Apple’s quest to dominate AI might involve a surprising ally: Google. According to Bloomberg’s Mark Gurman, the Cupertino-based giant is in early talks to use Google’s Gemini AI as the backbone for a revamped Siri launching next year. This could help Apple close the gap in generative AI, where it’s lagged behind rivals.
Apple has faced engineering hurdles, delaying a major Siri update originally set for this spring. The new version promises seamless voice control using personal data. Due to ongoing challenges, Apple’s AI leader John Giannandrea has been removed from the Siri project, with responsibilities now under Craig Federighi, head of software, and Mike Rockwell, who leads the Vision Pro team.
They’re exploring external help after initial chats with Anthropic (Claude) and OpenAI (ChatGPT). Anthropic was a top pick, but high costs broadened the search. Google has even started training a custom model for Apple’s servers. Apple’s running a “bake-off” between two Siri prototypes: Linwood (using in-house models) and Glenwood (third-party tech).
Apple’s AI team is experiencing some turbulence. Back in July this year, key architect Ruoming Pang left for Meta, enticed by a compensation package reportedly exceeding $200 million. A number of his colleagues have since departed as well, and many remaining staff are looking at other opportunities, spurred by the possibility of Apple leaning more on external AI tech. This potential pivot to third-party models marks a departure from Apple’s traditional focus on in-house development for privacy reasons. However, any such models would still run on Apple’s secure Private Cloud Compute infrastructure.
Could Gemini supercharge Siri? It’s reportedly a key option as Apple races ahead. What are your thoughts—share below! Stay updated with the latest news on this by downloading the Appleosophy App from the App Store or by visiting our website.
