Apple’s own research sheds light on Siri’s ability to catch up with AI

With the introduction of the new iPad Mini, Apple has made it clear that an AI-powered software experience is the way forward. And if that meant making the same internal upgrades to a tablet that cost almost half as much as its flagship phone, the company would still be moving forward.

However, its ambitions with Apple Intelligence lack competitiveness, and even by Apple’s own standards the experience has failed to delight users. In addition, the staggered rollout of the most ambitious AI features – many of which are still in the future – has left a poor impression on enthusiasts.

Now it appears that the reason for the delays has something to do with quality and performance, as Apple’s own testing shows. “The research found that OpenAI’s ChatGPT was 25% more accurate than Apple’s Siri and could answer 30% more questions,” a Bloomberg report said.

As a reminder, Apple’s position on Siri is pretty unique. Siri, for example, gets improved natural language understanding and deeper integration with apps and local files. However, there are tasks that cannot be completely completed. In such situations, queries are seamlessly shifted to ChatGPT.

This is part of a deal Apple made with OpenAI. Now it would make sense that Siri wouldn’t be able to do quite the same internet-connected tasks as ChatGPT, especially since Siri and ChatGPT are two completely different products. However, Apple doesn’t just use OpenAI’s tech stack for Siri.

According to OpenAI, ChatGPT will also help users with “image and document understanding.” Writing tools – which have already found their way into tools such as Notes and Safari – also use the ChatGPT Kitty. In addition, image generation is also handled by OpenAI technology.

Given its heavy reliance on ChatGPT, one might think it’s because Apple isn’t quite on the leaderboard with its own AI technology stack that could compete with Google’s Gemini or Meta. This assumption won’t be entirely implausible, and even Apple’s team seems to be okay with the status quo.

“In fact, some at Apple believe its generative AI technology is — at least so far — more than two years behind industry leaders,” the Bloomberg report adds. But it’s not just about progress, but also about the pace of introduction.

Take a look at Galaxy AI, Samsung’s take on an AI ecosystem that’s already deployed on a variety of its phones and computers using Google’s Gemini stack. Chinese smartphone manufacturers have been offering generative AI features such as image generation and a next-generation assistant for some time.

At this point, it seems almost certain that Apple’s strategy with Apple Intelligence has been rushed, apparently to allay investor concerns that the company is lagging in the AI ​​competition. So far, what little we’ve seen of Apple’s “AI revolution” has been anything but revolutionary.

The best implementation of Apple Intelligence to date has been notification summaries and prioritization, but these are useful features rather than something that would reshape the software experience for users. It would be interesting to see Apple bring new energy to its AI approach next year.

So far, the company has made no such announcements, and even the promises it made at its developer conference earlier this year have yet to be fulfilled.