Invest Like the Best with Patrick O'Shaughnessy
Ben Thompson on Big Tech, China, and the AI Boom Running Out of Money - [Invest Like the Best, EP.487]
· Summarized
Key takeaways
- A balanced and realistic view of global AI leadership recognizes the importance of geopolitical stability, supply chain diversification, and capital availability.
- The history of infrastructure projects like railroads and shipping provides valuable lessons on the cyclical nature of tech and capital markets, especially regarding capacity and energy constraints.
- Major tech firms are adopting diverse strategies in AI, from leveraging ecosystems and content to building hardware and enterprise platforms, each aligning with their core strengths.
- The AI compute industry largely functions as a commodity, and supply shortages or excess capacity significantly influence margins, prices, and competitive advantage.
- Long-term success in AI will depend on owning customer access, deriving value from data and infrastructure, and managing geopolitical and regulatory risks.
Summary
The Future of AI: Strategic Opportunities and Risks
Ben Thompson discusses the geopolitical nuances of the AI race, emphasizing that a US victory could create global dangers due to overreliance on a single superpower. He stresses that dependencies on China for critical supply chains like TSMC are profound and unlikely to be fully replaced without conflict. Thompson warns against magical thinking about manufacturing independence and highlights the importance of realistic assessments, including the challenge of diversifying production and the interdependence of global supply chains. He advocates for more openness and innovation, cautioning against imitation of China's autocratic model, and envisions a long-term equilibrium where the US and China maintain competitive but balanced AI development.
The Capital Constraint and Historical Analogies
Thompson draws parallels between current tech cycles and historical infrastructure buildouts, such as railroads of the 1870s, illustrating how capital scarcity often hampers expansion despite technological potential. He discusses how energy and capital shortages could slow AI progress and emphasizes that despite advancements, funding shortfalls and the need for sustained investment remain critical barriers. Thompson underscores that, historically, long-term projects like railroads succeeded due to financing and expansion of energy capacity, hinting at similar patterns in AI infrastructure development.
Big Tech Companies and AI Strategy
Coverage of major firms reveals diverse strategies: Amazon leverages its ecosystem for internal and external growth, Microsoft builds middleware and enterprise platforms reminiscent of IBM, Meta pivots towards entertainment and social engagement, and Apple remains cautious, focusing on device excellence and on-device AI. Thompson discusses how each company's approach reflects their core competencies and market positioning, with some playing offense in AI development and others focusing on leveraging existing strengths. The discussion highlights the importance of owning customer access, building robust ecosystems, and balancing innovation with risk management.
Hardware, compute, and market dynamics in semiconductors
Thompson examines the semiconductor industry's boom-bust nature, emphasizing how capacity constraints, memory cycles, and geopolitical issues impact supply and demand. He discusses the conservative stance of TSMC and the dangers of overbuilding capacity, drawing lessons from history like the shipping industry. The narrative stresses that AI compute is a commodity, with significant implications for pricing, margins, and competitive dynamics. Thompson notes that cost and power shortages are delaying AI's full potential and warns that market risks will continue to shift as companies try to balance capacity investments with future demand.
The Rise of New Winners and Strategic Industry Moves
Thompson identifies Meta, Google, and Nvidia as frontrunners in AI, contrasting their strategies: Meta's dependence on content, Google's ecosystem strength, and Nvidia's hardware dominance. He discusses the playbook of building supply chains, leveraging internal data, and investing in infrastructure that enables external sales and partnerships. The conversation underscores that in AI, owning data, infrastructure, and customer relationships remains essential for long-term success, but also warns of risks associated with over-investment, regulation, and geopolitical friction.
Sponsors
- Ramp · Ramp makes managing company spend smarter by default, with better data, decisions, and economics over time.
- Vanta · Vanta continuously monitors your security posture and streamlines audits, helping companies win enterprise deals and build trust.
- WorkOS · WorkOS provides infrastructure for B2B and AI-native companies to sell to enterprise, covering security, governance, and more.
- Rogo · Rogo is the AI platform for finance, building agents for Wall Street that understand diligence, modeling, and analysis.
- Ridgeline · Ridgeline offers a modern operating system for investment managers, handling trading, compliance, and reporting in real-time.
Show notes
My guest today is Ben Thompson, the founder and author of Stratechery. Ben is one of my favorite business thinkers and I love talking to him about everything happening in markets and technology.
We go through every important company, including OpenAI, Nvidia, Intel, Apple, Microsoft, Google, and Amazon. We also discuss why he thinks it would be dangerous for the United States to win the AI race outright, what container shipping and the railroads of the 1870s tell us about the buildout, and why the binding constraint on all of this may be capital rather than compute.
Please enjoy my conversation with Ben Thompson.
For the full show notes, transcript, and links to mentioned content, check out the episode page here.
-----
Become a Colossus member to get our quarterly print magazine and private audio experience, including exclusive profiles and early access to select episodes. Subscribe at colossus.com/subscribe.
-----
Ramp’s mission is to help companies manage their spend in a way that reduces expenses and frees up time for teams to work on more valuable projects. Go to ramp.com/invest to sign up for free and get a $250 welcome bonus.
-----
Trusted by thousands of businesses, Vanta continuously monitors your security posture and streamlines audits so you can win enterprise deals and build customer trust without the traditional overhead. Invest Like the Best listeners get a special offer of $1,000 off Vanta when you go to vanta.com/invest.
-----
WorkOS is the infrastructure B2B and AI-native companies use to sell to enterprise. It covers everything enterprise security requires: SSO, SCIM, RBAC, Audit Logs, AI governance, and more. Trusted by 2,000+ fast-growing companies, including OpenAI, Anthropic, Cursor, and Vercel.
-----
Rogo is the AI platform for finance. They're building agents for Wall Street that are trained to understand how bankers and investors actually do work: from diligence and modeling, to turning analysis into deliverables. To learn more, visit rogo.ai/invest.
-----
Ridgeline has built a complete, real-time, modern operating system for investment managers. It handles trading, portfolio management, compliance, customer reporting, and much more through an all-in-one real-time cloud platform. Visit ridgeline.ai.
-----
Editing and post-production work for this episode was provided by The Podcast Consultant.
Timestamps:
(00:00:00) Welcome to Invest Like The Best
(00:02:16) Winning the AI Race With China
(00:08:28) Timing, Capital, and the Railroads
(00:11:34) Berkshire, Google, and Absolute Profits
(00:14:23) Verifiable and Unverifiable Domains
(00:20:20) Aggregation Theory in the AI Era
(00:22:06) The Real Cost of Inference
(00:25:40) Why Consumer AI Needs Advertising
(00:30:08) Compute Shortages and Commodity Markets
(00:35:46) Memory Cycles and Boom Bust Dynamics
(00:42:08) TSMC, Intel, and Where Risk Goes
(00:44:51) The Best Setups in Big Tech
(00:52:14) The Frontier Model Contenders
(00:54:27) Microsoft's IBM Playbook
(01:00:45) Meta, Attention, and Advertising
(01:07:29) NVIDIA, Commodities, and Power
Follow this show on AudioBrevity and new episodes get summarized for you. Create a free account