Gavin Baker - Nvidia v. Google, Scaling Laws, and the Economics of AI - [Invest Like the Best, EP.451] – Invest Like the Best with Patrick O'Shaughnessy
Gavin Baker - Nvidia v. Google, Scaling Laws, and the Economics of AI - [Invest Like the Best, EP.451]
Bullish and bearish opinions expressed in this episode, paired with supporting transcript quotes. The quote confirms what was said—not whether the opinion is correct.
Bullish
$NVDA— Nvidia's Blackwell chip is critically important to AI progress, with its technical complexity creating a significant competitive moat that ensures continued dominance despite product transition challenges.
$NATGAS— Natural gas is the only viable near-term power solution for AI data centers given nuclear's construction timeline constraints, with abundant U.S. supply from fracking ensuring long-term availability.
$GOOGL— Google has a temporary competitive advantage in AI from a pre-training perspective, positioning itself as the lowest cost producer of tokens through its TPU infrastructure.
$AMD— AMD's MI450 chip represents a significant competitive threat to Nvidia that will dramatically reduce AI inference costs, improving AMD's position in the AI chip market.
$SPACEX— Data centers in space represent the most important development in the next 3-4 years, offering 6x solar irradiance, free cooling, and faster networking, creating a massive future opportunity for SpaceX.
$AVGO— Broadcom captures massive economic value earning 50-55% gross margins on Google's $15 billion TPU production, though faces long-term risk if Google brings production in-house.
$TSLA— Tesla's Optimus robot will benefit from xAI providing the intelligence module while Tesla Vision serves as the perception system, creating synergies across Musk's companies.
Bearish
$SAAS— Traditional SaaS companies are guaranteed to fail in AI by refusing to operate at sub-35% gross margins like AI natives, repeating the same mistakes brick-and-mortar retailers made with e-commerce.
$META— Meta has completely failed to deliver on AI promises despite massive investment, with Zuckerberg being 'as wrong as it was possible to be' about having the best AI in 2025.
$QUANTUM— Public quantum computing stocks are in a bubble lacking fundamental support, with the real leaders being Google, IBM, and Honeywell rather than the publicly traded companies.