Y Combinator President Garry Tan: Betting on New Founders Against Frontier Lab Commercial Teams
Y Combinator President Garry Tan stated that YC continues to fund over 800 teams, almost all of which are first-time founders, because their core judgment is that "the founder themselves is the variable." He also pointed out that the opposite of this bet is the commercialization teams within frontier AI labs.
This statement directly reveals a hidden divide in the current AI industry: on one side are frontier model companies represented by OpenAI and Anthropic, which attempt to promote commercialization through internal teams; on the other side are external entrepreneurs backed by organizations like YC, who build applications and new product forms around these foundational capabilities.
Source: Public Information
ABAB AI Insight
This is essentially a competition over "ownership of innovation." Frontier labs possess models and computational power, giving them a technological advantage, but their organizational structure is closer to that of research institutions or platform companies; whereas entrepreneurs can quickly experiment in uncertain markets and are better at defining specific scenarios and product paths. The tension between the two determines whether value in the AI era is confined within platforms or diffused to external ecosystems.
YC's large-scale bet on "first-time founders" is fundamentally about amplifying the coverage of exploratory space. The generality of AI capabilities means that application layer opportunities are extremely dispersed, making it difficult for a single company's internal team to exhaust them. By investing in numerous small teams, YC is building a decentralized "experimental network" to capture new paradigms through probabilistic advantages, rather than relying on the linear progress of a few mature teams.
The so-called "commercialization folks awkwardly" reflects the structural constraints large companies face in commercialization. Internal teams are often limited by existing products, brands, and risk controls, making it difficult to make radical attempts; whereas external entrepreneurs, unburdened by historical baggage, can more quickly align with user needs. This difference is particularly evident when technological paradigms are just changing.
From a longer-term perspective, this division resembles a replay of the relationship between operating systems and application developers. Foundational model companies are closer to the "new operating system," while entrepreneurs are the application layer on top of it. Who ultimately captures greater value depends not only on technological leadership but also on the degree of ecosystem openness and profit-sharing structures, which is the core of the current game.