Uncapped with Jack AltmanInside the Mind of the Investor Who Backed Josh Kushner, Peter Thiel, and Marc Andreessen | Ep. 34
CHAPTERS
- 0:00 – 0:21
TrueBridge’s lens on exceptional investing: contrarian thinking + conviction
Mel Williams opens by framing what TrueBridge has learned from decades of venture data: the best investors are willing to think from first principles and act against consensus. Just as important, they double down aggressively when early evidence proves a bet is working.
- •Exceptional investors are contrarian/first-principles driven
- •Conviction is the differentiator: ‘push chips on the table’ when it works
- •Concentrating into winners is emotionally and professionally hard
- •These traits show up repeatedly in top-performing venture histories
- 0:21 – 2:09
2025 pulse check: AI opportunity meets frothy early-stage pricing
Jack asks for a sentiment read on venture in 2025. Mel is excited about an early AI wave that can drive 10–15 years of value creation, but he notes froth—especially in formation/early rounds where product-market fit is unproven.
- •AI wave feels early and durable over the next decade-plus
- •Overall excitement paired with caution about frothy conditions
- •Valuations feel particularly high at the earliest stages
- •Founder credibility is enabling large raises before strong evidence exists
- 2:09 – 3:39
Why early-stage can look pricier than growth: multiples, growth rates, and uncertainty
They unpack a counterintuitive dynamic: later-stage rounds sometimes look ‘healthier’ than seed/Series A. Rapid AI-driven revenue ramps make traditional multiples harder to apply, while early rounds price in a lot of future success despite limited proof.
- •Recent growth rounds can be below peak-style revenue multiples
- •Public comps and scale leaders complicate ‘right’ valuation benchmarks
- •AI companies can hit unprecedented revenue growth rates
- •Formation-stage pricing is high despite weaker PMF evidence
- 3:39 – 4:34
Outside AI: calmer markets and more milestone-driven capital staging
Mel contrasts AI with other sectors, where he sees more rational pricing and attractive entry points. In non-AI categories, capital tends to be staged against progress and revenue rather than hype-driven scarcity.
- •Market feels less frothy outside AI
- •Valuations and deal pacing look more ‘reasonable’ in non-AI sectors
- •Founders remain strong; good companies are still being built
- •Capital is staged in response to milestones and traction
- 4:34 – 5:33
If this goes wrong: PMF failures, ‘carnage,’ and simultaneous record value creation
Jack asks what a downside scenario looks like. Mel predicts many well-funded AI startups won’t achieve product-market fit, creating significant losses, while the same cycle still produces unprecedented category-defining winners—classic venture power law.
- •Big risk: heavily funded companies that never reach PMF
- •Both can be true: lots of failures + more value creation than ever
- •Dot-com analogy: wipeouts alongside Amazon/Google-scale outcomes
- •Power-law dynamics are intensifying in this cycle
- 5:33 – 6:18
Why winners may be even bigger now: software economics, faster adoption, and no ‘sleepy incumbents’
Mel explains why he expects outsized winners: software’s marginal cost is low, and buyers adopt faster than prior cycles. Enterprises and consumers are proactively experimenting with AI, accelerating growth curves and raising competitive stakes.
- •Lower marginal cost of software amplifies scaling potential
- •Enterprises are actively budgeting to try AI tools
- •Consumers adopt quickly (e.g., ChatGPT)
- •Incumbents are more alert, increasing competitive speed
- 6:18 – 11:41
Signal dominates 2025: talent aggregation and the flywheel of capital, customers, and hires
Jack observes that elite labs and leading tech firms are unusually compelling employers, changing startup recruiting dynamics. Mel argues ‘signal’ is magnified: strong brands attract talent, capital, and customers faster than ever, reinforcing power-law outcomes.
- •Top AI companies and big tech are pulling in startup-minded talent
- •Signal is ‘magnified’ in today’s market
- •Brand/signal accelerates capital raising, hiring, and customer acquisition
- •This flywheel strengthens dominant companies and firms
- 11:41 – 14:08
Bigger VC platforms: why brands can justify massive funds—and why founders accept lower price for signal
They debate whether the most powerful VC firms should manage much more money. Mel explains how brand signal tangibly helps founders (follow-on funding, talent, customers, regulatory credibility), and why lifecycle investors reduce fundraising burden over time.
- •Top firms can both ‘be the signal’ and win signal-created deals
- •Brand signal provides concrete founder advantages beyond capital
- •Founders may trade valuation for higher probability of becoming a winner
- •Lifecycle funds’ deep pockets reduce repeated fundraising cycles
- 14:08 – 16:51
The venture math debate: fund size matters, but capability-to-size fit matters more
Jack references opposing views: ‘venture physics’ vs. ‘companies will be enormous.’ Mel agrees with both: it’s harder to 10x huge funds, yet data shows top firms can be the largest and still return best—if scale matches strategy, access, and conviction.
- •Large funds face mathematical headwinds, but winners drive outcomes
- •Long-term data: largest fund can also be highest returning
- •Key evaluation: fund size relative to team, strategy, access, and conviction
- •Self-reinforcing cycles: performance → capital → talent → access
- 16:51 – 18:19
Where firms get hurt: rapid fund-size doubling and the difficulty of increasing check size and concentration
Mel cites research (Josh Lerner) showing the biggest risk is abrupt fund growth rather than absolute size. When a firm doubles, it must write much larger checks and concentrate more heavily—requiring a psychological and process shift that many can’t make quickly.
- •Risk factor: fund size increases that are ‘more than doubling’
- •Doubling forces larger average check sizes and bigger winner exposure
- •Conviction and governance must evolve to support concentration
- •Firms often struggle to adapt their decision-making cadence
- 18:19 – 20:46
What great investors share: first-principles contrarianism + doubling down on winners
They connect concentration back to investor quality, using Peter Thiel as an example of contrarian underwriting. Mel reiterates the two defining traits: original, non-consensus thinking and the ability to scale exposure when a bet proves out.
- •Exceptional investors invest when others won’t (contrarian entry)
- •First-principles underwriting creates ‘signal’ rather than chasing it
- •Conviction manifests as increased exposure over time
- •High concentration is uncomfortable but historically correlated with top funds
- 20:46 – 23:09
Why TrueBridge wants seed exposure: platforms historically struggle at seed, but seed returns can be accretive
Jack asks why bother with smaller or less obvious managers if platforms are so strong. Mel argues that big firms repeatedly struggle to do seed well (and seed can create downstream signaling issues), so specialized seed managers provide both exposure and returns.
- •Platform firms have a long history of inconsistent seed performance
- •Seed programs can create negative signaling for downstream rounds
- •Specialists focused solely on seed can outperform in that niche
- •TrueBridge seed returns have been additive to overall portfolio results
- 23:09 – 24:59
How TrueBridge picks emerging/seed managers: people over markets, unique angles, and personal brand
Mel explains their underwriting at seed: avoid ‘picking markets’ over 15-year horizons and instead back investors with judgment, differentiated access, and a repeatable right-to-win. For seed, personal brand matters because managers lack the ‘Sequoia card.’
- •Seed selection is more people-driven than market-driven
- •Look for track record evidence + repeatable decision quality
- •Unique angle/network that yields proprietary deal flow/right-to-win
- •Ability to build a personal brand is crucial at seed
- 24:59 – 30:12
Big wins and big misses: Founders Fund early, First Round passed, and joining a16z in Fund II
Mel shares concrete examples of difficult calls: backing Amplify and Emergence early, and the firm-defining bet on Founders Fund’s first institutional raise. He also recounts passing on First Round Fund I due to portfolio construction, and initially passing on a16z until the platform vision became real by Fund II.
- •Affirmative, non-obvious bets: Amplify (Sunil Dhaliwal) and Emergence (Jason Green)
- •Best unconventional decision: investing in Founders Fund early (Thiel/Howery/Parker)
- •A painful miss: passing on First Round Fund I due to concentration constraints
- •Passing on a16z Fund I, then investing in Fund II after platform proof
- 30:12 – 33:06
Why mediocre firms keep raising: brand durability, luck vs. skill, and slow feedback loops
Jack asks why the venture ‘long tail’ persists. Mel points to structural reasons: a huge, growing LP supply base; most LPs struggle to separate luck from skill; and long fund timelines allow multiple new funds to be raised before performance is fully known.
- •‘Hard to kill a good (or even regular) brand’ in venture
- •Highly diversified and expanding LP capital supply sustains firms
- •Most LPs can’t reliably distinguish luck from skill
- •10+ year feedback loops let weak performers raise additional funds
- 33:06 – 36:52
TrueBridge’s own concentration strategy: shrinking manager count, annual force-ranking, and exit criteria
Mel describes how TrueBridge has concentrated over time—from 18 core managers in Fund I to ~11–12 today. They force-rank annually, reallocate toward top managers, add new high-potential entrants, and remove managers for team change, strategy drift, or mismatched fund growth.
- •TrueBridge has systematically concentrated its manager portfolio
- •Annual force-ranking keeps bottom slots ‘at risk’
- •Primary reason to remove managers: reallocate to top-ranked relationships
- •Other exit triggers: team changes, strategy drift, fund-size/capability mismatch
- 36:52 – 40:33
Advice to young LPs: build your network, follow signal early, and earn the right to become signal
Mel closes with career guidance: network quality is the core advantage in a hard-to-verify asset class. He advises new LPs to follow signal rather than trying to be it immediately—becoming signal requires years of cycles, pattern recognition, and deep relationships.
- •LP performance is tightly tied to network depth and authenticity
- •Networks generate deal flow, insight, and better attribution of skill
- •Early-career advice: follow signal; being signal is exceptionally hard
- •Becoming signal takes time, cycles, and accumulated pattern recognition
