The Twenty Minute VCThinking Machines Co-Founder Joins Meta for $3.5BN, Industry Venture's $665M Acquisition
CHAPTERS
Venture veterans’ cold open: wealth, incentives, and one-shot games
The episode opens with candid observations about how extreme money changes behavior and decision-making in tech and venture. The panel frames many choices—especially in today’s AI market—as option-like bets, and hints at the tension between loyalty and liquidity.
Pre-show banter: co-investing with Arthur Rock and feeling old in venture
Harry needles Rory about having co-invested with legendary VC Arthur Rock, prompting stories about Rock’s intensity and presence. The group riffs on aging in venture, career arcs, and the LP paradox of questioning whether successful managers are still “hungry.”
Goldman Sachs acquires Industry Ventures: what $665M buys
The panel breaks down Goldman’s acquisition of Industry Ventures (base ~$665M with earn-out potential) and why it makes sense for both sides. They emphasize Industry’s long grind building a scalable secondary/fund-of-funds business and Goldman’s need for higher-fee private assets.
Valuing asset managers: AUM multiples, revenue multiples, and real economics
Jason questions how a $7B AUM manager sells for under $1B, leading to a valuation lesson. Rory and Roger explain typical AUM-to-enterprise-value ranges and how fund-of-funds/secondaries economics differ from buyout platforms like KKR/Carlyle.
Distribution power: Goldman’s Apex and “productizing” private exposure
The conversation turns to why Goldman, specifically, is a strong acquirer. Roger highlights Goldman’s Apex platform for ultra-high-net-worth clients, and the group debates the quality of private market products that wirehouses pitch to individuals.
AI founder exits: Thinking Machines co-founder leaves after raising $2B for Meta package
The panel reacts to a high-profile AI talent move: after helping raise $2B at a ~$10B post valuation, a co-founder reportedly leaves for Meta with ~$3.5B in stock compensation. Jason frames it as a troubling cultural shift, while Rory emphasizes how extraordinary liquidity offers change the game.
Protecting teams and investors: vesting, cliffs, repurchase rights, and co-founder risk
Rory shifts from morality to mechanics: what founders and investors can do to reduce damage from key departures. The group discusses extended vesting, cliff vesting, and penalties for leaving—especially when the real asset is a small set of elite engineers.
SoftBank’s $5B margin loan on Arm to buy more OpenAI: Masa being Masa
SoftBank’s leverage move is framed as classic Masayoshi Son—high conviction, high risk tolerance. The panel assesses how leverable Arm is, whether margin calls are plausible, and why this is a comparatively “low-octane” Masa bet.
Data centers vs offices: is AI capex a bubble or an inevitable buildout?
Harry raises the striking comparison that more data centers are being built than offices, questioning whether this signals a bubble or a structural shift. Rory argues the technology trajectory is clear and scaling laws have held, but the ultimate constraint will be economic returns on capital.
Token hunger on the ground: vibe coding, AI agents, and demand that outruns supply
Jason offers operator-level evidence: building multiple apps quickly and deploying AI agents still feels compute-constrained, with massive unmet token demand. The group debates whether efficiency gains will reduce compute needs or simply accelerate software creation and increase consumption.
Where the alpha is in 2025: compressed “non-obvious” window and regulated-market nuance
The panel explores how the investment sweet spot is shrinking: companies can go from inception to “obvious” very quickly, forcing investors to choose between acute uncertainty or paying up. Roger argues for arenas with legal/regulatory complexity and slower diffusion curves, where speed-to-code is not the only moat.
Prediction markets (Polymarket vs Kalshi): regulatory arbitrage, funding, and “king-making” limits
A discussion on Polymarket and Kalshi frames them as regulatory arbitrage plays that are rapidly scaling—largely via sports-betting-like demand—under favorable political conditions. The group challenges whether VC capital can “pick a winner” here, concluding that users don’t care who funded the platform, only that it works and pays out.
Portfolio management deep dive: concentration vs diversification and follow-on discipline
The episode closes with a detailed debate on portfolio construction in an era where exit timelines are longer and growth-stage funds concentrate more. Roger outlines a model of moderate initial diversification (20–25 bets) followed by heavy concentration into a few breakouts, insisting every follow-on must stand alone on risk-adjusted expected value.
