Interview, Fireside Chat
Who Wins in the Next Decade of VC | Margin and Price Matter Less in a World of AI | Sameer Dholakia
- Intelligence is predicted to surpass the oil market in size and power all human sectors, driving a trillion-dollar super cycle characterized as a "magic" moment larger than the combined Internet and Cloud eras.
- The AI market is forecasted to consolidate around a handful of capital-intensive frontier providers, with hundreds of application apps building on this infrastructure while both open and closed token models are expected to generate massive business scales.
- Specific company forecasts include Legora reaching $5 to $10 billion in revenue from the legal sector, Agora growing to $200 million ARR from $20 million, Fireworks becoming a $250 billion business, and FieldGuide potentially becoming a $10 to $20 billion enterprise.
- Valuation expectations include OpenAI growing from $40 million to approximately $2 billion before investment, while specific capital allocations involve nearly $200 million in Lease AI and potential 20% fund stakes in generational companies like Anthropic.
- Investment strategies anticipate higher seed checks beyond the historical $3 to $5 million range, with seed funds potentially failing due to inability to afford necessary shots on goal in a high-priced market.
- The venture capital industry is expected to consolidate into giant platforms or extreme specialists, with firms targeting 5x fund returns by investing in de-risked late-stage companies that compound at hyper-rates.
- Market dynamics suggest companies will grow at 10 to 15 times NTM revenue multiples, blow through entry valuations, and maintain private status longer due to the private asset class size, with a potential IPO market unlock in 2028.
- Future timelines indicate robotics and embodied AI are currently circa 2021 with rapid acceleration expected in a few years, while AI is predicted to accelerate drug discovery in the next decade to a pace previously requiring half a century.
- Sovereign AI models are expected to emerge as large businesses for governments and regulated industries, alongside enterprise trends of owning intelligence rather than renting it, with specific open-source US-based models potentially reaching $500 billion via talent monetization.
- Risks include the capital intensity creating a self-fulfilling moat where only those with the biggest war chests can compete for talent, though capital alone is deemed insufficient without great products, and the difficulty of finding early-stage returns necessitating 50x fund returns to offset lower hit rates.