“State a hypothesis and the test that would disprove it.”
**Hypothesis:** Labor-saving machines create net job growth by freeing workers for higher-value roles.
**Disproving test:** Measure employment in the affected sector 5 years post-invention. If jobs decline despite economic growth, the hypothesis fails.
**Second-order consequence:** A transcription machine eliminates typists. Freed capital floods the market with cheap digital documents, collapsing professional writing rates. Journalists, editors, and copywriters see income plummet. Training costs to "pivot" exceed most workers' savings. Entire communities dependent on these jobs suffer concentrated unemployment, even as GDP grows. The machine succeeds; the workers fail.
“No instruments. Reason from speed, heading, and time elapsed.”
# The Unplanned Consequence
Consider a machine that automates writing—say, email composition. Initial benefit: effortless communication.
Second-order consequence: **information overload accelerates**. With zero friction to sending messages, people transmit more. Recipients drown in volume. Signal-to-noise ratio collapses. What once took thought now requires none, so trivial messages proliferate. The collective attention economy crashes. Paradoxically, making one job effortless makes *receiving* that output harder. Organizations respond by adding filters, which require new labor. The time saved vanishes into managing the debris of easy production. Efficiency at the individual level creates inefficiency at scale.