Apple WWDC 将于周一开幕,今年将揭示如何将 AI 模型深度集成至操作系统中。已知是专为 Apple 开发的 Gemini 版本,拥有 1.2 万亿参数,但本地运行更小的模型;复杂查询则调用 Apple 自有服务器上的大模型。悬念在于:Apple 版 Gemini 实际表现如何?集成深度怎样?是否会推出与 Siri 对话的语音模式或 iOS 语音控制?
Apple's WWDC starts on Monday. This year it promises to be particularly interesting.
The secret will be revealed of how Apple has finally integrated a proper AI model deep into its operating system.
As is known, it's a Gemini version, developed specifically for Apple, with 1.2t parameters, although a much smaller model runs locally. For complex queries, however, the larger model is accessed on Apple's own servers.
The exciting question remains: how good is Apple's Gemini really, and how deeply is it integrated into the OS? Will there be a voice mode to talk to Siri? Will iOS be controlled by voice? I'm looking forward to Monday.