Descriptions:
Bloomberg Tech interviews Navin Chaddha, managing partner at Mayfield, a VC firm that has deployed over $3 billion into AI companies across its history. Chaddha argues that AI is simultaneously making it easier than ever to launch a startup and harder than ever to build one that lasts — a dynamic that makes early-stage conviction bets more important, not less. Mayfield runs a deliberately selective portfolio of 8–10 investments per year, targeting inception-stage founders before they have even formalized a company.
Chaddha breaks down where value is concentrating across the AI stack: he estimates roughly 70% of value is currently accreting at the semiconductor layer, where veteran engineers dominate; the model layer rewards PhD researchers and postdocs; and the agent and application layer is where younger, first-time founders are emerging as a major force. He notes that Mayfield bets on “jockeys, not racetracks” — half of successful portfolio companies pivot significantly from their original idea, so founder quality outweighs initial thesis.
On exits, Chaddha predicts a near-term wave of strategic acquisitions in the $1–50 billion range alongside mid-cap IPOs targeting $5–10 billion market caps, with companies staying private longer because growth capital is available at scale. He frames AI as a 100x opportunity but emphasizes that power law dynamics mean returns will concentrate among a tiny fraction of investments — making selectivity, not volume, the defining variable for firms like Mayfield.
📺 Source: Bloomberg Tech · Published August 20, 2026
🏷️ Format: Interview
