Apple’s AI Struggle: Can He Save It? 🍎🔥

Tech & Science

Amar Subramanya is poised to become one of Apple’s most closely watched hires in recent years, joining the company just over six months after stepping into the role of Corporate Vice President of AI at Microsoft’s newly formed AI division. Prior to this, Subramanya spent sixteen years at Google, starting as a Staff Research Scientist and progressing to Principal Engineer before being appointed Vice President of Engineering in 2019. Notably, he contributed to two highly publicized releases: Gemini, which debuted in December 2023, and Imagen 3, released in August 2024. Following the launch of the latest Gemini model, OpenAI CEO Sam Altman cautioned his team to anticipate “rough vibes” and “temporary economic headwinds,” effectively acknowledging that the company’s advancements were being closely matched by OpenAI and that investor sentiment reflected this reality. Google’s pursuit of AI capabilities, while yielding a model approaching OpenAI’s intelligence level, wasn’t without its challenges. Recall Google’s initial bet on Bard; it took nearly three years— coinciding with ChatGPT’s release— to implement a significant course correction. While it remains too early to definitively determine whether Google will surpass OpenAI, the key takeaway is that the company now possesses a state-of-the-art AI model, coupled with the reach of all its Google-owned platforms for deployment.

Subramanya faces significant pressure from day one to address the perceived failures of his predecessor. Despite this uphill battle – compounded by a team operating at an all-time low morale – it is reasonable to expect he will need to bring in reinforcements swiftly, if he hopes to recover from Apple’s recent, relentless string of losses, particularly those sustained over the past year. Earlier today, several employees made the difficult decision to leave the company, and some may even be former colleagues of Subramanya who could prove valuable if Apple utilizes Gemini as the foundation for its revamped Siri. Furthermore, there’s an external expectation, though not directly his responsibility, that he will primarily focus on resolving Apple’s underlying AI technology. Ultimately, the success of that effort – and how it’s presented to the public – will depend entirely on his ability to carefully balance the need for rapid progress with Apple’s ambitious AI vision, potentially requiring the team to rebuild significant portions of its work. Looking ahead, rebuilding confidence – both internally and externally – will constitute the bulk of his responsibilities, but this will undoubtedly be just the initial step in a longer-term strategy.

For some time, it was commonly assumed that Apple’s involvement in machine learning was primarily focused on playing catch-up to its competitors. However, this perception doesn’t fully capture the essential function of any company striving for significant advancements in artificial intelligence: to cultivate an ecosystem where research and practical application can mutually reinforce each other. Ironically, this was precisely the goal John Giannandrea set forth before ChatGPT dramatically reshaped the entire industry.