Apple is in the AI business. After two years of watching its competitors — Google and Microsoft — surge ahead with their own incorporations of generative AI, Apple CEO Tim Cook announced “Apple Intelligence” at its Worldwide Developers Conference on June 10. The new features include information retrieval across apps, automated proofreading, enhanced photograph search softwares, and user image generation. With perhaps the most significant enhancement, Apple will provide its 2.2 billion active users the option to integrate ChatGPT into their iPhone experience.
What does this news mean for Apple’s business model and the future of the AI landscape? Milan Miric, associate professor of data sciences and operations, shared his thoughts on the ripple effect of Cook’s announcement.
Interviewer: What did you expect from Apple AI before the keynote address? Did Apple meet your expectations?
MM: If we think about Apple in the competitive context of other technology companies, they dominated the personal hardware space, whereas Microsoft and Amazon Web Services (AWS) have dominated the enterprise space. Their success has come from being able to sell products such as the iPhone and Mac, and even though they might sell additional subscription services, it’s really about products targeted at consumers who are using their own money to buy devices.
For this reason, my expectation was they would build on AI that is complementary to these core technologies. From what I heard in the keynote, that is exactly what they are doing. The first hour was mostly focused on introducing very incremental product features (such as lock at the app level rather than the device level, which is positioned as a broad safety issue, but I think it's more just to prevent interpersonal snooping). This signaled to me their focus remains very much around products, and using AI to augment those products.
What stood out to you about Apple’s approach to AI?
MM: In the Apple Intelligence case, this seems to be a broad bucket of tools that do things one might see in other software, but integrates it again around Apple’s consumer hardware products. For instance, the ability to track what is on your screen and on your device is reminiscent of what other applications are doing such as Grammarly or Gmail. These are not radically new by any means. However, what Apple is doing is integrating it as part of their OS rather than as an add-on you can get. The same is true for the integration of tools to their software suite.
What makes Apple’s AI philosophy different from their competitors, like Google, Microsoft, and Samsung?
MM: What I think is at the crux of this is a positioning around security. You buy an expensive device from Apple, but you know they are making money from selling you that device and not from selling you ads or selling your data like their biggest competitors. Therefore, similarly even if they are scanning your data, email, personal photos, etc., they are implicitly and explicitly guaranteeing this will be used privately, and for your own benefit, rather than theirs.
This integrates smartly with their overall approach. For example, they have their own AI chips (e.g. M2) which enables your hardware the opportunity to run this software locally (so you don’t need to send it to an external server). Then, they are also offering a personal cloud, again with the idea that it’s not going to be accessed by a third party and only intended as a means to sell you better hardware.
