The biggest highlight at this year’s WWDC (aside from that quirky calculator app for the iPad) was Apple Intelligence—Apple’s big play on AI. Among the suite of AI features unveiled, the standout was the impressive improvements to Siri. Often criticized as the weakest voice assistant, Siri’s generative AI makeover could soon leave Google Gemini and dedicated AI hardware like the Rabbit R1 in the dust.
The New Siri: How Voice Assistants Should Be
It’s hard to believe I’m saying this, but Siri might finally be setting the standard for voice assistants. While there’s been a lot of hype about AI’s potential over the past year, practical applications have been scarce. Apple, however, nailed it by showcasing real-world use cases, albeit with some questionable cartoonish AI stickers of your contacts. Apple has demonstrated where the true potential of on-device AI lies.
The logical next step for smart assistants is to handle tasks within apps using voice commands. Rabbit R1 is already dabbling in this space, though it’s currently limited to a small set of apps. I had hoped Google would perfect this with Gemini, but Apple has leapfrogged ahead, making Siri the benchmark for in-app actions with simple, natural language commands.
In a demo, Apple showed how the new Siri can understand your phone’s context, helping with tasks like summarizing articles. While this isn’t groundbreaking (Gemini can do this too), the real magic was in the Photos app. Siri can edit photos based on simple commands like “make it pop,” without needing to fiddle with editing tools. This is a step beyond what Gemini is currently offering, making Siri’s cross-app functions appear more advanced, like attaching a photo to an email draft with a quick voice command.
Siri vs. Gemini and Rabbit: The Showdown
Rabbit aims to make traditional apps obsolete, using a minimalist UI while handling processing on its servers. This is close to the future vision of AI, but it has two major flaws. First, changing user habits is challenging. A middle ground, where users can continue using their apps but delegate tasks to AI, is more practical. Second, Rabbit exists in its own ecosystem, limiting its ability to personalize responses compared to AI on current smartphones. This gives Apple and Google a significant advantage.
Both tech giants are well-positioned to make in-app AI interactions a reality. Apple, however, has taken the lead with its supercharged Siri. At last month’s I/O keynote, Google showcased Gemini’s advancements, including deeper integrations and real-time visual searches. But Apple has forged ahead, demonstrating practical AI applications that Google has yet to fully deliver.
A Lot to Like – And Some Concerns
As expected, Apple emphasized privacy. Most AI processing happens on the device, with cloud queries claimed to be as private as those on your phone. Apple’s generative and language models are optimized for smartphone resources, and its partnership with OpenAI means Siri can leverage the latest GPT-4 model, with user consent before forwarding anything to OpenAI.
However, these polished demos don’t tell the whole story. The real-world performance of these features remains to be seen, especially since many of Siri’s new tricks will roll out over the next year. Additionally, the demos only showcased first-party apps for in-app actions. Apple’s App Intent API aims to involve other developers, but the initial implementation might face some challenges.
This Siri overhaul should give Google some food for thought. While it was amusing that Apple borrowed Android’s Material You icon themes, it would be smart for Google to emulate Siri’s in-app actions. Otherwise, Gemini risks falling behind again, which is likely causing some anxiety at Google HQ post-WWDC. As for Rabbit, it might need to rethink its strategy unless it’s saved by a major player, which wouldn’t be an easy journey either.
Apple has set a new bar with Siri’s AI capabilities, and it’s a wake-up call for its competitors. The race to perfect the smart assistant is heating up, and it’s going to be fascinating to watch how Google and Rabbit respond.
