. He points to the cost of delivering a powerful generative AI architecture. With that in mind, the shortcut offered by using these “basic AI architectures” can be hard to resist.
There’s also one key way Apple can’t compete with Google: “Even if Apple had its own Gemini model, it likely wouldn’t have the infrastructure to serve its massive customer base,” Bajarin said in an email.
Maybe it’s not about the infrastructure, or about having a solid foundation on which Apple could build a custom model (think of how Apple Silicon is built on the ARM foundation). Maybe it’s a financial decision.
Follow the money
Gurman argues that Apple and Google could do this to redress the balance if the European Union forces Apple to get rid of Google Search and let consumers choose the default. Aside from the confusion and frustration this could cause, there are the billions in lost revenue that Apple would have to recoup. Yes, Google could pay a similar amount to put its Gemini AI front and center on the iPhone.
Yet it all makes no sense to me.
Apple’s goal with the iPhone 16, iOS 18 and future iPhones is to differentiate its products from Android phones. It wants people to switch and they will only do so if they see a tangible benefit. If the generative tools on the iPhone are the same as on the Google Pixel 8 Pro (and 9) or Samsung Galaxy S24 Ultra (and S25 Ultra), why switch?
I’ve seen reports that Apple has met with multiple AI companies (while they continually acquire smaller ones). I believe the goal here is to learn how they do it, and add to that what Apple’s AI efforts are still missing to not only achieve parity, but also outpace the AI competition – and that includes Google Gemini.