Executive Summary
A16z's investment team identifies three converging infrastructure primitives that bypass traditional distribution bottlenecks. Guy Willett argues stablecoins are narrow banks that must evolve into on-chain credit origination to scale beyond tokenized fiat, reducing loan servicing costs from 1-3% annually while enabling composability impossible in traditional finance. Oliver Shoe sees autonomous labs emerging through AI reasoning combined with robot learning, creating collaborative human-AI-robot systems where interpretability becomes the key differentiator for research applications. James de Costa reveals the greenfield strategy: AI-native startups selling to other AI-native startups at formation, exploiting the structural disadvantage incumbents face when serving low-revenue, high-cost early-stage customers. These aren't incremental improvements but foundational shifts that create entirely new market dynamics. The thesis connects through distribution arbitrage - native-first builders can serve customers incumbents cannot economically reach, then scale alongside them. Willett's synthetic dollars backed by infrastructure assets, Shoe's interpretable lab automation, and de Costa's startup-to-startup sales model all exploit the same dynamic: building for audiences incumbents ignore creates compounding advantages. This aligns with our past research on infrastructure value creation, where regulated platforms like Coinbase benefit from complexity while payment infrastructure captures creator economy value. The convergence suggests 2026 will favor companies that build native-first solutions rather than blockchain copies of existing systems.
Key Insights
what Guy Willett, Oliver Shoe, James de Costa said“it can drastically reduce back-office costs like loan servicing, which in many cases can take one to three percent of the outstanding credit facility itself every year”
This is a preview. Log in to see the full analysis including investment opportunities, risks, catalysts, and detailed insights.