Remember that clunky virtual assistant who always seemed a step behind the competition? Siri, the voice behind Apple devices, is known for being helpful at times, but not exactly known for its brilliance. Well, fret no more Apple fans, because a recent research paper by Apple researchers suggests Siri might be on the verge of a major upgrade.
The paper, titled “Re ALM: Reference Resolution As Language Modeling,” outlines how Apple’s AI system is being designed to take a big leap forward. The key here is understanding context. Imagine you’re looking at a picture of the Eiffel Tower on your phone and ask Siri, “What’s the weather like there?” In the future, Siri, powered by this new tech, might not just search the web for a generic weather forecast. Instead, it would understand you’re referring to the Eiffel Tower on your screen and provide a weather report for Paris.
This “reference resolution” essentially allows Siri to analyze what you’re doing on your device, be it an image, song, or even an alarm, to understand the context of your questions. With this newfound contextual power, Apple claims its AI can outperform even the mighty GPT-4, the brains behind the popular ChatGPT Plus tool.
This is a big deal. GPT-4 is a large language model (LLM) considered one of the best AI models in the world right now. Beating it, even in a specific task like reference resolution, is no small feat for Apple’s on-device AI. Imagine Siri seamlessly connecting the dots between what you’re looking at and what you’re asking, a far cry from the occasional misunderstandings of today.
And Siri isn’t the only underdog getting a potential boost. Samsung’s Bixby, another virtual assistant with a bit of a reputation (though perhaps not as stellar as Siri’s), is also rumored to be getting its own generative AI features. This means no matter if you’re an iPhone or Android user, the future of AI assistants seems to be getting brighter.
Apple’s research is impressive, but there’s still room for improvement. The paper itself acknowledges that while their approach is effective, there are limitations. For instance, complex questions requiring a nuanced understanding of an image’s layout might still trip up the new Siri. The researchers propose splitting the screen into grids and encoding those positions to improve understanding, hinting at exciting future developments.
The buzz around Apple’s next-generation AI is undeniable. Many expect it to be the star of the show at WWDC 2024, Apple’s annual developer conference. With Apple’s marketing chief himself calling the event “Absolutely Incredible,” it seems they’re confident in their progress.
While we’ll have to wait and see how the new Siri stacks up against the likes of Chat GPT’s voice function when iOS 18 drops, one thing’s for sure: Apple is making significant strides in the AI race. This is great news for Apple users who might have been watching the progress of their competitors with a hint of envy. Siri might finally be ready to shed its reputation as a not-so-smart assistant and become a truly intelligent companion on your Apple device.
