Humans Are Becoming Horses

TL;DR

Humans are becoming horses — not metaphorically but economically — because just as general-purpose mechanical muscles made horses unemployable after 1915 (not lazy, just useless), general-purpose mechanical minds like Baxter ($25K robots costing pennies/hour), self-driving autos, Watson, and self-teaching stock-trading bots are making human labor structurally unnecessary across physical, white-collar, and creative work, threatening to push unemployment past the 25% Great Depression threshold since the top 32 job categories (45% of the workforce) all existed 100 years ago and are already automation targets.

Key Points

• 1.Baxter, the general-purpose robot, costs less than one year of a human worker's salary and can learn tasks by watching — making him the "1980s computer," a cheap beginning rather than the peak of robotic capability.
• 2.The U.S. transportation industry alone employs ~3 million people, extrapolating to ~70 million jobs globally — all classified in the video as effectively "over" once self-driving autos scale, since humans represent roughly one-third of total costs for transportation companies.
• 3.IBM's Watson is already practicing medicine at Sloan-Kettering, diagnosing lung cancer — with an advantage over human doctors because it can track every patient worldwide, read all current research, and cross-reference every drug interaction simultaneously.
• 4.Legal discovery — reviewing millions of emails and documents for patterns — is already performed by bots in many law firms, outperforming humans in speed, cost, and accuracy because "bots don't get sleepy reading through a million emails."
• 5.The background music in the video itself was composed by a bot named Emily Howell, who can generate unlimited original music for free, and listeners in blind tests cannot distinguish her compositions from human composers.
• 6.New job categories only appear at #33 on the list of jobs ranked by workforce size, meaning that despite hundreds of new job types since 1776, genuinely new work remains a negligible fraction of actual employment — undermining the standard argument that automation always creates enough new jobs to compensate.