Sequoia's AI Ascent — The Cognitive Revolution
transcript
show notes
Guillermo Flor's summary of Sequoia Capital's AI Ascent event — a gathering of 150+ top AI minds, including Demis Hassabis, Andrej Karpathy, Greg Brockman, Boris Cherny, Dmitri Dolgov, and Jim Fan.
The core thesis: AI is not another software wave but a new foundation layer — a computation revolution, not a communication revolution — and the biggest, fastest wave in the sequence from semiconductors to mobile. Three inflections frame the moment: ChatGPT (2022), reasoning models, and long-horizon agents. Commercial AGI is defined as a system that can be dispatched to a job, recover from failure, and persist to completion, delivering 10-40x improvements rather than 10-40%.
Also covered: Pat Grady's calibration (waves are additive; roughly $10 trillion of services potentially addressable by machines, with US legal services alone around $400 billion); moats rebuilt around customer intimacy and workflow integration, with the diffusion gap between fast lab capability and slow enterprise adoption; Sonya Huang's agents framework (2026 as the agents story, from tab autocomplete through background agents to dark factories; services as the new software); and Constantine Buhler's cognitive revolution — the forecast that machines could perform 99.9% of cognition within a generation, closing on Protagoras: value concentrates in human experience, relationships, and attention.
Source: Guillermo Flor (@guilleflorvs) on X, Nov 2, 2026. Full breakdown: "Sequoia AI Ascent 2026: The future of AI" at theaiopportunities.com (AI MARKET FIT).