Core Insights - Apple plans to pay approximately $1 billion annually to acquire a high-performance AI model developed by Google, which has 1.2 trillion parameters, to enhance its Siri voice assistant technology [1][2] - The collaboration aims to revamp Siri's core architecture, with a new version expected to launch in spring next year [1][2] - Apple has evaluated multiple third-party models before selecting Google's Gemini, intending to use it as a transitional solution until its own AI model is competitive [1][3] Group 1 - The Siri transformation project, codenamed "Glenwood," is led by Mike Rockwell and Craig Federighi, and will be integrated into iOS 26.4 [2] - The Gemini model will enhance Siri's capabilities in summarization and task planning, while some functions will still utilize Apple's proprietary models [2] - The model will operate on Apple's private cloud servers, ensuring user data remains isolated from Google's infrastructure [2] Group 2 - Apple acknowledges its lag in the AI sector by adopting Google's technology, demonstrating a willingness to catch up [3] - Despite the collaboration, Apple does not view Gemini as a long-term solution and is accelerating the development of its own trillion-parameter cloud model [3] - The Chinese version of Siri may abandon the Gemini technology route due to restrictions on Google services, potentially using a self-developed model in conjunction with Alibaba's content filtering [3]
为Siri植入“超级大脑”!传苹果(AAPL.US)拟10亿美元年费购谷歌(GOOGL.US)Gemini作过渡方案