Apple Poaches 36 Google AI Experts for On-Device AI Push

Key Takeaway

• Apple has recruited over 36 AI specialists from Google since 2018, aiming to strengthen its AI and Machine Learning team.
• The company's AI strategy focuses on deploying generative AI directly on mobile devices, diverging from cloud-based services.
• Despite challenges in AI deployment, Apple continues to innovate with the development of "MM1" models and a cautious rollout approach.

AI Talent Acquisition

Apple has been actively recruiting artificial intelligence experts, notably poaching at least 36 specialists from Google since 2018, when John Giannandrea was hired as its top AI executive. This hiring spree is part of Apple's broader strategy to bolster its global AI and Machine Learning team, which now includes a secretive research laboratory in Zurich, Switzerland. The Zurich-based "Vision Lab" emerged following Apple's acquisition of two local AI startups, FaceShift and Fashwell, and focuses on advanced AI models that integrate text and visual inputs. Despite its aggressive recruitment and expansion, Apple has maintained a low profile regarding its AI ambitions, contrasting with the publicized multibillion-dollar AI investments by rivals like Microsoft, Google, and Amazon.

Strategic AI Developments

Apple's AI strategy appears to be centered on deploying generative AI on mobile devices, enabling AI chatbots and apps to operate directly on the device's hardware and software. This approach could revolutionize how AI is integrated into consumer technology, moving away from cloud-based services. Apple CEO Tim Cook has emphasized the company's responsible investment and innovation in AI technologies, highlighting a long-standing commitment to AI, as evidenced by products like Siri and the company's early adoption of neural network technologies. The acquisition of Perceptual Machines in 2016 and the recruitment of notable AI researchers underscore Apple's dedication to developing foundational AI technologies.

AI Innovations and Challenges

Apple's AI team, now comprising former Google luminaries such as Samy Bengio and Ruoming Pang, has been instrumental in developing a family of AI models known as "MM1," which utilize both text and visual inputs. This development aligns with Apple's history of acquiring AI startups to enhance capabilities in image and video recognition, data processing, and content curation. However, the company faces challenges in AI deployment, including the need for more powerful chips with increased memory to support AI models on devices and the inherent risks of language models generating incorrect or problematic responses. Apple's cautious approach to AI rollout reflects a balance between innovation and control over new technologies.

Street Views

• Sumit Sadana, Micron Technology (Bullish on AI smartphones):

  "The next big thing is going to be ‘AI smartphones’ — and these will require a lot more Dram."

• Erik Woodring, Morgan Stanley (Bullish on the next iPhone):

  "What we’ll be looking for at WWDC are previews of one or two AI features that can become game changers for the average consumer."

Management Quotes

• Tim Cook, CEO of Apple:

  "Apple has been doing research across a wide range of AI technologies and investing and innovating responsibly around the new technology."

• Chuck Wooters, former Siri team member at Apple:

  "During the time that I was there, one of the pushes that was happening in the Siri group was to move to a neural architecture for speech recognition. Even back then, before large language models took off, they were huge advocates of neural networks."