How LP Deployment to Venture Will Change in 2025 & Are Endowment Funds in Trouble?

Rory O’Driscoll (Scale) and Jason Lemkin debate the $3B Windsurf/OpenAI deal, LP allocation stress, and why a guaranteed 5x should always get funded.

• OpenAI’s Windsurf acquisition (~$3B, ~1% of market cap) is framed as an SVP-level bet to close coding gap vs. Anthropic, not a bet-the-farm move.
• Windsurf was a Chrome plugin called Kodium 90 days before becoming a $3B acquisition target — fastest PMF-to-exit arc either investor had seen.
• Multi-stage funds (e.g., GreenOaks led Windsurf seed and doubled down at A) are systematically squeezing seed managers’ ownership from ~15% to ~10% while seed fund sizes have doubled.
• A $500-600M seed fund return on Windsurf is less than one-third of the fund needed for 5x — illustrating how brutal venture math has become even on headline wins.
• Ivy League endowments facing the denominator effect should dramatically revise cash planning for 2025, per O’Driscoll, as venture distributions slow and allocations look distorted.
• Jason Lemkin’s investment rule: if you have 100% conviction of a 5x, do the deal regardless of valuation or ownership — the regret of missing a sure thing outweighs sizing discipline.
• AI products at $20/month (Windsurf, Highfield) vs. $20K+ entry cost for legacy enterprise tools is driving 100% of early adopters into market simultaneously — explaining anomalous growth rates.
• Lemkin uses an AI agent with trigger rules (growth metrics thresholds) to surface deals automatically, calling it better than a human associate for pre-screening.

2025-04-24 · Watch on YouTube