In response to skyrocketing seed valuations, VCs are shifting their portfolio construction models. Instead of targeting a specific ownership percentage, the key decision is now what percentage of the total fund to deploy into a single deal. The focus has moved from ownership to the magnitude of the bet relative to the fund size.

Founders must understand that taking venture capital means their startup is now a financial instrument for the VC's fund. The VC's return expectations become the startup's required trajectory, a critical alignment in an AI era where investors expect astronomical outcomes.

Sam Lessin predicts massive losses for seed VCs backing companies branded as "AI businesses." These ventures are too capital-intensive and commoditizable to generate traditional venture returns, even if they become massive. AI should be a tool, not the business model itself.

Pre-product AI startups are commanding billion-dollar valuations because the barrier to entry has skyrocketed. To build a competitive new foundation model, a startup must be able to raise approximately $2 billion before even launching a product. This forces VCs to place massive, early bets on a very small number of elite, pedigreed founders.

Despite OpenAI's massive success, its capital-intensive nature means early seed investors see returns around 25x. While good, this isn't the massive fund-returner many assume, highlighting the risk of capital-consumptive businesses for seed funds, even when they become unicorns.

A new startup strategy involves acquiring traditional businesses and dramatically increasing their margins by integrating AI. This approach requires a unique blend of M&A, operational change management, and AI expertise, differing from typical venture-backed company creation.

AI companies raise subsequent rounds so quickly that little is de-risked between seed and Series B, yet valuations skyrocket. This dynamic forces large funds, which traditionally wait for traction, to compete at the earliest inception stage to secure a stake before prices become untenable for the risk involved.

Conventional venture capital wisdom of 'winner-take-all' may not apply to AI applications. The market is expanding so rapidly that it can sustain multiple, fast-growing, highly valuable companies, each capturing a significant niche. For VCs, this means huge returns don't necessarily require backing a monopoly.

AI startups' explosive growth ($1M to $100M ARR in 2 years) will make venture's power law even more extreme. LPs may need a new evaluation model, underwriting VCs across "bundles of three funds" where they expect two modest performers (e.g., 1.5x) and one massive outlier (10x) to drive overall returns.