The $10 Trillion AI Revolution: Why It’s Bigger Than the Industrial Revolution
Sequoia’s Konstantine Buhler argues AI compresses the 144-year Industrial Revolution into a few years and targets a $10T US services market that is 0.2% automated today.
- US services market is $10T; only ~$20B automated by AI today—Sequoia’s core investment thesis.
- AI mirrors the Industrial Revolution: steam engine (1712) to assembly line (1913) took 144 years; Sequoia expects AI to compress that to a few years.
- ‘FLOPs per knowledge worker’ is the new production function; portfolio companies forecast minimum 10x increase, optimistic case is 1,000–10,000x.
- Work is shifting from 100% certainty/low leverage to 1,000%+ leverage with lower certainty—Rocks (AI sales agents) cited as live example.
- Persistent memory has no equivalent of scaling laws yet; vector DBs and long context windows haven’t cracked it—Sequoia sees a major open opportunity.
- Academic AI benchmarks are dead; real-world validation is the new gold standard (Expo ranked #1 on HackerOne against human hackers).
- Open source AI is at a ‘precarious’ moment—two years ago it seemed competitive with frontier models; Sequoia now less certain but still betting on it.
- AI security is a sleeper theme: hundreds of AI security agents per human and per agent are possible, unlike physical-world constraints.
2025-08-28 · Watch on YouTube
