According to Techmeme, at a Tokyo forum, SoftBank CEO Masayoshi Son said he “was crying” over selling his company’s $5.83 billion stake in Nvidia, a move he justified because he “just had more need for money to invest in OpenAI.” Separately, Apple announced that its AI boss, John Giannandrea, is retiring and will be replaced by Amar Subramanya, who most recently was a Vice President of AI at Microsoft and spent 16 years at Google before that. The press release framed it as “an exciting new chapter” for Apple’s AI ambitions. Industry observers, like Gene Munster, suggest this leadership change means the new Siri is likely still on track for a spring release, but also that Giannandrea was likely asked to retire as Apple pushes for more aggressive AI development.
The Real Story Behind Apple’s AI Shakeup
Here’s the thing: when a key executive “retires” right before the biggest product launch in their domain in a decade, it’s rarely just a peaceful farewell. The consensus from folks like Mark Gurman and Gene Munster is pretty clear: Giannandrea was probably pushed out. Why? Because Apple’s AI strategy, especially with Siri, has been seen as lagging for years. Bringing in a former Microsoft AI VP, Amar Subramanya, is a massive signal. It’s not just about getting fresh blood; it’s about getting someone from the epicenter of the current AI platform wars. Microsoft, with its deep OpenAI partnership, has been setting the pace. Apple is basically hiring from its perceived competition to catch up. The final line of that press release about “shaping the future of AI” isn’t corporate fluff—it’s a mission statement for Subramanya’s entire tenure.
Son’s Nvidia Regret Is a Lesson in Timing
Masayoshi Son crying over selling Nvidia is a stunning admission from one of tech’s most famous investors. He sold a $5.83 billion stake to fund bets like OpenAI. But look at the math now. That stake would be worth many, *many* times that today. It’s the ultimate “what if” story. But it also highlights the brutal trade-offs in venture capital and big tech investing. You need liquidity to swing for the fences on the next big thing, which for Son was AGI and OpenAI. So he sold a known, skyrocketing winner to fund a potential, world-changing future winner. The fact that he’s openly emotional about it tells you everything. He knew Nvidia was good, but even he couldn’t foresee just how stratospheric it would become. It’s a painful reminder that in tech, being right isn’t enough; you have to be right *and* have the conviction to hold on.
What This All Means for the AI Arms Race
So we have two parallel stories about resource allocation in the AI gold rush. Apple is reallocating its human capital, swapping out leadership to get more aggressive. SoftBank reallocated its financial capital, selling a golden goose to plant new seeds. Both actions are admissions that their previous positions weren’t optimal for the current moment. For Apple, the pressure is now squarely on Subramanya. The “new Siri” and Apple’s broader AI integration at WWDC aren’t just nice-to-haves; they’re existential for the iPhone’s relevance. As swyx and others note, the infrastructure for reliable, fast AI on-device is a monstrous challenge. Can a new boss quickly change a company’s ingrained culture and technical debt? That’s the billion-dollar question. Meanwhile, Son’s regret shows that even the savviest investors are struggling to navigate this market’s volatility. Everyone is scrambling, and the landscape is being rewritten by the week.
