The VC's Lens: How AI Is Rewriting the Rules of Defensibility

Episode notes

Every company building AI right now is asking the same question: if the models keep getting better and anyone can access them, what actually makes us defensible? Avi Bharadwaj writes the checks that answer that question. As an Investment Director at Intel Capital, he focuses on the software infrastructure layer of AI, backing companies like Scale AI, Bria, TrueFoundry, and Twelve Labs. In this episode of Talking AI, Avi sits down with Matt Paige to break down exactly where moats are showing up as frontier models commoditize intelligence. He walks through five specific layers of defensibility for application companies (unique data, workflow and system of action, product reimagination, integration, and trust and compliance) and explains why the infrastructure between the model and the application is where most enterprise AI projects actually stall.

• More Agents Than Employees: How Zapier Disrupted Itself Before AI Could90 / 100
• 99% Correct Is Still Failure: The Last Mile for Mission-Critical AI62 / 100
• The State of AI 2026 Mid-Year Reality Check
• Context, Control, Collaboration: Why Capability Was Never the Bottleneck
• Past the Productivity Ceiling: Rebuilding the Enterprise from First Principles