Why $1B Exits are Dead
Summary
A discussion between a16z's David George and VenCap's David Clark reveals AI's profound impact on venture capital and the tech industry. They highlight that top 1% exit values have surged from \$10 billion in 2020 to \$32 billion in 2024, a 10x increase in 24 months. Frontier AI companies like Anthropic and OpenAI are adding revenue monthly at a pace comparable to Meta, Google, or Microsoft, despite less than 5% enterprise diffusion. This unprecedented scale, coupled with infrastructure constraints in compute and data centers, is forcing investors to re-evaluate assumptions about value capture and defensibility. The conversation also covers the shift towards native AI applications, the accelerating pace of value creation, and the growing challenge for VCs to identify durable winners in a rapidly evolving market.
Key takeaway
For venture capitalists evaluating AI investments, recognize that traditional metrics for scale and defensibility are rapidly shifting. Focus on companies demonstrating hyper-growth and those positioned within the "token path," while also preparing for higher early-stage loss ratios. Your firm must adapt to provide extensive platform support as AI startups face "big company problems" much earlier.
Key insights
AI is accelerating company scale and exit values, forcing venture capital to rethink traditional investment assumptions.
Principles
- Top 1% tech exits 10x'd to \$32B in 24 months.
- Frontier AI models drive unprecedented revenue growth.
- Supply constraints (compute, power) limit AI bubble risk.
In practice
- Prioritize companies in the "token path" for value capture.
- Anticipate proactive, native AI applications over skeuomorphic.
- Support lean, aggressive, native AI company structures.
Topics
- Venture Capital
- AI Startups
- Frontier AI Models
- Exit Valuations
- Data Center Constraints
- AI Ecosystem
Best for: CTO, VP of Engineering/Data, Executive, Investor, Entrepreneur, Director of AI/ML
Related on AIssential
Editorial summary, takeaway, and curation by AIssential. Original article published by The a16z Show.