Apple developing AI features to bridge gap with Google, OpenAI
What's the story
Apple is said to be developing a variety of AI-driven features, including an enhanced version of Siri, which could be unveiled as early as next year.
Bloomberg's Mark Gurman reported that Apple executives were "caught off guard" by the sudden surge in AI interest within the industry and have been working to keep pace since late 2022.
The initiative to incorporate AI-based capabilities into Apple's offerings is being spearheaded by senior vice presidents John Giannandrea, Craig Federighi, and Eddy Cue.
Details
Smarter Siri and iOS 18 AI features
According to Gurman, Giannandrea's group is working on a more intelligent Siri, deeply ingrained with AI, which could be launched by next year.
Concurrently, Federighi is supervising the incorporation of AI into the upcoming major iOS release. This involves a reimagined interaction between Siri and the Messages app, allowing users to handle intricate inquiries and complete sentences more efficiently.
Apple is also exploring the integration of AI into productivity tools like Pages and Keynote.
More
AI enhancements for Apple Music and Xcode
Cue is advocating for the implementation of AI elements in Apple Music, like automatically-generated playlists.
Additionally, Federighi's team is examining the possibility of integrating AI into Xcode to expedite developers' coding process, aligning it with services like Microsoft's GitHub Copilot.
Apple plans to allocate approximately $1 billion annually to AI research and product development in an effort to compete with industry giants like OpenAI, Google, and Microsoft.
Insights
On-device vs cloud-based AI debate inside Apple
Reportedly, Apple is currently debating whether to implement generative AI as a solely on-device experience, a cloud-based system, or a combination of both.
While on-device operation ensures privacy, it restricts the potential of large language models compared to server farm-based models.
Gurman has suggested that Apple will likely determine the approach on a case-by-case basis, with some features operating entirely on-device and others depending on a cloud infrastructure.