Uncapped with Jack AltmanUncapped with Jack Altman

Inside the Mind of the Investor Who Backed Josh Kushner, Peter Thiel, and Marc Andreessen | Ep. 34

Jack Altman and Mel Williams on trueBridge’s Mel Williams on AI froth, signal, and conviction.

Mel WilliamsguestJack Altmanhost
Nov 25, 202540mWatch on YouTube ↗
AI cycle excitement vs early-stage frothEarly-stage vs growth-stage valuation dynamicsPower-law returns and larger winnersSignal, brand, and consensus investing debateFund size, scaling risk, and concentrationSeed exposure vs platform-fund limitationsLP diligence: luck vs skill and network edge
AI-generated summary based on the episode transcript.

In this episode of Uncapped with Jack Altman, featuring Mel Williams and Jack Altman, Inside the Mind of the Investor Who Backed Josh Kushner, Peter Thiel, and Marc Andreessen | Ep. 34 explores trueBridge’s Mel Williams on AI froth, signal, and conviction Mel Williams, co-founder of TrueBridge (a major venture fund-of-funds), shares a 2025 market read: deep excitement about AI’s decade-plus opportunity alongside froth, especially at the earliest stages.

At a glance

WHAT IT’S REALLY ABOUT

TrueBridge’s Mel Williams on AI froth, signal, and conviction

  1. Mel Williams, co-founder of TrueBridge (a major venture fund-of-funds), shares a 2025 market read: deep excitement about AI’s decade-plus opportunity alongside froth, especially at the earliest stages.
  2. He argues today’s power-law dynamics are intensifying—winners may be bigger than ever, even as many overcapitalized startups fail to reach product-market fit.
  3. Williams describes how elite “signal” firms and individuals (e.g., Sequoia, Founders Fund, Andreessen, Thiel, Kushner) attract capital, talent, and customers faster, creating self-reinforcing advantages.
  4. He also lays out TrueBridge’s manager-selection framework: prioritize contrarian/first-principles investors with the conviction to concentrate, be wary of rapid fund-size step-ups, and use deep networks to separate luck from skill.

IDEAS WORTH REMEMBERING

5 ideas

AI is both a long-term wave and a near-term frothy market.

Williams expects AI to drive opportunities for 10–15 years, but notes valuations feel especially inflated at formation/seed where evidence of product-market fit is thin.

Growth rounds look healthier than early rounds—multiples are not at peak extremes.

He contrasts “frothy” early pricing with later-stage rounds that resemble more defensible revenue multiples relative to public comps, implying risk is front-loaded in the earliest deals.

Expect simultaneous ‘carnage’ and unprecedented value creation.

Many heavily funded companies won’t find product-market fit, yet the category’s biggest winners can still create more total value than prior venture eras—classic dot-com pattern amplified.

Signal is a competitive weapon that compounds advantages.

Top brands create signal that accelerates follow-on financing, recruiting, customer acquisition, and even regulatory posture—pulling resources toward a small set of firms and companies.

Exceptional investors share two traits: contrarian/first-principles thinking and conviction to concentrate.

TrueBridge’s pattern from decades of data is that top funds invest differently early and then “push chips in” on winners—accepting portfolio concentration many can’t stomach.

WORDS WORTH SAVING

5 quotes

We’re gonna see a lot of carnage over the next ten years, and we will see more value created over the next ten years than we’ve seen in the venture industry previously.

Mel Williams

Ninety percent of the market is chasing the heat and chasing the signal, and ten percent of the market is the signal.

Mel Williams

The two characteristics... most important for exceptional investors... [are] contrarian or first principles... [and] conviction... willing to push their chips on the table.

Mel Williams

It’s really difficult for LPs to distinguish between luck and skill.

Mel Williams

You’re only as good as your network... My second piece of advice would be follow the signal. Try not to be the signal.

Mel Williams

QUESTIONS ANSWERED IN THIS EPISODE

5 questions

When you say early-stage valuations are frothier than growth, what specific ‘red flags’ (terms, check sizes, ownership targets) most often signal trouble?

Mel Williams, co-founder of TrueBridge (a major venture fund-of-funds), shares a 2025 market read: deep excitement about AI’s decade-plus opportunity alongside froth, especially at the earliest stages.

You mentioned AI is 50–60% of venture activity—how do you think that share changes if model capability or unit economics plateau for a few years?

He argues today’s power-law dynamics are intensifying—winners may be bigger than ever, even as many overcapitalized startups fail to reach product-market fit.

What’s your practical diligence process for separating luck from skill—who do you call, what questions do you ask, and what evidence changes your mind?

Williams describes how elite “signal” firms and individuals (e.g., Sequoia, Founders Fund, Andreessen, Thiel, Kushner) attract capital, talent, and customers faster, creating self-reinforcing advantages.

In your view, what makes an investor ‘the signal’ versus merely benefiting from signal—can you give concrete behaviors you’ve observed?

He also lays out TrueBridge’s manager-selection framework: prioritize contrarian/first-principles investors with the conviction to concentrate, be wary of rapid fund-size step-ups, and use deep networks to separate luck from skill.

You cited fund-size doubling as a danger zone; what operating changes must a firm make (decision-making, reserves policy, partner incentives) to scale successfully?

Chapter Breakdown

TrueBridge’s vantage point: Fund-of-funds, Midas data, and the job of picking investors

Jack sets context on Mel Williams and TrueBridge’s role as a venture fund-of-funds with deep ecosystem visibility, including early backing of top firms and data work supporting the Forbes Midas List. They frame the episode around how an LP picks venture managers versus how VCs pick companies.

2025 venture sentiment: Early AI wave, real adoption, but frothy pricing at the front end

Mel describes optimism driven by AI’s long-run opportunity horizon and evidence of adoption at scale. At the same time, he sees frothiness—especially at the earliest stages—where credibility-driven rounds can clear at high valuations before PMF is proven.

Valuation inversion: Early-stage hotter than growth, and why multiples look ‘healthier’ later

They discuss how some growth rounds are being priced at more reasonable revenue multiples compared to prior peaks, while early-stage rounds can be aggressively priced. Rapid AI-driven revenue growth complicates conventional multiple-based valuation frameworks.

Outside AI: More rational markets, attractive valuations, and milestone-based capital staging

Mel argues that non-AI categories look notably less frothy, with more disciplined step-ups tied to progress and fundamentals. While AI dominates activity, there remain compelling opportunities in other sectors at more reasonable entry points.

If things go wrong: PMF failures, capital overhang, and ‘carnage + massive value creation’

Mel’s core risk is that many heavily funded startups won’t reach product-market fit, leading to painful write-downs and failures. Yet he believes both outcomes can coexist: significant carnage alongside unprecedented value creation, echoing past cycles like dot-com.

Why winners are bigger now: lower marginal costs, faster adoption, and incumbents fully awake

They explore why winner-take-most dynamics may be intensifying: software scales cheaply, buyers adopt faster, and enterprises are proactively experimenting. Consumers also onboard quickly (e.g., ChatGPT), accelerating early traction and compounding advantage.

Talent and signal concentrate at giants: OpenAI/Anthropic/Meta as magnets vs startups

Jack notes recruiting has shifted: top talent now views leading AI labs and major tech as compelling even relative to hot startups. Mel ties this to the magnified value of ‘signal,’ which pulls capital, talent, and customers faster than before.

Platform snowballs vs emerging edges: Why TrueBridge likes both ends of the barbell

Mel explains TrueBridge’s barbell preference: premier platforms that can win across stages, and distinct individuals with proprietary early-stage access. He argues growth-stage venture can be attractive risk-adjusted return if done with the right firms, while early-stage requires unique angles to compete.

The long tail problem: brand advantages, legacy software drift, and who gets hurt

Mel expresses concern for the long tail of venture firms and for legacy software companies that can’t pivot to AI. As signaling becomes stronger, top brands gain structural advantages that may squeeze mid-tier firms and slow adapters.

Consensus vs contrarian: ‘10% are the signal’ and the rest chase it

Reacting to the idea that good investing is increasingly consensus investing, Mel argues it’s both: most capital chases heat, while a small set of exceptional firms/people create the signal. Those signal-creators can be both contrarian early and dominant in competitive rounds later.

Fund size debate: venture math vs techno-optimism, and what actually breaks funds

Mel reconciles Josh Kopelman’s ‘physics of venture math’ with Marc Andreessen’s view that winners will be enormous. Fund size matters, but the more predictive risk factor is rapid step-ups in fund size relative to firm capability and conviction, especially when concentration requirements jump.

What makes exceptional investors: first-principles contrarianism + conviction to concentrate

Mel identifies two standout traits of elite investors: contrarian/first-principles thinking and the willingness to bet big when evidence emerges. They discuss how top-performing funds often end with extreme NAV concentration in a few breakout winners.

Why seed still matters: platforms struggle at seed, and TrueBridge focuses on people over markets

Mel argues many platform firms historically struggle to invest effectively at seed due to signaling and downstream conflicts, making specialist seed managers valuable. In selecting seed managers, TrueBridge emphasizes people—track record, unique angle, proprietary access, and personal brand—rather than trying to pick markets.

Decision case studies: early bets that worked, and misses corrected later

Mel shares examples of non-obvious early commitments (Amplify Partners, Emergence) and the firm’s most consequential early relationship with Founders Fund. He also describes missing First Round’s first fund due to portfolio construction and passing on a16z fund I before investing in fund II once the platform thesis was clearer.

Why mediocre venture firms survive: fragmented LP base, luck vs skill confusion, and long feedback loops

They explain the durability of venture firms: capital comes from many sources, many LPs struggle to separate luck from skill, and performance evidence takes years. By the time results are clear, managers may have already raised multiple additional funds.

TrueBridge’s own concentration strategy: fewer managers, forced ranking, and capital-constrained discipline

Mel outlines TrueBridge’s evolution toward a more concentrated manager roster, shrinking core managers over time while increasing allocations to the best. They actively force-rank managers annually and remove ones displaced by higher-conviction allocations, better newcomers, or manager-specific issues like team changes or strategy drift.

Advice for aspiring LPs: build an authentic network and ‘follow the signal’ before trying to be it

Mel’s key advice is that LPs are only as good as their networks—relationships drive access, diligence, and insight. He recommends that newer LPs follow strong signal rather than trying to create it prematurely, since becoming signal requires long experience and pattern recognition; the cost of missing early funds is comparatively lower for LPs.

EVERY SPOKEN WORD

Install uListen for AI-powered chat & search across the full episode — Get Full Transcript

Get more out of YouTube videos.

High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.

Add to Chrome