The New Economics of AI Investing

The traditional venture capital model is undergoing a fundamental shift. David George (a16z) and Aram Verdiyan (Accolade Partners) argue that the "power law"—where a small number of investments drive the vast majority of returns—is no longer just a feature of venture capital, but a systemic reality of the AI era. Unlike traditional software, where capital was often used to hire more people and build features, AI allows capital to be deployed directly into compute, which creates a compounding product advantage. This capability has fundamentally changed the risk-reward profile of the asset class.

Why AI is Categorically Different

AI is not merely the next evolution of SaaS; it is a paradigm shift targeting massive sectors of the global economy, including labor, healthcare, and transportation. While SaaS took 15 years to reach $100 billion in revenue, AI achieved this in four. The panelists suggest that the Total Addressable Market (TAM) for AI is significantly larger than traditional software because it targets the economic value of tasks performed by humans rather than just software IT spend.

Because of this, the "winner-take-all" dynamic is more pronounced. In this environment, being the category winner is essential, as second place often plays for scraps. However, the market is expanding so rapidly that multiple categories are emerging simultaneously, meaning that while individual categories follow a winner-take-all pattern, the overall opportunity set is vast and non-zero-sum.

Portfolio Construction and the "Death of the Middle"

Investors are increasingly moving away from the "middle" of the market. The panelists identify a clear bifurcation:

• Specialized Seed Funds: These firms can carve out a niche by identifying winners at the pre-seed stage before larger firms have enough signal to commit.
• Lifecycle Investors: Large firms like a16z leverage their massive operating resources (700+ employees) to derisk outcomes for founders, which creates a flywheel effect. Founders prefer partners who can help them land customers and talent, and this brand reputation attracts the best founders, which in turn attracts the best LPs.

For institutional allocators, the data is stark: only about 1% of firms have achieved consistent 3x net returns over the last two decades. Access to the top 5-10 category-defining companies is the primary driver of this performance. Without this access, investors are likely to underperform public markets, making the case for AI as a "core" rather than "satellite" allocation.

The Role of Capital as a Competitive Moat

Historically, throwing money at a startup often led to coordination issues and overhead. In the AI era, capital buys compute, and compute improves the product. This creates a unique feedback loop where capital directly reinforces a company's competitive advantage. This dynamic is why late-stage venture outcomes are now reaching valuations previously reserved for public companies, and why the largest opportunities in robotics, energy, and physical infrastructure are likely still ahead.