IM 823: Intelligent with Two G's - Google Has a Fox Problem - podcast episode cover

IM 823: Intelligent with Two G's - Google Has a Fox Problem

Jun 12, 20253 hr 12 minEp. 823
--:--
--:--
Download Metacast podcast app
Listen to this episode in Metacast mobile app
Don't just listen to podcasts. Learn from them with transcripts, summaries, and chapters for every episode. Skim, search, and bookmark insights. Learn more

Episode description

  • Interview with Stephen Witt
  • Altman's Gentle Singularity
  • Sutskever video: start at 5:50-6:40
  • Paris on Apple Glass
  • OpenAI slams court order to save all ChatGPT logs, including deleted chats
  • Disney and Universal Sue A.I. Firm for Copyright Infringement
  • Apple paper: The Illusion of Thinking: Understanding the Strengths and Limitations of Reasoning Models via the Lens of Problem Complexity
  • Futurism on the paper
  • Could AI make a Scorsese movie? Demis Hassabis and Darren Aronofsky discuss
  • YouTube Loosens Rules Guiding the Moderation of Videos
  • Meta Is Creating a New A.I. Lab to Pursue 'Superintelligence'
  • Meta and Yandex are de-anonymizing Android users' web browsing identifiers
  • Amazon 'testing humanoid robots to deliver packages'
  • Google battling 'fox infestation' on roof of £1bn London office
  • 23andMe's Former CEO Pushes Purchase Price Nearly $50 Million Higher
  • Code to control vocal production with hands
  • Warner Bros. Discovery to split into two public companies by next year
  • Social media creators to overtake traditional media in ad revenue this year

Hosts: Leo Laporte, Jeff Jarvis, and Paris Martineau

Guest: Stephen Witt

Download or subscribe to Intelligent Machines at https://twit.tv/shows/intelligent-machines.

Join Club TWiT for Ad-Free Podcasts!
Support what you love and get ad-free shows, a members-only Discord, and behind-the-scenes access. Join today: https://twit.tv/clubtwit

Sponsors:

Transcript

Inside Jensen Huang's NVIDIA Empire Primary Navigation Podcasts Club Blog Subscribe Sponsors More… Tech Inside Jensen Huang's NVIDIA Empire

Jun 13th 2025

AI-created, human-edited.

In a revealing interview on the latest episode of Intelligent Machines, hosts Leo Laporte, Jeff Jarvis, and Paris Martineau sat down with investigative journalist Stephen Witt to discuss his explosive new book, "The Thinking Machine: Jensen Huang, NVIDIA and the World's Most Coveted Microchip." What emerged was a portrait of one of tech's most successful yet complex leaders - a man whose neurotic perfectionism and explosive temper helped build the world's most valuable company.

While Jensen Huang presents as the polished, charismatic CEO at NVIDIA's GTC conferences, Witt's deep-dive investigation reveals a far more complex figure. "Jensen has a terrible temper," Witt explained to the hosts, describing how the NVIDIA CEO has trained his staff not to interfere when he erupts at employees in front of large audiences.

The eruptions, as Witt experienced firsthand, are methodical and devastating. "It's humiliating," he recounted of his own encounter with Huang's wrath. "He questions your professionalism, he questions your ability... Why am I even participating in this if you're just gonna waste my time in this way?" The tirade lasted 20 minutes, with PR staff frozen in place, too intimidated to intervene.

Yet this same man suffers from severe stage fright. "God, I hate public speaking," Huang confided to Witt before a routine presentation, despite being one of tech's most compelling keynote speakers. This contradiction - a terrified perfectionist who has created a culture of fear while inspiring fierce loyalty - lies at the heart of NVIDIA's success story.

The hosts were particularly fascinated by NVIDIA's transformation from a gaming GPU company to the backbone of the AI revolution. As Leo Laporte noted, "This company was originally made GPUs for gaming... it sounds like they were making toys, right?"

Witt explained how this "toy" business was actually preparation for something much bigger. In the 1990s, NVIDIA was one of 70 companies fighting for survival in what he called a "Battle Royale" scenario. Huang's strategy was ruthless: "He would make lists of everybody who worked at 3DFX, everybody who worked at all his competitors... they would strategize on how to poach these people away... They called it brain extraction."

The real breakthrough came with CUDA, NVIDIA's platform that could flip graphics cards into scientific computing tools. Wall Street hated the investment - scientists had small budgets and weren't reliable customers. But Huang, influenced by Clayton Christensen's "Innovator's Dilemma," saw it as necessary disruption. "Sometimes it is right to ignore your customers and to ignore your investors," Witt quoted from Christensen's work.

Today's AI boom has turned those despised scientific computing investments into gold. Jeff Jarvis pressed Witt on the current market dynamics, learning that NVIDIA's H100 chips sell for $30,000 each, with companies like OpenAI ordering hundreds of thousands of them. "It's a $10 billion order, if not more, and it all just goes right into NVIDIA's pocket," Witt explained, noting NVIDIA's 75% gross margins.

The scale is staggering. The Texas supercomputer being built costs $50 billion - "more expensive than China's Three Gorges Dam and about 10 times as much as the James Webb Space Telescope." All powered primarily by NVIDIA chips.

Paris Martineau's questions revealed one of NVIDIA's most important strategic insights. While competitors can reverse-engineer chips, NVIDIA's real advantage lies in its software ecosystem and community engagement. "AMD can design the silicon just as well as we can," Witt quoted NVIDIA's perspective. Their competitive moat comes from embracing the scientific community and solving the world's hardest computing problems, often at a loss, to stay on the bleeding edge.

Perhaps most revealing was Witt's psychological portrait of Huang. The CEO's success stems from what Witt called living in "a mental torture chamber of his own divisement." When asked to throw a ceremonial first pitch at a baseball game in Taiwan, Huang warned the crowd he was terrible and asked them to look away. Then he spent six months having his wife pitch to him in their backyard until he could throw properly.

"He's a guy who actually lives in a mental torture chamber... constantly beating himself up for his perceived shortcomings and then creating programs for self-improvement," Witt observed. This neurotic perfectionism, combined with genuine terror of failure, drives both his personal excellence and his company's relentless innovation.

The interview also touched on NVIDIA's next big bet: edge computing and robotics. Huang envisions a world where AI brains live in every robot and autonomous vehicle, unable to rely on cloud computing due to latency requirements. NVIDIA's "Omniverse" platform aims to create high-fidelity physics simulations - digital twins of reality - where robots can train safely before deployment in the real world.

As Jeff Jarvis noted, it's "the Matrix, but we're not in it" - every permutation of possible futures being computed to train the next generation of AI systems.

The hosts' discussion with Witt illuminated how one man's psychological complexity and strategic vision created the infrastructure for our AI future. From gaming graphics to scientific computing to artificial intelligence, NVIDIA's journey reveals how technical innovation, strategic patience, and yes, even toxic leadership traits can combine to reshape entire industries.

As the interview concluded, it was clear that understanding Jensen Huang means understanding the uncomfortable truth that genius and dysfunction often go hand in hand in Silicon Valley's highest echelons. NVIDIA's success story isn't just about superior technology - it's about how fear, paranoia, and an obsessive drive for perfection can create the foundation for a technological revolution.

For anyone trying to understand how we arrived at the current AI moment, Witt's portrait of Jensen Huang and NVIDIA provides essential context for the human drama behind the headlines and stock prices.

"The Thinking Machine: Jensen Huang, NVIDIA and the World's Most Coveted Microchip" by Stephen Witt is available now. Listen to the full interview on the Intelligent Machines podcast.

Share: Copied! Intelligent Machines #823
Jun 11 2025 - Intelligent with Two G's
Google Has a Fox Problem All Tech posts Contact Advertise CC License Privacy Policy Ad Choices TOS Store Twitter Facebook Instgram YouTube Yes, like every site on the Internet, this site uses cookies. So now you know. Learn more Hide Home Schedule Subscribe Club TWiT About Club TWiT FAQ Access Account Members-Only Podcasts Update Payment Method Connect to Discord TWiT Blog Recent Posts Advertise Sponsors Store People About What is TWiT.tv Developer Program and API Tip jar Partners Social Contact Us
Transcript source: Provided by creator in RSS feed: download file
For the best experience, listen in Metacast app for iOS or Android