Bill Maris: How Google Could Crush AI Competitors, Why Small Funds Win, and AI's Atari Stage
Episode Details
In a comprehensive episode of the All-In Podcast, guest Bill Maris joins co-hosts David Friedberg and David Sacks to unpack his career and investment philosophies. Bill Maris shares how he started by running Data Centers from his apartment, a move that seemed unconventional but captured his vision of the future—a concept he illustrated using historical photos provided by Stewart Butterfield. Maris later collaborated with Rich Miner, co-founder of Android, to launch Google Ventures under the umbrella of Google. By relying heavily on data and Machine Learning—the approved corporate nomenclature for early AI—he developed a highly successful portfolio strategy, earning accolades in TechCrunch. He proved that applying Computer Science yields major results, evident in GV’s investment in Climate Corp, founded by David Friedberg, which achieved a massive exit to Monsanto. Transitioning to modern Venture Capital, Bill Maris explains his rationale for creating Section 32. He argues that giant entities like Andreessen Horowitz and Founders Fund distort the market. These mega-funds require colossal M&A and IPOs to generate meaningful DPI, ultimately pressuring the Startup Founder with inflated valuations. In contrast, Section 32 deliberately maintains smaller fund sizes, successfully backing companies like CrowdStrike, Cohere, and Coinbase. On the topic of artificial intelligence, Maris warns that Google could deploy Capital as a weapon by dramatically discounting its Gemini models, thereby exerting fatal margin pressure on competitors like OpenAI and anthropic. Comparing the current state of AI to the primitive game Zork, he asserts the industry is in its infancy but will rapidly achieve seamless Ambient Computing, fueled by underlying hardware like GPUs. Lastly, the discussion delves into the challenges and opportunities of Deep Tech. As the founder of Calico and an investor in Flatiron, Maris focuses extensively on Computational Biology. He points to investments in New Limit, a company spearheaded by Blake Byers and Brian Armstrong. However, he highlights a massive hurdle in the US: extreme regulatory burdens enforced by the FDA, alongside the political gutting of the CDC and NIH. Coupled with restrictive H-1B visa policies, this hostile climate is causing a brain drain. Promising researchers from places like India are increasingly migrating to China, where biotech research proceeds rapidly. Despite these friction points, Maris remains optimistic about Deep Tech, lauding Elon Musk for successfully overcoming massive engineering challenges at SpaceX and Tesla.
The episode explores the shifting landscape of venture capital, highlighting the performance advantages of smaller, selective funds over mega-funds. It also examines the future of AI through the lens of 'ambient computing' and the potential for large incumbents to use capital as a weapon to disrupt current AI startups.
Portfolio lens: A cautious, data-driven approach favoring infrastructure and computational efficiency while hedging against the risks of capital-intensive, model-centric AI startups.
Generated with gemini-flash-lite-latest on 6/26/2026, 1:22:21 AM. For research only. Not financial advice.AI Infrastructure and Enabling Technologies
Focus investment on the 'picks and shovels' of the AI revolution—physics engines, controllers, and hardware—rather than the large language models themselves.
The guest argues that the current AI landscape is in an 'Atari command line' stage and that long-term value will accrue to the underlying platforms that make ambient computing possible, similar to how gaming evolved through hardware advancements.
- Comparison of AI's current state to early gaming (Zork) and the expectation of rapid evolution to 'ambient computing' within five years.
- Focus on physics engines, controllers, and GPUs as the essential components for the next phase of AI.
- Advancements in silicon and specialized AI hardware.
- Development of more sophisticated physics engines for simulation.
- Rapid obsolescence of hardware standards.
- Potential for large incumbents to vertically integrate and commoditize these enabling layers.
- Analyze the capex spending of major cloud providers on non-compute AI infrastructure.
- Evaluate the competitive moat of specialized controller and physics engine software providers.
AI Model Margin Compression
Large incumbents with massive capital reserves may aggressively discount AI tokens to capture market share, creating significant margin pressure on pure-play AI model startups.
The guest suggests that Google could use 'capital as a weapon' to offer Gemini tokens at a fraction of the cost of competitors, potentially rendering current business models of AI startups unsustainable.
- The guest explicitly mentions the possibility of Google discounting token prices to 20 cents on the dollar to exert fatal pressure on competitors.
- The observation that current AI startups are burning investor cash to grab market share.
- Aggressive price-cutting announcements from major cloud AI providers.
- Quarterly earnings reports showing high spend commitments vs. revenue for AI-heavy companies.
- Antitrust regulatory intervention preventing predatory pricing.
- Startups successfully differentiating via proprietary data or specialized workflows that are not easily commoditized.
- Monitor token pricing trends across major AI API providers.
- Assess the burn rates and runway of leading independent AI model labs.
Computational Biology and In Silico Research
Computational biology holds the potential to bypass traditional, slow clinical trial processes, provided that reliable in silico simulations of human cells can be achieved.
While traditional biotech is hampered by regulatory burdens and slow clinical trials, computational approaches offer a path to accelerate discovery and reduce costs.
- The guest highlights the shift from fringe longevity science to mainstream interest in computational biology.
- The mention of companies like New Limit as examples of the new wave of biotech investment.
- Breakthroughs in realistic human cell simulation.
- Increased adoption of AI-driven drug discovery platforms.
- Extreme regulatory hurdles enforced by the FDA.
- The 'anti-science' climate and brain drain in the US, potentially slowing domestic innovation compared to China.
- Track the progress of in silico cell modeling research.
- Evaluate the regulatory landscape for AI-assisted clinical trial designs.
Watchlist
- Token pricing for Gemini, GPT-4, and Claude
- H-1B visa policy changes
- FDA regulatory approvals for AI-driven biotech
- Public market performance of recent 'Deep Tech' IPOs
Open Questions
- How will the public market react when the current crop of private AI unicorns finally go public?
- Can the US reverse the 'brain drain' of top scientific talent to China and India?
- Is the 'barbell' strategy of combining small venture funds with large growth vehicles mathematically viable over the long term?
Key Topics & People
.jpg?width=300){kind=avatar}
Large facilities housing servers and GPUs for AI compute, which are increasingly hard to power and zone.
Investor and podcast host offering insights on geopolitical conflicts, US policy, and politics.
Host of the All-In Podcast conducting the interview with Ryan Cohen.
The podcast hosting the interview with GameStop CEO Ryan Cohen.
Investment funding provided to startups and early-stage companies.
.jpg?width=300){kind=avatar}
Co-founder of Android who helped conceptualize Google Ventures.
Founding CEO of Google Ventures and founder of Section 32, focusing on life sciences and deep tech.
A venture fund founded by Bill Maris, operating smaller funds to achieve top-decile returns.
The venture capital investment arm of Alphabet Inc., founded by Bill Maris.
An agriculture tech company founded by David Friedberg, marking GV's first ex-Google investment.
Technology news publisher that highlighted Google Ventures' strategies.
The foundational field of computation driving modern innovation.
Data analysis and computational modeling to understand biological systems.
The corporate term Google preferred to use before AI became mainstream.
Subsidizing products to aggressively underprice competitors.
Entrepreneurs launching businesses influenced by VC valuations.
Liquidity events necessary for VC funds to return capital.
The integration of computing seamlessly into everyday surroundings.
A venture capital firm known for highly concentrated, massive returns.
A prominent venture capital firm mentioned in contrast to smaller funds.
A cybersecurity company and portfolio investment of Section 32.
Co-founder of Coinbase and co-founder of New Limit.
Investor and co-founder of the life sciences company New Limit.
.jpg?width=300){kind=avatar}
Technology entrepreneur who provided historical photographs for Bill Maris's presentation.