Meta AI Chief Alexandr Wang has credited former Apple executive Daniel Gross with helping set up an unexpected partnership between Apple and Meta. This collaboration enables Meta’s Muse Spark Open artificial intelligence model to operate on Apple’s premium Mac Studio computers.
Speaking to Axios, Wang stated that Gross—who is currently employed at Meta—was instrumental in uniting the two firms. Wang noted that the partnership has been in progress for several months.
Daniel Gross Helped Bridge Apple and Meta
Wang highlighted Gross’s efforts in fostering cooperation between the two tech giants. “I think he played a really big role in kind of getting the two companies to work nicely together,” Wang told the outlet. He also offered a playful comment regarding the restored relationship. “We’re excited for mommy and daddy to be playing nice again,” he remarked.
Apple and Meta share a lengthy history of conflict, focusing heavily on user privacy and digital platform oversight. These frictions became more pronounced in 2018 when Apple CEO Tim Cook publicly criticized Facebook’s management of the Cambridge Analytica data breach.
Subsequently, Meta CEO Mark Zuckerberg characterized Cook’s comments as “extremely glib.” Relations worsened further in 2021 following Apple’s rollout of privacy updates that limited cross-app user tracking.
Also Read: Xcode for Apple Developers: Tools for Building iOS Apps
Meta’s Muse Spark Open Runs on Mac Studio
The joint initiative focuses on Muse Spark Open, an AI model boasting 816 billion parameters. Meta identifies this as its most advanced open-weight model developed in the US so far. The system functions across a cluster of two or four Mac Studio computers powered by Apple’s M5 Ultra chips, utilizing Apple’s MLX framework to share processing tasks between the devices.
Designed for intensive functions, Muse Spark Open handles complex operations like lengthy coding projects, multi-step reasoning, and video evaluation. Developers can access and download the model via services like Hugging Face, adherence to licensing terms permitting. Running the software locally also provides privacy benefits; Wang explained that Meta is unable to view the prompts or documents input by developers while utilizing the model on personal hardware.
“Meta has no control or visibility to any of that,” the Meta AI chief stated. Even so, operating the system demands significant hardware outlays, requiring a minimum of two Mac Studio units, which presents a high cost barrier for individual developers.
