This will likely be the year Apple will start using the magical Generative AI term for its upcoming software features. While we heard a lot about Siri upgrades Cupertino is said to be planning, the news aggregator and part-time leaker yeux1122 – on the Korean blog Naver – said to have a developer’s source about the latest progress in applying Generative AI to Siri.
According to them, Siri’s Generative AI version will be announced at WWDC 2024. It will be Ajax-based, as previously reported by Bloomberg and The Information. Ajax is a Large Language Model trained inside Apple for years now, and it’s said to be better than ChatGPT 3.5. While it was first thought that Cupertino would keep this chatbot for internal usage, yeux1122 shares that it will add “more personalization and natural conversation capabilities to Siri.”
It will help with efficient management between various Apple devices, offer Apple-specific creation services, link between various external services, and even subscription-based options.
While this all seems a little bit foggy, The Information previously shared Apple has two teams to develop language or image models. One of the teams might be working on software that generates “images, video, or 3D scenes,” while the other is working on “long-term research involving multimodal AI which can recognize and produce images or videos.”
The publication said that one of these LLMs could “eventually interact with customers who use AppleCare.” At the same time, the Siri team plans to incorporate these language models to make complex shortcut integrations a lot easier.
That said, what yeux1122 is reporting could make a bit more sense with the other stories’ insights. In addition, they previously wrote that Apple is preparing to use a Large Language Model to completely revamp Siri into the “ultimate virtual assistant,” becoming the “killer AI app for accessibility across Apple’s vast array of devices.”