Intersection of INDIE & Deep Tech — A Conversation with Michael Dempsey, Compound Managing Partner
My friend, Michael Dempsey, Managing Partner at Compound, and I have an ongoing thread about the intersection of indie ideals and his areas of focus in deep tech. Most of these conversations happen over meals or online, so we decided to dive in a bit with the cameras on.
Our conversation was as fun as it was wide ranging.
I’ve watched for years as Michael has built Compound into a research focused, thesis-driven firm that has moved from the edges to the center of some of the hottest investment areas today. We get into details behind his early conviction around AI juggernauts like RunwayML and Wayve, to new themes they’re exploring in crypto and biology.
Some quick takeaways and highlights to look for —
Our conversation was as fun as it was wide ranging.
I’ve watched for years as Michael has built Compound into a research focused, thesis-driven firm that has moved from the edges to the center of some of the hottest investment areas today. We get into details behind his early conviction around AI juggernauts like RunwayML and Wayve, to new themes they’re exploring in crypto and biology.
Some quick takeaways and highlights to look for —
- There’s a trend of deep tech startups pursuing overly ambitious “narrative rotations” by constantly expanding into new areas, which Michael aptly calls a “Ponzi scheme of ambition.” This is driven by the need to justify raising large amounts of venture capital.
- His bias towards building one big business in a massive market rather than narrative expansions. For shelling point investments, like SpaceX and Palantir, the focus is on owning and expanding a principle or market over time, rather than narrative rotations.
- Compound’s approach to investment themes that are considered “too early or too ridiculous” with no real customers yet, and funding those contrarian opportunities.
- The value of developing specific year-by-year theses on how certain sectors will play out, acknowledging that 80% may be wrong but aiming for the 20% that are very right. You only really need to be right twice per fund.