20VC: Nebius Co-Founder on AI Infrastructure Bubbles | The Real Impact of Open Source on OpenAI & Anthropic | How Price Elastic is Demand for Compute | Could Nebius Sell 10x More Compute If They Had It & more with Roman Chernin
June 8, 2026
AI Summary
5 min readThe AI Infrastructure Race Is Just Beginning
When DeepSeek released a cheaper, more efficient model in early 2025, Nebius stock dropped 40% in a week. That same week, the company had its best sales week ever. "People on the market were concerned that market is going down and infrastructure companies like Nebius not needed because if AI is so much cheaper, maybe it's a bubble," recalls Roman Chernin, co-founder of the $66 billion AI infrastructure company. "But at the same time, we never had the best commercial week in the history of the company because so many people figured out that they can run inference in their production workloads with DeepSeek and economics will work." That paradox — cheaper intelligence driving more consumption, not less — is the central dynamic Chernin sees shaping the next decade of AI infrastructure.
The Bubble Question and the Real Demand Curve
Continue reading the full summary in the app — free to try.
Read Full Summary →Free • No credit card required
Never miss an episode of The Twenty Minute VC (20VC): Venture Capital | Startup Funding | The Pitch
Get every new episode summarized in your inbox — free, ~5 minutes to read.
No spam. Unsubscribe anytime.
What you'll learn
- 1 (04:20) **Episode Introduction & AI Infrastructure Bubble Debate** - Roman Chernin, co-founder of Nebius, joins to discuss the AI infrastructure landscape, the company's $66B market cap, and Leo Ashenbrenner's large position.
- 2 (04:58) **Is AI Infrastructure a Bubble?** - Roman argues we are at the very beginning, not a bubble, citing limited enterprise adoption and only one proven use case (coding) at scale.
- 3 (07:12) **The Open Source Threat to Frontier Model Providers** - Roman explains that the shift to open source is already happening, but it doesn't hurt OpenAI/Anthropic because they push the frontier while open source eats behind them.
- 4 (12:37) **Where Nebius Wants to Move Faster: The Four Layers of Product** - Roman outlines the four dimensions of building Nebius: capacity, product, customer success, and capital.
- 5 (19:07) **Price Elasticity of Compute Demand** - Roman confirms that if Nebius had 10x more capacity, they could sell it, but the key is building a diversified customer portfolio across all four product layers.
- 6 (23:13) **Price Elasticity of Demand** - Roman reveals they recently raised prices and still have pipeline pressure, but notes that for inference, there is a limit where customer economics break.
- 7 (26:04) **The Product Layer: Following Customer Needs** - The main question is how customers evolve, from training to inference, from labs to enterprises, and from using models to building agents.
+ Full timestamped outline available in the app
Show Notes
Roman Chernin is Co-Founder and Chief Business Officer of Nebius, one of the fastest-growing AI infrastructure companies in the world. Today, Nebius operates some of the largest AI compute clusters globally and serves leading AI labs, enterprises, and developers. Today, Nebius has a market cap of $57BN.
AGENDA:
00:00 — Why AI Infrastructure Is Not a Bubble
05:00 — The Real Impact of Open Source on OpenAI & Anthropic
11:00 — Jevons Paradox: Why Cheaper AI Creates More Demand
13:00 — The Four Layers of AI Infrastructure Explained
19:00 — If Nebius Had 10x More Capacity Tomorrow
26:00 — The Shift from Training to Inference and Agents
31:00 — How Token Factory Cuts AI Costs by 70%
44:00 — Sovereign AI, Europe, and the Future of Model Building
49:00 — Competing Against Hyperscalers with 10x More Capital
59:00 — The Biggest Threat to Nebius Isn't Competition—It's Consolidation
More from this podcast
The Twenty Minute VC (20VC): Venture Capital | Startup Funding | The Pitch →