Big Ideas 2026: New Infrastructure Primitives - a16z Podcast Recap

Podcast: a16z Podcast

Published: 2025-12-26

Duration: 20 minutes

Guests: Guy Willette, Oliver Shu, James da Costa

Summary

The episode examines three major shifts in infrastructure: the evolution of programmable money, the integration of AI into scientific research, and a new distribution strategy for AI-native startups.

What Happened

The episode kicks off with Guy Willette discussing how programmable money is moving beyond stablecoins to on-chain credit origination and synthetic financial products. He emphasizes the potential for decreased operational costs and increased scalability compared to traditional finance models. Willette sees an emerging role for private credit funds to facilitate on-chain loans, significantly reducing back-office expenses traditionally associated with loan servicing.

Next, Oliver Shu introduces the concept of autonomous labs, where AI and robotics collaborate to enhance scientific research. He discusses the importance of interpretability in AI-driven research, noting that the integration of AI reasoning and lab automation is a step toward more efficient scientific processes. Shu highlights various startups and government collaborations that are working on closing the loop in scientific experimentation.

The episode then shifts to James da Costa, who outlines a new distribution strategy for AI-native startups. He argues that selling to other startups at their formation allows for a growth trajectory as these companies scale. Da Costa uses Stripe as a case study, showing how it initially targeted startups and grew as its customers grew.

Willette elaborates on how stablecoins currently resemble narrow banks and proposes a shift towards natively originating credit on-chain. He believes this shift can lead to more efficient financial products, particularly in emerging markets where derivatives often outperform underlying assets in trading volume.

Shu discusses the uneven progress across various scientific fields and the potential of AI-driven science to speed up research in life sciences, chemistry, and material science. He notes collaborations like the Genesis mission and partnerships with companies like DeepMind as significant steps forward.

Da Costa points out that incumbents often struggle with startup clients due to their rigid structures and suggests that new startups offer a unique opportunity for growth. He highlights the role of accelerators in providing a steady flow of new customers for startups adopting this strategy.

The episode concludes with a synthesis of how these new infrastructure primitives - financial, scientific, and distributional - are creating new compounding markets. Each speaker illustrates how these foundational changes enable entirely new systems to emerge and grow.

Key Insights

• Programmable money is evolving to include on-chain credit origination and synthetic financial products, potentially reducing operational costs and increasing scalability compared to traditional finance models.
• Autonomous labs, where AI and robotics collaborate, are enhancing scientific research by integrating AI reasoning with lab automation, which could lead to more efficient scientific processes.
• AI-native startups are adopting a distribution strategy of targeting other startups at their formation, allowing them to grow alongside their customers, as demonstrated by Stripe's initial focus on startups.
• Stablecoins currently function similarly to narrow banks, but there is a push towards originating credit directly on-chain, which could result in more efficient financial products, especially in emerging markets.