Выпуск
Episode #116: As Soon As You Think You're Invulnerable: Jensen Huang and Nvidia's Board
Stewart Squared · 19 ноября 2026 г. · 58 мин
In this episode, host Stewart Alsop III and his father, Stewart Alsop II dive into the bombshell resignation of Nvidia CEO Jensen Huang amid controversy over workplace affairs, examining whether this represents a catastrophic governance failure or something more calculated. The discussion ranges from Nvidia's board composition—featuring some directors who've been there since the company's 1993 Sequoia and Sutter Hill funding round—to the strategic implications for the entire AI chip industry. They explore how Anthropic has deliberately positioned itself as Nvidia-independent through partnerships with Google TPUs and Amazon Trainium, potentially benefiting from Nvidia's current chaos as Anthropic heads into its IPO. The conversation also touches on the fundamental tensions between board oversight and founder-CEO power, the geopolitical dimensions of GPU access to China, and whether national security concerns played a role in Huang's ousting from the company he built into a $5 trillion powerhouse. Key Insights 1. Jensen Huang's removal from Nvidia represents a catastrophic governance failure at precisely the wrong time for the company. Despite building Nvidia from the ground up and positioning it perfectly for the AI revolution through strategic decisions like developing CUDA and aggressively pursuing the GPU market, the board forced him out over personal conduct issues. This appears to be a massive strategic mistake comparable to Apple firing Steve Jobs, as Huang was irreplaceable in his understanding of how to maintain Nvidia's dominance in an increasingly competitive landscape where companies like Anthropic are actively working to reduce dependence on Nvidia chips. 2. The Nvidia board demonstrated poor governance by maintaining the same directors for decades, including original venture capitalists who have been on the board for thirty years since the company's founding. This lack of board turnover and absence of proper succession planning created a situation where the board failed to address warning signs, including a New York Times article about Huang's affairs over a year ago. The nominating and corporate governance committee, which should have been monitoring these issues, was clearly not paying adequate attention despite the company's five trillion dollar valuation. 3. Anthropic has strategically positioned itself to be independent of Nvidia through a diversified chip strategy using Google TPUs, Amazon Trainium, and only limited Nvidia GPUs. Unlike OpenAI which built its own data centers dependent on Nvidia chips, Anthropic leases data centers and has developed the capability to train models across multiple chip architectures. This strategy makes Anthropic potentially the key actor that will change the semiconductor landscape and loosen Nvidia's control over the AI infrastructure market, which explains the massive six times oversubscription demand for their upcoming IPO. 4. The host suspects national security agencies were behind Huang's removal, pointing to Marco Rubio's role in publicizing allegations about one of Huang's affairs involving a potential Chinese spy. He believes both parties' national security wings allied to take Huang out. The guest finds that believable but says the deeper story is hubris: Huang had begun to think of himself as infallible. 5. Dario Amodei emerges as a strategic genius comparable to Elon Musk in business acumen, having built Anthropic with a clear strategy of maintaining independence from any single chip provider while developing frontier AI models. His approach contrasts sharply with OpenAI's dependence on Nvidia infrastructure, and his partnership structure with Amazon and Google provides both capital and chip diversity. The company's ability to reduce token costs and train models faster while maintaining quality leadership positions them as potentially the winning foundation model company, which explains why institution