Facts vs Feelings with Ryan Detrick & Sonu Varghese

Talking About AI with Steve Hou and Kai Wu (Ep. 178)

Mar 11, 2026

Guest

Kai Wu

Guest

Steve Hou

Kai Wu, founder and CIO at Sparkline Capital with a quant investing background. Steve Hou, a Bloomberg quant researcher formerly at AQR with a PhD in economics. They unpack the AI stack: chips, compute, models, and software. They discuss surging compute demand, model commoditization versus orchestration, energy and supply constraints, and which industries are most exposed to automation.

Ask episode

AI Snips

Chapters

Transcript

Episode notes

INSIGHT

AI Lowers Goods Prices But Not Immediate Service Inflation

AI is deflationary for manufactured and automatable goods but may not lower aggregate inflation quickly because sticky services (childcare, healthcare) resist automation.
Steve Hou stresses timeline and composition: goods fall fast, labor‑heavy services do not.

INSIGHT

Hidden Winners Are Industrial Data Center Suppliers

AI demand has created unexpected winners among low‑profile industrial suppliers that supply data centers (transformers, optical cables).
Steve Hou cites Siemens, Schneider and obscure cable makers booked out for years due to hyperscaler vetting.

INSIGHT

Enterprise Software Moats Are Distribution Not Code

Enterprise software moats are often distribution, sales, and service, not superior code, so incumbents can adopt AI and defend positions.
Kai Wu argues buyers pay for brand, contracts, and support, which limits instant displacement by vibe‑coding.

Get the Snipd Podcast app to discover more snips from this episode

Get the app

Artificial intelligence continues to dominate headlines and investment flows, but understanding the technology behind it requires looking beyond the hype and into the structure of the ecosystem itself. In this episode of Facts vs Feelings, Ryan Detrick, Chief Market Strategist at Carson Group, and Sonu Varghese, Chief Macro Strategist at Carson Group, sit down with Steve Hou of Bloomberg and Kai Wu of Sparkline Capital to explore how the AI economy actually works, from the infrastructure powering it to the applications beginning to reshape industries.

The conversation moves through the full AI stack, including semiconductors, computing power, models, and software layers, while also examining how competition, innovation, and investment are shaping the next phase of the technology cycle.

Key Takeaways

• The AI stack matters: Chips, infrastructure, models, and applications each play a distinct role in the ecosystem

• Compute demand keeps expanding: AI adoption continues to drive demand for semiconductors and data infrastructure

• Competition is accelerating: Innovation across companies may push AI models toward commoditization

• Productivity gains will vary: Some sectors may see faster AI-driven improvements than others

• Markets are pricing the shift: Investor expectations around AI continue shaping technology and equity markets

Steve Hou and Kai Wu are not affiliated with CWM, LLC. Opinions expressed by this individual may not be representative of CWM, LLC.

Jump to:

0:02 — Opening And Guest Intros

1:46 — Kai And Steve’s Quant Backgrounds

6:56 — Two ChatGPT Moments And AI Agents

10:45 — Compute Demand And Industrial Tailwinds

17:03 — Models Commoditize, Orchestration Rises

23:39 — AI, Inflation, And Energy As Constraint

31:17 — Europe, Korea, And Defense Capacity

38:02 — Software’s Reset And Duration Risk

46:30 — Timelines, Diffusion, And S-Curves

53:05 — Active Selection Across Regions

59:15 — Building Firms With AI Force Multipliers

1:03:49 — Mentors, Simplicity, And Implicit Knowledge

1:05:44 — Closing And Disclaimers

Connect with Ryan:

• LinkedIn: https://www.linkedin.com/in/ryandetrick/

• X: https://x.com/RyanDetrick

Connect with Sonu:

• LinkedIn: https://www.linkedin.com/in/sonu-varghese-phd/

• X: https://x.com/sonusvarghese?lang=en

Questions about the show? We’d love to hear from you! factsvsfeelings@carsongroup.com