Descriptions:
In this Bloomberg Technology interview, Bennett, co-founder of A-Star, breaks down how his early-stage VC firm navigates a startup funding landscape increasingly dominated by mega-rounds. A-Star has grown from a $300 million debut fund to its current $450 million Fund III while staying committed to traditional seed investments of $2–$5 million — well below the headline-grabbing rounds being raised by AI research labs.
Bennett describes a clear bifurcation in the market: younger founders building application-layer products on top of OpenAI and Anthropic are still raising conventional seed rounds, while researchers spinning out of established model labs — such as Thinking Machines Lab and Safe Superintelligence — are commanding $2 billion or more before shipping any product. A-Star has largely avoided the latter category, which is led by firms like Andreessen Horowitz, focusing instead on the long-compounding venture model where early bets can return 100–300x over a decade.
The conversation also touches on how incentives shift as fund sizes grow, pushing larger firms toward later-stage deployment. Bennett argues that despite the noise around comically large pre-product raises, the traditional seed market remains healthy for application-layer founders building on top of foundation models. The interview offers a grounded perspective on capital efficiency and partnership strategy in an AI funding environment that many argue has become detached from fundamentals.
📺 Source: Bloomberg Technology · Published May 12, 2026
🏷️ Format: Interview
