Apple Can ‘Distill’ Google’s Big Gemini Model

Before we get to today’s column, we wanted to flag OpenAI CEO Sam Altman’s major reorg, the company’s new "Spud” model and its decision to shut down the Sora video app and application programming interface, which we scooped on Tuesday. (For what the company’s organizational structure looks like now after the changes, check out our updated OpenAI Org Chart.)

The latter decision seems in line with product chief Fidji Simo’s mandate to cut down on costly “side quests.” Sora was a major user of servers, and servers are in short supply. So RIP Sora, and thanks for making this one cat video.

On to today’s column… 

Apple’s AI partnership with Google to help the iPhone maker improve its Siri AI assistant and other long-promised AI features is deeper than previously known. 

While we have reported that Apple can tweak, or fine-tune, a version of Google’s Gemini AI so that it responds to queries the way Apple wants, the agreement gives Apple a lot more freedom with Google’s tech.

In fact, Apple has complete access to the Gemini model in its own data center facilities. Apple can use that access to produce smaller models that power specific tasks or are small enough to run directly on Apple devices so they can run the tasks faster, said a person who has direct knowledge of the arrangement.