Elad Gil: 91% of Generative AI Unicorn Valuations Concentrated Within 1.5-Hour Radius of San Francisco Bay Area
Investor Elad Gil pointed out while sharing Shreyan Jain's analysis of 2026 unicorn data that generative AI unicorn valuations are highly concentrated in the San Francisco Bay Area, which accounts for 91% of the global market value of private AI companies within a 1.5-hour radius. This proportion is particularly exaggerated in the dimension of "purely generative AI unicorns," with almost no other city capturing more than 2% of the share.
Public data from CB Insights and Elad's further analysis show that the overall market value of global unicorns increased from approximately $4.4 trillion to about $5.8 trillion within a year, with the average valuation of a single unicorn rising to over $4.4 billion, reaching a six-year high. Among these, the share of generative AI in the total market value of unicorns jumped from about 2% in 2023 to 22% in 2026, with almost all incremental growth contributed by companies in the U.S., particularly in the Bay Area.
Supplementary data from English media and real estate consulting firms indicate that the Bay Area concentrates the vast majority of global AI startups, venture capital, and core talent, with San Francisco and San Jose becoming the core of the "AI supercluster," attracting significant investments in AI-related office space and infrastructure, further reinforcing the local closed loop of "AI–talent–capital–real estate."
Source: Public Information
ABAB AI Insight
This is not "city luck," but a typical supernode of resource self-reinforcement: when AI capital, talent, data centers, and customers are highly concentrated within the same physical radius, the efficiency of obtaining synergies from any new resources entering the Bay Area is far higher than in other regions. This "AI supercluster" structure is more extreme in the Bay Area than in New York, London, or Beijing, directly leading to the Bay Area AI unicorns still obtaining disproportionate valuation premiums even amid a tightening global capital environment.
From a long-term structural perspective, the Bay Area is becoming the "natural pricing layer for AI capital." When the largest AI model companies, AI infrastructure, application platforms, and top venture capital are concentrated in the same geographic circle, capital markets will default to the "pricing home for AI assets" being in the Bay Area, thereby assigning an additional "Bay Area premium" in valuation models. This premium, in turn, attracts more AI capital, forming a self-circulation of "capital—valuation—capital."
A deeper change is the re-centralization of global innovation geography. After once being hotly debated as "remote work and decentralization," the infrastructure of the AI cycle relies heavily on computing power, data, and talent density, making the Bay Area once again the "control point of the physical layer of AI." This means that AI companies far from this core area face long-term constraints of "longer signal links, thinner capital and talent," making it difficult to compete equitably in valuation and financing rhythm, even if their technical capabilities are strong.