The Twenty Minute VCThinking Machines Co-Founder Joins Meta for $3.5BN, Industry Venture's $665M Acquisition
At a glance
WHAT IT’S REALLY ABOUT
Venture news dissected: secondaries M&A, AI talent wars, and portfolio strategy
- Goldman Sachs’ $665M acquisition of Industry Ventures is framed as a rational “asset manager” deal valued as a percentage of AUM and enabled by Goldman’s distribution to private-wealth clients.
- A Thinking Machines co-founder leaving after raising $2B to join Meta for $3.5B sparks a debate on founder responsibility versus one-shot incentives, and what governance/vesting protections investors and co-founders should implement.
- SoftBank’s reported $5B margin loan against Arm to buy more OpenAI is treated as classic Masa-style leverage, raising the broader question of where AI infrastructure capital will come from and under what return expectations.
- The panel argues AI data-center buildout is driven by validated scaling laws and insatiable demand, but the ultimate constraint may be economic returns rather than technical feasibility or usage.
- They contrast concentrated versus diversified AI investing and go deep on portfolio management: start diversified, then concentrate via follow-ons using new information, while treating each check as an option with risk-adjusted expected value.
IDEAS WORTH REMEMBERING
5 ideasIndustry Ventures’ sale price fits standard asset-management valuation logic.
They interpret the deal as ~10% of $7B AUM (with earn-out) and roughly “10x revenue / ~20x earnings” for a high-margin, brand-driven platform, comparable to StepStone/Hamilton Lane-style models.
Some investment businesses are sellable; most venture partnerships are not.
Productized, fee-stream platforms (secondaries/fund-of-funds) can be acquired wholesale, while small venture firms are “people businesses” where selling out the partners destroys the underlying asset.
Mega personal payouts can turn careers into a one-turn prisoner’s dilemma.
When an individual is offered more liquid value than their startup equity’s risk-adjusted value, the multi-period reputational game collapses and “bad behavior” becomes more likely, forcing investors to price that risk rather than moralize it away.
Co-founder departure risk should be addressed explicitly in formation docs.
They highlight longer vesting (e.g., six-year vs four-year), cliffs, repurchase rights, and penalties for leaving for competitors as increasingly important when teams are the core asset and poaching is plausible.
AI infrastructure buildout is unlikely to be limited by demand—only by returns.
The group accepts that token demand will expand to fill supply (efficiency gains just increase usage), but argues capital markets will slow spending if marginal ROI fails, even if scaling laws continue to hold.
WORDS WORTH SAVING
5 quotesIn the face of unprecedented wealth, I'm shocked to discover that most people behave badly.
— Rory O’Driscoll
Something's broken in the way that we're evolving as humans if everything ultimately reduces to what's in it for me.
— Jason Lemkin
Once you're not playing a multi-period game, and when someone offer you $3.5 billion, you're no longer playing a multi-period game. You're playing a one and done, right? You're gonna get bad human behavior.
— Rory O’Driscoll
Polymarket was founded by a solo founder in his toilet during lockdown.
— Jason Lemkin
Everything in life you can price as an option.
— Rory O’Driscoll
High quality AI-generated summary created from speaker-labeled transcript.
