TLDR
Bill Maris, founder of Section 32 and former Google Ventures CEO, shares four lessons from his career: glimpse the future through conviction, embrace apparent insanity, trust computer science over intuition, and keep funds small because they mathematically outperform large funds. In a panel discussion, he and other venture capitalists debate the broken incentives of mega-funds, the risk of private companies staying private too long, the coming compression in AI model pricing, and the shift of scientific talent to China. Maris argues that small, focused funds deliver superior returns and that AI infrastructure — not just larger models — will drive the next wave of innovation. The conversation also covers life sciences, deep tech, and the unfairness of forcing overpriced private shares onto public retirement accounts. The episode blends personal anecdotes with data-driven venture capital insights. The panel includes David Sacks and moderator Jason Calacanis. The discussion highlights the bimodal nature of venture returns and the need for incentive realignment. Maris emphasizes that 95% of top-decile venture performers are funds under $750 million. He also warns that Google could slash token prices, pressuring AI startups. The episode is a masterclass in venture strategy and macro trends. The tone is candid, contrarian, and deeply analytical. The insights are grounded in Maris's experience building Google Ventures and Section 32. The panel explores the tension between early-stage craft and late-stage asset gathering. The conversation ends with Sacks predicting a pendulum swing back to smaller funds. The episode is essential listening for founders and investors alike.
For you
PRO members can sign in and request more information — or their own version of this digest, tuned to what they care about.
Sign in to unlockSurprising Insights
Google initially prohibited using the term "AI" and forced "machine learning" because AI was considered science fiction 100 years away.
The U.S. is losing scientific talent to China due to funding cuts and anti-science policies, while China recruits top scientists from Europe and India.
Companies staying private longer forces overpriced shares onto public retirement accounts (401ks), transferring risk to ordinary investors who didn't participate early.
Venture returns are extremely bimodal: 75% of funds lose money, and a handful of winners must sell to public markets to realize gains, creating a potential bubble.
A $5B fund returning 1.01x makes more absolute profit for GPs than a $500M fund returning 3x, creating perverse incentives to raise ever-larger funds.
Key takeaways
Small venture funds (under $750M) mathematically outperform large funds due to focus, alignment, and exit math — 95% of top-decile performers are small funds.
Applying computer science and data-driven portfolio construction (simulations, backtesting) yields superior venture returns; don't bet against computer science.
Entrepreneurs who see the future often appear insane to others; conviction and willingness to take personal risk are prerequisites for breakthrough investing.
Venture capital incentives are broken: mega-funds profit from management fees even with mediocre returns, while entrepreneurs get inflated valuations that may not be sustainable.
AI development will follow the gaming industry trajectory: from brittle command-line interfaces to immersive ambient computing within five years, driven by infrastructure (physics engines, GPUs, controllers) not just larger models.
Notable quotes
“With a smaller fund, I have the advantage to be very selective in the companies that I invest in, the people that I hire.”
▸ Notable quotes · 0:12“To see the future, sometimes you need to be a little bit insane.”
▸ Notable quotes · 4:39“Don't bet against computer science.”
▸ Notable quotes · 7:56“Small funds outperform large funds. This is simply the math.”
▸ Notable quotes · 9:53Chapters
0:00 — Introduction and Bill Maris Background: Bill Maris is introduced as founder of Section 32 and former Google Ventures CEO; he outlines four lessons from his career starting from 1997.
1:13 — Lesson 1: Glimpsing the Future: Maris recounts quitting Wall Street after seeing a server in a closet, founding a web hosting company in his Vermont apartment, and tarrying a roof in a thunderstorm to protect his servers.
4:39 — Lesson 2: Insanity to See the Future: Using inauguration photos from 1989 and 1993, Maris illustrates how entrepreneurs who glimpse the future (like the man with a laptop in 1993) appear insane to contemporaries.
5:57 — Lesson 3: Don't Bet Against Computer Science: At Google Ventures, Maris used machine learning (then forbidden to call AI) to simulate portfolio construction and determine optimal fund size, achieving ~4.1x returns.
8:13 — Lesson 4: Small Funds Outperform Large Funds: Maris presents data showing funds under $750M average 4.76x DPI vs 1.42x for larger funds; he explains the exit math that makes large funds structurally disadvantaged.
10:53 — Panel Discussion: Venture Capital Strategies: Moderator and Maris debate barbell strategies, late-stage investing, and the ethics of companies staying private while claiming public benefit.
15:46 — AI and Gaming Analogy: Maris compares AI's evolution to gaming from Zork to photorealistic worlds, predicting ambient computing in five years driven by infrastructure, not just larger models.
17:17 — Life Sciences and Biotech: Discussion on computational biology, FDA constraints, and the flight of scientific talent to China due to U.S. funding cuts and immigration policies.
19:38 — Deep Tech and Incentive Structures: David Sacks joins to discuss fund size dynamics, broken GP/LP/entrepreneur incentives, and why the pendulum will swing back to smaller funds.
Speakers
Bill Maris — Founder of Section 32, former CEO of Google Ventures
Moderator (Jason Calacanis) — Host, interviewer
David Sacks — Venture capitalist, co-host