Elon Musk vs Sam Altman | The Implosion of Thinking Machines | Can VC Survive Public Pricing?
Jason Lemkin and Rory O’Driscoll on 20VC debate whether VC can survive public-market pricing, dissect Thinking Machines’ collapse, and size up ClickHouse at $15B and Replit at $9B.
- Thinking Machines raised at $50B valuation but is imploding after multiple co-founder departures; Lemkin frames it as a seed deal gone wrong, not a catastrophe.
- Investors likely have redemption clauses triggered by team departures — returning ~$1.6B of $2B after 20% spend is the rational exit, not a long restructuring march.
- ClickHouse was doing ~$50M ARR a year ago; at $15B valuation the bet is near-100% AI market penetration and 2-3 more years of 3-4x growth to pencil out.
- Replit went from an unfinishable product at $2.5B to a product Lemkin calls 50x better; projected $700-900M ARR by end of year would justify the $9B round.
- Lemkin argues non-technical CEOs cannot run AI labs: without researcher-level credibility, top AI talent simply won’t stay regardless of compensation.
- OpenAI dropped vesting cliffs to poach talent from struggling labs — top AI researchers leave millions behind to work on problems they find intellectually compelling.
- Rory: public markets are sifting, not collapsing — Palantir trades at 70x forward sales at 45% growth while slow-growth SaaS gets crushed; venture must stay in hot sectors.
- Adobe acquired SEMrush (AEO/GEO play) rather than a smaller AEO startup because enterprise sales motion requires existing customer base, not frontier technology alone.
2026-01-22 · Watch on YouTube
