a16z rewrites venture capital with compute-driven monopolies
By Klaus Berger · Reporting from Frankfurt ·
Ron and David George argue that AI is transforming venture capital by shifting trillions in labor spending toward massive compute investments and extreme power law dynamics.
# How AI Is Rewriting the Power Law of Venture Capital
Ron and David George argue in the a16z podcast, "How AI Is Rewriting the Power Law of Venture Capital," that AI is not just another software update. Instead, it targets the daily grind of paralegals, accountants, and coders—a labor market worth $30 trillion, roughly the annual GDP of France and the UK combined. According to David George, because the US spends forty-fold more on labor than on software, the market for AI could dwarf traditional SaaS by ten times.
Can compute buy a monopoly?
Ron and George say AI has fundamentally altered the "power law," where a few winners take almost everything. In this new era, capital compounds advantage because investors spend it on compute, which immediately sharpens the product. This creates a self-reinforcing loop of scale more aggressive than any previous tech cycle.
To drive this growth, a16z has built an engine of 700 employees to help founders scale from seed to IPO. George says this massive operating resource allows a firm to win the best deals and shield founders from early volatility. In this environment, mid-sized VC firms vanish because founders prefer the orbit of a giant.
The Moral Hazard of Huge Checks
The idea that a fund can shield a startup from risk by simply growing larger is magical thinking. When a firm institutionalizes moral hazard by writing $500 million checks into a single company, it does not eliminate risk; it piles it into one basket.
The logic for this concentration is that only the largest funds have the firepower to support a market leader in a world of soaring compute costs. If the power law is truly this extreme, then diversifying is a waste of time. However, this ignores the history of market shifts. We have seen this rhetoric before: the belief that a new paradigm makes cash flow and customer acquisition costs obsolete. When the "power law" becomes the only reason to ignore the math, the market stops pricing risk and starts betting on a miracle.
Legacy software is hitting a wall
The discussion turns sharp when the speakers address the stagnant software companies of the 2016–2021 vintage. These firms grew at moderate rates and saw their valuations peak during the 2021 buyout boom, but they are now faltering.
| Cohort | 2021 Valuation | Current Status | | :--- | :--- | :--- | | 2016–2021 Vintage | 25–32x EBITDA | 50% Value Loss | | AI-Driven SaaS | >10x Revenue | Accelerating |
This is where the balance sheet collides with the physical world. The public market is finally discounting legacy software, rewarding only those that accelerate growth through AI. The a16z team notes that putting Sears on a website did not make it Amazon. Yet, they suggest the solution is to allocate more venture capital. They overlook the fact that the current bottleneck is not a lack of money, but a lack of copper wires, cooling fans, and high-voltage substations—problems a VC engine cannot solve.
The belief that we can inflate our way toward a $100 trillion horizon—a sum exceeding three times the current global GDP—ignores the truth that stability is a policy, not an accident. By prioritizing growth over profit and concentrating capital into a few super-funds, the venture ecosystem is building a house of cards. We are replacing the disciplined selection of value with a preference for brand-name orbits. History suggests that when the "power law" justifies ignoring the rulebook, the eventual correction is not a dip, but a collapse.