Sequoia's cognitive revolution essay compares AI to industry

Sequoia Capital partner Konstantine Buhler published "The Cognitive Revolution" on September 2, 2026. His thesis is that the AI transition rhymes with the Industrial Revolution. Machines took over most physical work over two centuries. AI will take over most cognitive work, and it will happen faster. The essay is a long-range frame that founders can use to place their product in a historical pattern.

Physical work and cognitive work

Buhler splits work into physical and cognitive. He writes that over about two centuries, physical work went from 99% done by muscle to 99.9% done by machines. He argues cognitive work is at the start of the same curve. A century ago, 99% of cognitive work was done by humans. He expects 99.9% to be done by machines in the near future, because machines will compute far more, not because people will think less.

He estimates that cognitive and physical work could split a world economy of about $120 trillion roughly equally. The physical half has been mechanizing for 200 years. The cognitive half has barely started.

Falling cost and rising demand

Buhler maps the inputs. Coal and iron became electricity and compute. The Bessemer process became algorithms such as the Transformer. He writes that the cost of cognition is collapsing faster than the cost of mechanical work did after 1800.

He cites William Jevons, who observed in 1865 that more efficient engines increased Britain's coal use rather than reducing it. Buhler expects the same for cognition. Most problems currently go unexamined because thinking is expensive. As the cost falls, demand rises.

Coding agents as the Spinning Jenny

The Spinning Jenny, built in 1764, moved spinners from doing the task to supervising it. Buhler sees AI coding as the equivalent: the first skilled cognitive task where the machine does most of the work and the engineer directs and reviews it. He argues the defining product of this revolution, its equivalent of the automobile, has probably not been built. His candidates include machine-speed science and personal agents.

Jobs and Engels' pause

Buhler notes that over 75% of American workers farmed in 1800, against about 1% of employment today. New categories of work replaced the old ones. He also cites the economic historian Robert Allen on "Engels' pause." Between 1780 and 1840, British output per worker rose 46% while real wages rose 12%. Between 1840 and 1900, output per worker rose 90% and real wages rose 123%.

His concern is speed. The Industrial Revolution gave workers decades to adjust. Buhler writes that this one may offer five years, and that education and safety nets were designed for the old pace.

Education and health

Buhler cites Benjamin Bloom's 1984 finding that a student tutored one-to-one outperformed 98% of students in a conventional classroom. He argues the cost of a patient, knowledgeable tutor is approaching zero. In health, he points to drug discovery as search at a scale humans cannot manage, and to thousands of diseases left untreated because investigating each one was never economic.

What stays human

Buhler argues that what stays economically human longest is wanting things, choosing between them, being accountable and being trusted. In his words, the machine can draft the treaty, but someone still has to sign it.

Founder takeaways

• Target latent demand. Look for work that goes undone today because it costs too much, not only work done today.
• Build for supervision. Products should let one person direct and check many machine tasks.
• Sell accountability. Where a human must sign off, design the product around that sign-off.
• Look beyond coding. Buhler sees coding as the first act, with larger markets still ahead.

Sources

1. The Cognitive Revolution
   Sequoia Capital (Konstantine Buhler) · Sep 2, 2026