A lot is riding on next week's WWDC 2024 keynote. The importance of this presentation is much higher than the standard post-event market trends. The pressure on Tim Cook and his team to deliver is actually even higher than it was before the Vision Pro launch last year.
On Monday, Apple will unveil its AI plans, a topic that has had a huge question mark hanging over Cupertino for the past few years as rivals like Google and Microsoft have embraced generative AI. There is a broad industry consensus that systems powered by large-scale language models like ChatGPT and Gemini will have a major impact on how we interact with our devices.
Apple is expected to announce a partnership with OpenAI to bring its intelligence to iPhones and Macs. Apple's short-term strategy is to deeply integrate generative AI with existing features, with Siri at the center. Since its introduction in 2011, Apple has been pushing to make its voice assistant an integral part of all its operating systems.
But 13 years later, Siri hasn't been the revolution Apple promised. There are many reasons for this, but the main one is competence. The concept of an artificial voice assistant has been around for decades, but there's a reason no one has been able to fully crack it. As phone makers and app developers have turned smartphones into every-devices, the job of these assistants has become increasingly complex.
Stanford Research's research was impressive, but the technology needed for a frictionless experience wasn't quite there yet. Siri co-founder Norman Winarski addressed the underlying problem in 2018, noting that Apple's original plan was for a much more limited assistant that would handle things like entertainment and travel. “These are hard problems, and when you're a company with a billion users, the problems get even harder,” Winarski said at the time. “Maybe they're looking for a level of perfection that's just not achievable.”
Generative AI hasn't reached that level of maturity yet, at least not yet. Hallucinations are still an issue, which is why it feels like we're still just taking our first steps, even after all the hype over the past few years. If anything, Google has been overly aggressive in some areas, a prime example of this being their decision to put Gemini results at the top of searches.
The world's dominant search engine, prioritizing it over trusted resources, needs to be as accurate as possible, which means not telling users to eat glue. Google calls the Gemini results a product of its “Search Labs,” but most users probably won't understand what that means in terms of product maturity, nor will they be willing to click through to find out more.
Over the past few years, I have met several researchers who use the word “magic” to describe the “black box” results surrounding large-scale language models. This is not a criticism of all the great research being done in this field, but rather just a recognition that there is a lot we still don't know about this technology.
Arthur C. Clarke put it this way: “Any sufficiently advanced technology is indistinguishable from magic.”
But where Google has been more intentional is with its integration of Gemini into Android. Rather than completely replacing the Assistant, Google is integrating its generative AI platform into various applications. Users can also make Gemini the default by assigning it to the Assistant button on their Pixel devices. At least so far, this implementation requires deliberate action on the part of the user.
Gemini hasn't completely conquered Android yet, but Google is clearly hinting that it will one day completely replace Google Assistant. I half expected such an announcement at I/O last month, but it's good to see that Google ultimately chose to give Gemini some more time to mature.
Whether the Assistant name sticks is ultimately a branding decision. Apple, on the other hand, is very attached to the name Siri — after all, the company has spent more than a decade marketing the product to consumers — but sooner or later, generative AI will dominate the smart assistant space.
Voice assistants in general are facing an existential crisis. Smart speakers are the precursors to broader platforms like Siri, Alexa, and Google Assistant. Shipments are declining after surging during the pandemic. It's unfair to label the sector as on the brink of doom, but without the right push, that's likely in the long term.
While generative AI is poised to be its logical successor, the first set of hardware devices built around these models, like the Humane Ai Pin and Rabbit R1, are only a testament to how far the category needs to go before it can be considered a consistent experience for mainstream users.
Apple will finally show its hand on Monday. After the electric car collapse, there are rumors that the company will move some employees to AI generation, but all signs point to Apple having ceded a significant lead to its competitors. Therefore, the most logical strategy for the company would be to partner with a leading company like OpenAI.
Shortly after the Siri acquisition was announced, Steve Jobs was asked if the company was trying to beat Google at what it does best: “We're an AI company,” Jobs said. “We're not going to get into the search business. We're not interested in it. Other people are doing that well.”
The company's approach to generative AI is at the same stage today — at this stage, Apple can't compete with OpenAI's proprietary tech, so it's partnering with them instead — but even the best of the current models will be a while away from being ready to completely replace your current smart assistant.