Latest / The CEO Diary with Fexingo: Leadership Lessons, Executive Decisions, and Corner Office Stories / How Masayoshi Son Built SoftBank on the Go Big or Go Home Bet
Transcript
- Lucas: So back in 2000, a Japanese telecom executive worth around six hundred million dollars meets a barely thirty-year-old Chinese entrepreneur in a cramped office above a secondhand furniture shop. The guy has a crazy pitch — his company has almost no revenue, but he wants to build the world's biggest marketplace. Luna: Wait — is this Masayoshi Son meeting Jack Ma? Lucas: Exactly. Son, who built SoftBank from a software distributor into a telecom and internet giant, sat through Ma's pitch for just about six minutes. Then he offered forty million dollars for a stake in Alibaba. Luna: Six minutes. That's either insane conviction or legendary luck. Lucas: Son would argue there's no difference. That bet eventually became worth around seventy-five billion dollars at its peak — a return so large it funded nearly every major SoftBank move for two decades. And the way Son talks about it, he'll tell you he saw the future in those six minutes. Luna: So how much of Son's success is pattern recognition versus just being willing to take risks others wouldn't? Lucas: That's the core question. And to answer it, we need to look at what he calls his 'grand vision' — the idea that technology will go through paradigm shifts every thirty years, and he wants to bet on every one. Luna: I've heard him say he's a 'vision capitalist' not a venture capitalist. What does that mean in practice? Lucas: It means he writes checks that are ten to a hundred times larger than a typical VC. He doesn't do seed rounds or series A's. He wants to be the dominant investor in companies he believes will define the next era — and he wants to put in enough capital that his bet alone can change the company's trajectory. Luna: Give me an example beyond Alibaba. Lucas: In 2016, SoftBank acquired ARM Holdings, the British chip designer, for thirty-two billion dollars. That was a huge premium. At the time, people thought he overpaid for a company that was quietly licensing designs. But Son saw that ARM's architecture would power every smartphone, every IoT device, eventually every AI chip. His thesis was that data and compute would explode, and ARM would be the toll booth. Luna: And that bet has aged pretty well with the AI boom, hasn't it? Lucas: Extremely well. ARM's IPO in 2023 valued it at over fifty billion, and it's climbed since. But Son didn't hold it through SoftBank — he sold a chunk in 2020 during a cash crunch. That's the volatility that comes with his style. He'll make a massive bet, then need to unwind it to stay alive. Luna: Which brings us to WeWork. Lucas: Right. The WeWork saga is the flip side of Son's approach. He poured over ten billion dollars into WeWork, convinced that co-working was going to reshape real estate. He pushed founder Adam Neumann to grow faster, spend more, and pursue an IPO at a forty-seven billion dollar valuation. Luna: And we all know how that ended. The IPO collapsed, Neumann was ousted, SoftBank had to rescue it at a fraction of the valuation. Lucas: Son admitted later that he had 'lost perspective' on WeWork. His conviction overwhelmed his analysis. He saw a charismatic founder, a narrative about changing how people work, and he doubled down repeatedly. It's a case study in how the same instincts that produce Alibaba can produce WeWork. Luna: So is there a method behind the apparent chaos? How does Son actually decide which bets to make? Lucas: He has a mental framework he calls 'time machine investing.' The idea is that developed markets show what developing markets will look like in ten or twenty years. So he looks at what's happening in the US or China and invests in similar trends elsewhere. But the real engine is the Vision Fund — the hundred billion dollar fund he raised in 2017, mostly from the Saudi Public Investment Fund and Mubadala. Luna: That fund changed venture capital. Suddenly every startup wanted a SoftBank check because it came with a network and a stamp of approval — but also with pressure to spend fast and think big. Lucas: Exactly. The Vision Fund's model was to write hundred-million to billion-dollar checks, take board seats, and push for rapid growth. The thesis was that in a winner take most tech economy, being the biggest investor in the potential winner is the only rational strategy. Luna: But the returns have been mixed. The Vision Fund had some big wins — Uber, DoorDash, ByteDance — but also significant losses in WeWork, OYO, and a bunch of failed delivery startups. Lucas: Son himself said the Vision Fund's portfolio was 'a mess' in 2022. He took a step back, slowed down new investments, and focused on preserving capital. But then in 2023 and 2024, he started talking about a 'next big thing' again — artificial general intelligence. Luna: I read he wants to raise a hundred billion for a new ai focused fund. Lucas: He's been pitching it as 'Project Nazca,' named after the mysterious giant drawings in Peru. His vision is that AGI will be the biggest opportunity in human history, and SoftBank needs to be the lead investor. It's pure Son — audacious, slightly mystical, and requires enormous capital. Luna: What do other tech leaders think of him? Is he respected or seen as a cowboy? Lucas: Both. People like Elon Musk have called him a 'legend' but also questioned his judgment. Jensen Huang from NVIDIA has a close relationship with Son — they've done several deals together. But many institutional investors are wary. The Vision Fund's track record makes them cautious about putting money into another mega-fund. Luna: So what's the leadership lesson from Son? Is it 'bet big or go home'? Lucas: I think the real lesson is about managing your own psychology. Son is unafraid to be wrong. He's said publicly that he expects seven out of ten of his bets to fail. But the three that succeed need to be massive. That's a hard mindset for most CEOs, who are trained to minimize failure, not embrace it. Luna: It does echo what we heard about Jensen Huang — the idea of 'the urgency of failure'. Lucas: Exactly. But where Jensen focuses on execution and product cycles, Son focuses on capital allocation and conviction. He's essentially a macro-bettor on technology trends, using SoftBank as his personal vehicle. And he's been doing it long enough that you can't dismiss it as luck. Luna: Before we go deeper into his AI bets, can I ask — what keeps these deep dive conversations going ad-free? Lucas: Honestly, it's listeners like you who chip in a couple of dollars a month. That's what keeps the shows independent and free of sponsor interruptions. Luna: Yeah, if you've gotten something out of episodes like this one, buy me a coffee dot com slash fexingo genuinely makes a difference. Lucas: So back to Son's AI play — he's been meeting with Sam Altman, with Demis Hassabis at DeepMind, with leaders across the chip ecosystem. He wants SoftBank to own the infrastructure layer for AGI. Think data centers, energy, chip design — all through massive equity stakes. Luna: That sounds like he's trying to build something akin to a national grid for AI, but privately owned. Lucas: That's exactly the vision. And it's why he's so fixated on ARM. He believes ARM's energy-efficient chip architecture will be the standard for AI inference, which is the process of running AI models after they're trained. If he's right, SoftBank's thirty-two billion dollar acquisition becomes the cornerstone of a trillion-dollar ecosystem. Luna: But the timing matters. AI investment is frothy right now. A lot of money is chasing a future that may not materialize as quickly as expected. Lucas: Son would say that's exactly when you should double down — when others are hesitant. His track record shows he's early more often than he's wrong, but being early can be painful. In the late 1990s, he invested heavily in internet companies before the dot-com crash. SoftBank lost nearly ninety-nine percent of its market value at one point. Luna: And yet he survived, rebuilt, and came back stronger. Lucas: That resilience is part of the story too. Son has a long-term time horizon. He often says he thinks in three-hundred-year plans. While that sounds grandiose, it gives him the psychological capacity to ride out crashes that would break most CEOs. Luna: So for a founder listening, is Son's approach replicable? Or is it unique to someone with his personality and resources? Lucas: I think the principle is replicable: be willing to take asymmetric risks where the upside is enormous and the downside is manageable. For Son, the downside was losing his fortune but not his ability to raise again. For a smaller founder, that might mean betting the company on one product pivot or one key hire. The magnitude scales, but the logic doesn't. Luna: And the other piece is having a clear thesis. Son doesn't invest randomly — he has a worldview about technology cycles, and every bet fits into that narrative. Lucas: Right. The narrative is what gives him conviction. Without it, you're just gambling. With it, you're making a calculated bet on what you believe the world will look like. Whether that's Alibaba in 2000, ARM in 2016, or AGI in 2026, the through line is consistent. Luna: It's fascinating to watch someone who's been written off multiple times keep coming back with an even bigger bet. Lucas: That's the corner office story that keeps me interested. Son isn't a perfect CEO — his oversight of WeWork was disastrous, and his personal investing style can create chaos inside SoftBank. But he's a reminder that in business, the biggest rewards often come from the most uncomfortable decisions. Luna: And that the same conviction that produces your biggest wins can also produce your biggest losses. Lucas: Exactly. The question every leader has to answer is whether they can live with that trade-off. Son clearly can, and that's what makes him one of the most fascinating CEOs of the last three decades.