Why Apple's AI Head's Poaching By Meta May Be Less Important Than It Seems

News of Meta's poaching Apple's AI head may be less significant than initial reports suggest. 

This is because Ruoming Pang and his Apple team were already under fire for the relatively unimpressive performance of their work compared to that of OpenAI, Anthropic and others. There have been persistent reports that Apple had already decided to use OpenAI and Anthropic for its AI reboot of Siri, which may have prompted Pang's willingness to defect. And while Meta is throwing a lot of money around (tens of millions each annually for Pang and others), Zuckerberg's ego and grandiose ambitions (Metaverse, anyone?) have often led to less than anticipated returns. JL

Maxwell Zeff reports in Tech Crunch and MacRumors reports:

Apple’s head of AI, Ruoming Pang, is leaving to work at Meta. Pang was in charge of a team with 100 employees working on Apple's large language models. Meta lured Pang with a deal worth tens of millions of dollars per year. Meta has also scooped up AI experts from OpenAI, Anthropic, and Scale AI. Models developed by Pang's team are used for Apple features like email summaries, Priority Notifications, and Genmoji, but haven’t been a huge success. They’re less capable than OpenAI's, Anthropic's, and Meta's. Apple has reportedly considered Anthropic or OpenAI for an AI version of Siri, rather than using its own models. Those discussions impacted the morale of the team Pang is leaving. Other engineers on the team are also mulling offers from other companies.

Apple’s head of AI models, Ruoming Pang, is leaving the company to work at Meta, Bloomberg reported on Monday. This marks the latest high-ranking AI executive Meta CEO Mark Zuckerberg has scooped up to lead his new AI superintelligence unit.

Pang previously ran Apple’s in-house team that trained the AI foundation models that underpin Apple Intelligence and other on-device AI features, according to the report. Apple’s AI models haven’t exactly been a huge success — they’re far less capable than what OpenAI, Anthropic, and even Meta offer. Apple has reportedly even considered tapping third-party AI models to power its forthcoming AI-enabled Siri upgrade.

Sources told Bloomberg that Pang’s departure might be the first of many in Apple’s troubled AI unit. 

 

Pang was in charge of a team with approximately 100 employees that work on Apple's large language models. Models developed by Pang's team are used for Apple Intelligence features like email summaries, Priority Notifications, and Genmoji.

 

Meta lured Pang with a deal worth tens of millions of dollars per year. Meta has also scooped up AI experts from OpenAI, Anthropic, and Scale AI.

Rumors suggest that Apple has been considering relying on technology from Anthropic or OpenAI for a future LLM version of Siri, rather than using the models it has been working on. Those discussions have reportedly impacted the morale of the foundation models team that Pang is leaving. Other engineers on the team are also mulling offers from other companies.

 

Apple is struggling to catch up to competitors that have more advanced AI features, including Google and Samsung. Earlier this year, Apple delayed the ‌Apple Intelligence‌ ‌Siri‌ features that it had promised at WWDC 2024, and the delay led to a restructuring of its AI teams. Apple's AI efforts are now being overseen by Apple's software chief Craig Federighi and Mike Rockwell, who led Vision Pro development.

Nevertheless, Pang could bring expertise in designing small, on-device AI models to Meta, joining an array of talent Zuckerberg has poached in recent months, including leaders from Google DeepMind, OpenAI, and Safe Superintelligence.