← Home
The Twenty Minute VC · July 25, 2026 · 1h 37s

20VC: Mercor CPO on Revenue Concentration from Frontier Labs | Why Large Enterprise is Scared to Partner with Frontier Labs | Why Small Specialised Models is the Future with Osvald Nitski

Osvald Nitski, Chief Product Officer at Mercor, discusses the AI data marketplace's explosive growth and the future of the industry. Topics include whether open-source models threaten data providers, why large enterprises remain hesitant to work directly with frontier labs, and the shift toward small, specialized AI models. The conversation also covers Mercor’s customer concentration risk, AI-generated code fueling a cybersecurity arms race, and the timeline for robotics’ mainstream breakthrough moment.

This summary was generated from show notes and public descriptions, not from a full transcript review. Details may contain inaccuracies.

Canon

Nitski discusses when robotics will achieve a breakthrough that transforms the field as dramatically as ChatGPT did for language.

Novel

Every Company Will End Up with Its Own Specialised AI Model00:09:00
Nitski explores the idea that enterprises will move toward small, specialized models rather than relying solely on massive general-purpose ones.

Highlights

Will Open-Source Models Kill Data-Provider Businesses?00:04:00
Osvald Nitski discusses whether the rise of open-source AI models threatens companies that sell proprietary data to train frontier models.
Why Enterprise Is Still Terrified of Working with Frontier Labs00:07:00
The conversation tackles why large enterprises remain hesitant to partner directly with frontier AI model companies.
Balancing AI Performance Against Exploding Token Bills00:11:00
The challenge for founders is optimizing AI model performance without letting token costs eat margins.
What It Now Takes to Be a Great Product Manager in an AI‑Native World00:15:00
Nitski outlines the evolving skills required for product managers in an AI‑first environment.
Mercor's Dependence on a Handful of Frontier-Model Customers00:37:00
The interview addresses the risk of Mercor’s revenue concentration from a small number of large AI labs.
AI-Generated Code and Agents Fuel a Cybersecurity Arms Race00:47:00
The conversation raises concerns that AI-generated code and autonomous agents are accelerating cyber threats.
Was this useful?