How AI Is Rewriting the Power Law of Venture Capital
A16Z partners discuss how AI is fundamentally reshaping venture capital dynamics, creating more extreme power law distributions where capital directly compounds competitive advantages. They argue that venture capital—particularly in frontier AI—should become a core allocation for most institutional investors, and that portfolio construction, access, and position sizing now matter more than ever.
Summary
This A16Z podcast episode features Jem Kha, David George, and Accolade Partners' Adam Verdean discussing how AI is rewriting venture capital's power law. The conversation begins with striking data: only 20 out of 3,000 US venture capital firms have achieved consistent 3x net returns over two decades, and this dispersion is becoming even more extreme.
The core thesis is that AI is fundamentally different from previous technology paradigms. Unlike traditional startups where excess capital becomes a liability through coordination problems, frontier AI companies can convert dollars directly into compute, which directly improves products and reinforces competitive advantages. This creates a compounding effect unavailable in prior eras. David George emphasizes that AI is attacking every facet of the GDP simultaneously—transportation, labor, services, capital, and coordination—representing roughly $30 trillion in addressable markets, compared to previous paradigms that hit narrower slices.
The speakers discuss how AI's total addressable market extends far beyond software. While healthcare IT spending is $60-100 billion annually, AI targets the trillion-dollar labor market in healthcare (claims, billing, administration). The economic value of tasks being automated becomes the real TAM, not just software licensing. This challenges traditional SaaS valuation models.
On portfolio construction, they stress that consistency in venture comes from having access to category-defining companies across every vintage. Of the 20 top-performing firms, most achieved this through vertically integrated approaches combining early-stage, seed, growth, and sometimes late-stage capabilities. Large funds can now generate venture-like returns in late-stage if they can size positions at 5-10% of fund value, which was previously impossible. This requires early-stage relationships to maintain through later rounds.
The speakers address LP incentive misalignment: GPs get fired for omission (missing the next Facebook), while LPs only get fired for commission (investing in vendors). This creates opposing risk tolerances. Additionally, LPs lack incentives to embrace change—they can match benchmarks by investing conservatively and keep their jobs, while missing frontier opportunities still leaves them acceptable.
They discuss the "death of the middle"—dedicated mid-stage funds struggle because founders prefer either highly specialized early-stage experts or large integrated platforms. Small specialized firms can coexist by moving pre-seed earlier than large firms; large diversified platforms win through scale and ecosystem benefits.
On legacy software companies: those from pre-ChatGPT vintages valued at 25-32x EBITDA now face existential challenges. They can't IPO at prior valuations, PE buyers demand AI-native capabilities, and many aren't growing fast enough. Some may adapt (like Intercom), but many face liquidation or significant write-downs. The speakers note that simply adding AI customer service to legacy workflows often destroys NPS and creates negative spirals, especially if debt-laden.
Regarding timing and valuation concerns, they argue that early-stage AI companies often show misleading traction signals (companies going from zero to $5M ARR in months with no renewal data), making evaluation extremely difficult. They emphasize founder judgment and deep customer understanding over financial metrics as signals for real adoption.
The speakers identify major untouched opportunities: consumer AI will move beyond chatbots to proactive, agent-based interfaces; robotics will likely exceed language models in value creation over the next decade; autonomy is barely started (fewer than 10,000 Waymos deployed); healthcare drug discovery and care delivery are largely untouched; and physical world domains (defense, manufacturing, energy, data centers) represent massive opportunities. A critical constraint emerges: energy and infrastructure (grid access, transmission, permissioning) may be the real bottleneck, not demand.
They conclude that venture capital allocation should increase substantially for institutional investors—possibly to 40%+ versus historical 9-15%—not just for direct venture returns but because AI exposure across all asset classes is becoming essential. The 20 firms that consistently win deserve concentrated LP capital, while diversification across many underperforming firms destroys returns.
About this episode
a16z’s Jen Kha and David George sit down with Accolade Partners’ Aram Verdiyan to discuss how AI is changing the power law of technology investing, why the largest companies can compound advantages in ways that weren’t possible before, and what that means for how investors construct portfolios. They explore why AI may be much bigger than traditional software, with applications reaching into labor, healthcare, transportation, services, and other major parts of the economy. David explains why capital itself can now reinforce an AI company’s advantage by buying more compute, while Aram makes the case that AI should increasingly be treated as a core allocation rather than a satellite position. The conversation also gets into the changing economics of venture and growth investing, how to distinguish real AI traction from early hype, what AI means for legacy software and private equity, and why some of the largest opportunities may still be ahead in robotics, autonomy, healthcare, energy, and physical infrastructure.
Key Insights
- Only 20 out of 3,000 US venture capital firms have achieved consistent 3x net returns over two decades, suggesting extreme performance concentration that is becoming even more pronounced with AI.
- AI companies uniquely benefit from additional capital because dollars convert directly into compute, which improves products, whereas traditional startups suffer coordination problems from excess funding.
- AI's addressable market extends far beyond software to actual labor costs and task value—healthcare's trillion-dollar administrative labor market is 10x larger than traditional healthcare IT spending categories.
- Large venture firms with integrated early-stage through late-stage platforms can now achieve venture-like returns in late-stage by sizing positions at 5-10% of fund value, but this requires early-stage relationships built over years.
- LP incentives are misaligned with GP risk-taking: GPs face consequences for missing opportunities while LPs face consequences only for bad investments, making LPs rationally conservative even when missing frontier returns.
- Pre-ChatGPT software companies valued at 25-32x EBITDA cannot go public at those valuations, have no PE buyers, and lack AI-native capabilities, creating a massive write-down and liquidation problem.
- Early-stage AI company traction metrics are highly misleading—companies reaching $5M ARR in months with no renewal cycles often receive valuations similar to actual recurring revenue businesses, making selection extremely difficult.
- Energy infrastructure and grid access, not demand, may be the actual bottleneck for AI scaling, as the US has slower renewable capacity deployment than other countries and regulatory constraints on transmission and permissioning.
Topics
Transcript
We've looked at the data of 3,000 venture capital firms in the US. Only 20 have achieved consistent 3x net returns over the last two decades. Right now, clearly the power law is more extreme than it has been in the last 10 to 20 years of technology investing. For the first time, you can take capital and throw it at a company, and it compounds their advantage. AI is attacking every facet of the GDP. Transportation, labor, services, capital, coordination. There hasn't been a technology paradigm that hits on 30 trillion in GDP at the same time. Elon has talked publicly about Brock Bot on Sam's side. He's talked about Astra and some of the long-running capabilities that are…
Full transcript available for MurmurCast members
Sign Up to AccessMore from The a16z Show
Who Grades the AI Models? | Ben Horowitz & Rayan Krishnan
VALS, an independent AI evaluation company, addresses the gap where public benchmarks fail to accurately measure model capabilities—evidenced by Meta's Llama 4 underperforming on private benchmarks while excelling on public ones. The podcast discusses how third-party evaluators are essential for both labs seeking credible performance proof and enterprises needing ROI justification for AI spending, while also exploring the role of standardized evaluations in policy and geopolitical AI governance.
OpenAI Researchers on the Future of Mathematical Reasoning
OpenAI researchers discuss how AI models are making progress on long-standing mathematical problems by combining literature knowledge, executing complex proofs with precision, and exploring multiple approaches without human cognitive biases. They present case studies in sphere packing, coding theory, and group theory, arguing that AI's ability to persist through difficult problems and leverage symmetry properties is fundamentally changing what mathematics gets solved and how it's practiced.
Can Open Source Keep AI Power From Concentrating?
Lucas Kaiser, co-author of the Transformer paper, discusses how AI power is currently concentrating in large companies due to the resource-intensive nature of current technology, but argues this is not inevitable. He believes research breakthroughs in algorithms and training methods could enable smaller players and distributed models to compete effectively.
Your AI Doctor Is Coming | Julie Yoo
Julie Yoo, a healthcare investor at Andreessen Horowitz, argues that AI will benefit healthcare more than any other industry because healthcare has historically underinvested in technology, allowing it to leapfrog legacy systems and adopt AI-native solutions directly. She identifies major opportunities in consumer healthcare, AI-native care delivery, robotics, and new payment models, predicting a future where individuals have personalized AI doctors available continuously.
Aaron Levie on Why Open AI Wins
Aaron Levie, CEO of Box, discusses why open-weight AI models benefit the entire AI ecosystem rather than threatening frontier labs, argues that the economic value in AI accrues to inference infrastructure rather than model weights, and explains why model routing will become the default enterprise AI strategy.