The 1.2-trillion parameter model will serve as the basis for the next-generation Siri.
A new report from Bloomberg says that Apple is finalizing a deal in which it will pay Google about a billion dollars a year to use its AI technology as the foundation for the long-overdue new Siri.
The new model is essentially a customized version of Google’s Gemini LLM, according to Mark Gurman’s report. Apple reportedly evaluated Claude and ChatGPT earlier this year and zeroed in on Google to create an LLM based on its Gemini model.
The version Apple will employ will have 1.2 trillion parameters, in comparison to the 150 billion parameters of the cloud-based Apple Intelligence model used today. The number of parameters is one measure of the size and complexity of an AI model, but not the only one.
