Descriptions:
a16z general partner Alex Rampel delivers a sweeping investment thesis on the AI application layer, framing the current moment as the fifth major technology product cycle — following the PC, internet, cloud, and mobile eras. Drawing on historical NASDAQ data and prior platform transitions, Rampel argues that AI is uniquely positioned to compound faster than previous cycles because it is built on top of already-pervasive infrastructure: 8 billion smartphone users, globally distributed cloud compute, and broadband connectivity that earlier waves had to build from scratch.
The talk uses Ramp, the corporate expense and card management platform, as a case study showing a sharp enterprise adoption inflection in early 2025. Rampel then examines Salient AI in depth — a voice agent startup automating debt collections calls. Salient replaces call centers plagued by 40-70% annualized employee churn with software that tracks compliance requirements across all 50 US states at the county level, operates in 21 languages, and achieves collection rates 50% higher than human agents. The company’s ingestion of every proposed statute as it moves through legislative processes is cited as a durable data moat.
A significant portion of the talk addresses how competitive moats evolve in the AI era, arguing that lower software creation costs make domain-specific data advantages more important rather than less. Rampel also directly counters published research claiming enterprise AI deployments are underperforming, presenting adoption data from forward-thinking companies as contradictory evidence. The presentation is broadly aimed at investors and founders evaluating where durable value will accumulate in the AI application stack.
📺 Source: a16z · Published January 19, 2026
🏷️ Format: Deep Dive
