Who Wins in the Next Decade of VC | Margin and Price Matter Less in a World of AI | Sameer Dholakia

Sameer Dholakia argues that AI is a larger supercycle than either the internet or cloud because intelligence will be a market bigger than oil and useful to every human and every company. He frames the current moment as “magic,” says the capital intensity of frontier models means the market will consolidate into a few winners, and views both closed and open models as large opportunities rather than an either-or choice. He also says the scale of the opportunity makes precise debates about token mix less important than backing the strongest platforms and applications.

He revisits Bessemer’s Anthropic investment, saying the firm joined the Series D at a $15 billion pre-money valuation and that it was the lowest-voted deal in firm history. In hindsight, he says Bessemer should have written a bigger check, but the conviction came from seeing a cloud-like market structure where a few foundation model leaders would sit underneath many AI applications. He also says the OpenAI opportunity he passed on was a useful lesson in how quickly conviction can be stretched by market growth.

On investing discipline, Dholakia says entry price still matters, but only in the context of whether a company can compound into a venture-scale outcome. He prefers judging deals on NTM metrics, revenue quality, and the ability to expand over time, and says capital is now a moat in AI because it drives scale, brand, talent, and market leadership. He closes by arguing that venture is splitting into mega-platforms and specialists, that firms should stay private longer, and that founders and investors should take some liquidity off the table when the opportunity allows.