The Peel with Turner Novak The Past, Present, and Future of AI, Robotics, Venture Capital, Crypto | Why Deep Tech Startups Mess up Value Capture | Michael Dempsey, Managing Partner, Compound
53 snips
Aug 14, 2025 Michael Dempsey, Managing Partner at Compound, shares his insights on the evolving landscape of technology and investment. He discusses the intersection of AI and robotics, the importance of value capture for deep tech startups, and the challenges in navigating venture capital. Dempsey reflects on the investment journey with Runway and contrasts traditional investment strategies with those needed for AI startups. He emphasizes the need for adaptability in venture firms and critiques the authenticity crisis emerging in venture content.
AI Snips
Chapters
Transcript
Episode notes
Backing Runway From Early Research
- Michael Dempsey relays how he first DM'd Runway founders after seeing early generative model demos and backed their seed in 2018.
- Runway pivoted from general AI creative tools to focused video workflows and later coupled product with model development to scale rapidly.
Ship Many Fragmented Experiments Fast
- Ship many small features quickly to learn user demand rather than perfecting a single integrated product early.
- This approach lets you test use cases and collect training data that inform model improvements.
Build Thesis-Driven Research First
- Do deep primary research across academia and R&D to form prescriptive investment theses before writing checks.
- Use thesis-driven forecasting to invest in long-tail, scientifically risky categories.