What Happened to Horses Is Happening to Us. (11 year old prediction)

TL;DR: As mechanical brains replace mental labor, humans are experiencing a fate similar to that of horses being replaced by cars in the early 20th century—a large number of jobs disappear, and the new jobs created are insufficient to absorb the unemployed. ## Physical Labor In the past, humans reduced physical labor by creating tools: from sticks to tractors, mechanical muscles are stronger and more tireless than humans. This enabled specialization and economic growth. Now, mechanical brains (software robots) are reducing the demand for mental labor. Old-style automation was expensive, custom, and bulky industrial robots. The new representative of automation is **Baxter**—a general-purpose robot. - It learns operations through vision, costing less than the average annual salary of a human worker. - It can do any task within its arm's reach. - Similar to computers in the 1980s: not the peak, but the beginning. - Operating cost is just a few cents of electricity, while humans require minimum wage. Even at one-tenth the speed, it's still cost-effective because the price is one-hundredth. Baxter is smart enough to take over many low-skill jobs. Machines dumber than it have already replaced jobs: a modern supermarket needs only one human to supervise 30 self-checkout robots. Hundreds of thousands of baristas worldwide are threatened by barista robots—a robot network that can remember each customer's preferences. Technological change often comes from things that were around a decade ago becoming cheaper and faster. Robots are undergoing this process. ## The Luddite Horse Imagine two horses talking in 1900: one worries that mechanical muscles make horses redundant, the other says there are more city jobs and cars might bring new work. But looking back from 2000: horse numbers peaked in 1915 and then continuously declined. Economics has no rule that better technology creates more and better jobs for horses. Replace "horse" with "human," and people think it's reasonable. Just as mechanical muscles pushed horses out of the economy, mechanical brains will do the same to humans. Not immediately, but technology improves and becomes cheaper at a rate biology cannot match. ## Cars Self-driving cars already exist and work properly, having driven hundreds of thousands of miles. They don't need to be perfect; they just need to be better than humans: human drivers kill 40,000 people in the US alone each year. Self-driving cars don't blink, text, or get sleepy. Calling self-driving cars "cars" limits the imagination. They are actually **Autos**—tools for solving "moving objects from point A to point B." Micro-Autos can work in warehouses, mega-Autos can work in mines. The US transportation industry employs about 3 million people globally, about 70 million. These jobs have already peaked. A common argument is that unions prevent this, but workers always lose. The economics winner: labor costs account for about one-third of a transportation company's total costs. Accidents, carelessness all cost money. Insurance companies prefer drivers who never have accidents. Autos are coming, and this is the first area where most people will see robots change society. There are many similar situations in the economy. ## The Future Shape It's easy to think that technology only eliminates low-skill jobs, and white-collar workers are safe harbors—but software robots are more invisible, faster, and cheaper. White-collar workers are more expensive and more numerous, providing greater incentives for automation. An automation engineer's job is to replace your job with software robots. The frontier isn't super smart programmers writing robots; it's writing robots that can learn: give a robot a bunch of tasks and examples of correct completion, and it learns how to do them. Even with just goals, robots can learn. The stock market is no longer a human activity: it's mainly self-learning trading robots trading with other robots. The New York Stock Exchange trading floor is essentially a TV studio. Robots have learned to write: many newspapers already have robot-written sports news, TPS reports, quarterly reports. Document processing, decision-making, writing—demand in many human job categories is declining. ## Specialized Robots The bulk of a lawyer's work is drafting documents, predicting litigation outcomes, and discovery. Discovery in many law firms is no longer human work: robots scan millions of emails in hours with higher accuracy. IBM's Watson crushed humans on *Jeopardy!* but its main role is becoming the best doctor: understanding patient descriptions, giving accurate diagnoses, and it already provides lung cancer guidance at Sloan Kettering. Human doctors misdiagnose at alarming rates; understanding drug interactions exceeds human knowledge. Robots can learn from the experience of all robots, read the latest research, and find correlations. When a doctor robot is in your phone, demand for general practitioners will decrease. ## Creative Robots Creativity might feel like magic, but the brain is a complex machine. Artistic creativity isn't what most jobs rely on. The number of people who truly make a living from writing or performing is extremely small; an economy based on poetry and painting is impossible. Background music is composed by robot Emily Howell, and in blind tests, people cannot distinguish human from machine. Chess was once thought a uniquely human skill, but machines beat us. All human talents will follow. ## Conclusion These are not science fiction. Robots are working right now in labs and warehouses. Horses didn't become unemployed because they were lazy, but because they became unemployable. Many smart, fully capable humans will become "unemployable," through no fault of their own. New jobs won't save us. The 1776 US Census tracked a few types of jobs; now there are hundreds, but new jobs are not a significant part of the labor force. In the list of jobs sorted by number of workers, transportation tops the list. All these jobs existed a hundred years ago, and almost all are automation targets. Only at position 33 do we find a new job. The Great Depression had 25% unemployment. Just the automation of jobs already feasible in the list above (which accounts for 45% of the labor force) would exceed that number. In the modern miracle garden of technology, new jobs are not a significant part of the economy. Automation is not bad, but inevitable—a tool for creating abundance with little effort. When most of the population doesn't need to be productively employed, how does the existing economic system distribute abundance? This is the core question the video raises. Source: https://www.youtube.com/watch?v=7Pq-S557XQU