AUTOMATION
The standard [Anti-Luddite argument is that as jobs are taken by automation in one field people start working in another field. This is reasonably backed by history but there is no evidence to assume it will hold true indefinitely and significant evidence it is now not holding true anymore.
The introduction of the steam engine and industrial machines first affected agriculture. Automated thrashers, cotton pickers, combines rendering most agriculture jobs obsolete decades to centuries ago. Agriculture employed (or enslaved) over 70% of the American population back in 1800
Today it only accounts for 2% of the labor force, yet produced enough food for over 60 times the 1800 population (not including food relief or sales to other countries around the world). So if agriculture went from 70% to 2%, all the workers had to go somewhere and they did: into manufacturing. But by 1950 the so called golden age of US manufacturing 35% of the labor force was working in factories, today that number is 9%. Yet US manufacturing has only risen in production and is still the worlds largest manufacturing center in gross production. [http://www.shopfloor.org/2011/03/u-s-manufacturing-remains-worlds-largest/18756]
Today 80% of the labor force in the USA is in the “Service sector”, but automation is making inroads in that as well. In both Federico Pistono book "Robots will Steall Your Jobs but that's OK" and Martin Ford book "The Lights in the Tunnel: Automation, Accelerating Technology and the Economy of the Future" they provide a break down of the most common jobs in the USA today. For example Cashiers are 2.24% of all American jobs, but now self-checkout lanes are being installed and can cut down the number of cashiers by 75%. Truck drivers (the most common job in the USA) are 2.61% but automated trucks will likely start taking those jobs as well. Even if a truck driver is required in the automated truck as backup the driver can now sleep behind the wheel while the truck gets further without having to stop so the driver can sleep, allowing fewer drivers to do more deliveries. Secretaries (2.22%) are being replaced by digital assistance. But no these people will move on to new fields, so the anti-Luddite claims. What new fields are there? In the list of jobs from most common to least you have to go down to 34th on the list to get to a career that has not existed more then 70 years ago: Computer software engineer at 0.74% of the job market.
Well maybe people working in service today could get work in intellectual and skilled labor fields. Unfortunately there are in fact a plethora of inroads automation is making in those fields as well: pharmacist replaced by robots, software that is replacing paralegals and lawyers in document review, software that can write journalistic sports and basic business reports, image recognition software that replaces radiologist. It is argued that these things augment and assist people in these fields but if automation helps one person do more of a task then would it not mean that now less people need to do those tasks?
Another way to look at this is as a ratio of number of products produces divided by number of people needed to produce said products. Over the centuries this ratio has been going up in general for most products. As a result everyone has been able to get a job even with more and more automation because automation just means more products per producing person. Everyone can be productive because productivity was matched by demand, but can this stay true forever? Every product requires some amount physical resources and energy to make as well as demand from end users. Lets look at energy (as material resources are just a matter of how much energy is needed to to extract, refine and move them): For the last 3 centuries world energy consumption has been growing by ~3% per year. Right now it is at 15x10^12 watts (15 Terawatts), at 3% growth it would be at 150 TW by 2091, by 2345 our energy consumption would be equal to all the energy in sunlight hitting the earth (2.7x10^17 W)! Even if we could produce all that power by nuclear fusion or something the waste heat would exceed solar heating of the planet! The surface temperature of the earth would raise to boiling, literally boiling in just a few years! If we keep projecting 3% growth by the year 3058 humanities energy consumption would exceed the energy output of the entire sun (3.8x10^26 W)! And by the year 3760 humanity energy consumption would exceed the output of every single star in the galaxy (3.8x10^35 W)! Hence why there is no such thing as sustainable growth, it is PHYSICALLY IMPOSSIBLE! Pistono does a good job explaining this exponential growth problem but fails to really connect it to automation, I will now do that.
Thankfully world total energy consumption is starting to stagnate (actually falling in 2009!) and consistently falling in many developed countries. World population growth is slowing (and will hopefully plateau this century) and some developed countries even have negative growth. But though that may save us from physical limits of energy and matter its not good for production. How are we to produce more products if there is not enough power and raw materials to make them and consumers to buy them? If automation keeps growing the ratio but the number of products can't keep go up, then number of people needed to make them must go down.
Imagine for a moment a restaurant where the owner gets a machine to replace most if not all the workers, anyone that has worked at a McDonalds as a teenager can tell you that it would not be very hard to replace the assembly line of human burger makers with a machines that cooks the meat paddies, drops them on the bread buns, drops down pickles, lettuce, tomatoes and squirt of condiments. Now lets say the workers ran off to start restaurant of there own using the same model. Thus multiplying the number of restaurants, everyone gets a high paying job but would there be enough consumers for them all? Agriculture has experience several fold increases in products to producers ratio but the number of consumers has increase to match it so the ratio has increased much less if we look at it from a products to consumer ratio rather the products to producers ratio. As is Americans are eating way too much, they are morbidly obese and dying of diabetes and heart disease: we would need to institute real vomitoriums to get people to “consume” more food! But this scenario does not just play out with foods. Imagine cellphone makers replace all there labors with machines, and that magically those labors got the education to design new cellphones and manage new cellphone companies, with a production to R&D ratio of greater then 100:1, could we all stand to buy 100 cell phones to keep all those new cell phone companies afloat?
Now some may argue that automation can't take ALL jobs, well at present and for the near future that seems true. Most automation is done by at best Weak Artificial Intelligence: very “smart” at a specific task and completely incompetent at anything else. The [ulr=http://en.wikipedia.org/wiki/Watson_%28computer%29]IBM's Watson supercomputer and DeepQA program[/url] can and has trumped the best humans at answering even cryptic questions from memorized knowledge, it certainly smart at that and will likely replaced tech support workers and research aids, but it can't create new know data, or assemble a product or drive a car. Other WAI though can do each one of those task from evolving software that discovers new technologies, to image recognizing factor robots to self driving cars, but WAI can only do those tasks they are specifically programmed and built for: they can't think about or are even aware of "anything" else, they have no consciousness. A Strong Artificial Intelligence that could solve or adapt to any problem a human can and/or could qualify as conscious is likely decades away, and some theologians and philosophers say its impossible because of some supposed mystical and supernatural property of the human brain. But all that is irrelevant: machines don't need to take all jobs, just enough, fast enough, its simply a matter of automation taking jobs and incomes faster then they can be replaced.
Imagine machines will soon replace most service jobs, what jobs are left? Maintaining the machines? Sure but only a few people are needed to do that verse all the number of people who did those jobs the machines replaced. Ok how about jobs as engineers, biochemist, inventors... but those require extensive educations, above average intelligence and creativity. Ok how about artist and craftmen? Aside for the need for creativity and skill again, not enough people will be around to buy all those arts and crafts. There won't be more people having plumbing problems and needing more plumbers, having electrical problems and needing more electricians, heck if you have a computer problem these day you usually contact a machine first that can answer some of your problems rather then a paid specialist.
If automation enters an industry but the industry can't grow to match (because of physical limitation at most), then its work force must shrink, we have the historical drop of agriculture and manufacture jobs as proof of that. The profits that was taken to pay wages now goes up to the owners of the businesses, and the unemployed workers must compete in a shrinking job market, this would cause incomes to drop, middle class to shrink and become poor, the poor grow poorer and the rich to grow richer. That is exactly what is happening today! Stock markets and capital gains are higher then they have ever been before, yet unemployment is up, people are getting lower and lower paying part time jobs. The middle class is shrinking into poverty.
Much of this is blamed on outsourcing, a problem only in that its giving someone else “your” job. Outsourcing is a problem for the working population of developed nations, it drives salaries down to the lowest common denominators of the world. But truth be known machines are now competing with developing country's cheap practically slave labor! For example Foxcon the Taiwanese electronic giant that assembles all Apple products of all things is beginning to to automate. It all a matter of when will machines cost less then a practical slave in a undeveloped country. Also unskilled impoverished labors can't compete in quality verse the precision of automatons, nor will automatons commit suicide or demand pay raises.
This can't continue for long. If more products could be produced maybe their price would drop and the ever poorer people could still buy them but as I covered before there are fundamental limits in energy and materials that we appear to already be hitting that limit production. As the people get poorer they can't buy the products anymore. So demand will drop, and that means the rich will start to see profits drop. In fact the economy of individual nations and thus every nation because of globalization will become increasingly unstable and fall into recessions one after another. Oscillating wildly or possibly even entering depressions until solutions are implement.
and I will talk about the solutions soon enough, some of which could lead to a future as wonderful as today is to a black slave of 1800 America, but for now, does anyone disagree with the automation being the fundamental cause of our economic woes?
The standard [Anti-Luddite argument is that as jobs are taken by automation in one field people start working in another field. This is reasonably backed by history but there is no evidence to assume it will hold true indefinitely and significant evidence it is now not holding true anymore.
The introduction of the steam engine and industrial machines first affected agriculture. Automated thrashers, cotton pickers, combines rendering most agriculture jobs obsolete decades to centuries ago. Agriculture employed (or enslaved) over 70% of the American population back in 1800
Today it only accounts for 2% of the labor force, yet produced enough food for over 60 times the 1800 population (not including food relief or sales to other countries around the world). So if agriculture went from 70% to 2%, all the workers had to go somewhere and they did: into manufacturing. But by 1950 the so called golden age of US manufacturing 35% of the labor force was working in factories, today that number is 9%. Yet US manufacturing has only risen in production and is still the worlds largest manufacturing center in gross production. [http://www.shopfloor.org/2011/03/u-s-manufacturing-remains-worlds-largest/18756]
Today 80% of the labor force in the USA is in the “Service sector”, but automation is making inroads in that as well. In both Federico Pistono book "Robots will Steall Your Jobs but that's OK" and Martin Ford book "The Lights in the Tunnel: Automation, Accelerating Technology and the Economy of the Future" they provide a break down of the most common jobs in the USA today. For example Cashiers are 2.24% of all American jobs, but now self-checkout lanes are being installed and can cut down the number of cashiers by 75%. Truck drivers (the most common job in the USA) are 2.61% but automated trucks will likely start taking those jobs as well. Even if a truck driver is required in the automated truck as backup the driver can now sleep behind the wheel while the truck gets further without having to stop so the driver can sleep, allowing fewer drivers to do more deliveries. Secretaries (2.22%) are being replaced by digital assistance. But no these people will move on to new fields, so the anti-Luddite claims. What new fields are there? In the list of jobs from most common to least you have to go down to 34th on the list to get to a career that has not existed more then 70 years ago: Computer software engineer at 0.74% of the job market.
Well maybe people working in service today could get work in intellectual and skilled labor fields. Unfortunately there are in fact a plethora of inroads automation is making in those fields as well: pharmacist replaced by robots, software that is replacing paralegals and lawyers in document review, software that can write journalistic sports and basic business reports, image recognition software that replaces radiologist. It is argued that these things augment and assist people in these fields but if automation helps one person do more of a task then would it not mean that now less people need to do those tasks?
Another way to look at this is as a ratio of number of products produces divided by number of people needed to produce said products. Over the centuries this ratio has been going up in general for most products. As a result everyone has been able to get a job even with more and more automation because automation just means more products per producing person. Everyone can be productive because productivity was matched by demand, but can this stay true forever? Every product requires some amount physical resources and energy to make as well as demand from end users. Lets look at energy (as material resources are just a matter of how much energy is needed to to extract, refine and move them): For the last 3 centuries world energy consumption has been growing by ~3% per year. Right now it is at 15x10^12 watts (15 Terawatts), at 3% growth it would be at 150 TW by 2091, by 2345 our energy consumption would be equal to all the energy in sunlight hitting the earth (2.7x10^17 W)! Even if we could produce all that power by nuclear fusion or something the waste heat would exceed solar heating of the planet! The surface temperature of the earth would raise to boiling, literally boiling in just a few years! If we keep projecting 3% growth by the year 3058 humanities energy consumption would exceed the energy output of the entire sun (3.8x10^26 W)! And by the year 3760 humanity energy consumption would exceed the output of every single star in the galaxy (3.8x10^35 W)! Hence why there is no such thing as sustainable growth, it is PHYSICALLY IMPOSSIBLE! Pistono does a good job explaining this exponential growth problem but fails to really connect it to automation, I will now do that.
Thankfully world total energy consumption is starting to stagnate (actually falling in 2009!) and consistently falling in many developed countries. World population growth is slowing (and will hopefully plateau this century) and some developed countries even have negative growth. But though that may save us from physical limits of energy and matter its not good for production. How are we to produce more products if there is not enough power and raw materials to make them and consumers to buy them? If automation keeps growing the ratio but the number of products can't keep go up, then number of people needed to make them must go down.
Imagine for a moment a restaurant where the owner gets a machine to replace most if not all the workers, anyone that has worked at a McDonalds as a teenager can tell you that it would not be very hard to replace the assembly line of human burger makers with a machines that cooks the meat paddies, drops them on the bread buns, drops down pickles, lettuce, tomatoes and squirt of condiments. Now lets say the workers ran off to start restaurant of there own using the same model. Thus multiplying the number of restaurants, everyone gets a high paying job but would there be enough consumers for them all? Agriculture has experience several fold increases in products to producers ratio but the number of consumers has increase to match it so the ratio has increased much less if we look at it from a products to consumer ratio rather the products to producers ratio. As is Americans are eating way too much, they are morbidly obese and dying of diabetes and heart disease: we would need to institute real vomitoriums to get people to “consume” more food! But this scenario does not just play out with foods. Imagine cellphone makers replace all there labors with machines, and that magically those labors got the education to design new cellphones and manage new cellphone companies, with a production to R&D ratio of greater then 100:1, could we all stand to buy 100 cell phones to keep all those new cell phone companies afloat?
Now some may argue that automation can't take ALL jobs, well at present and for the near future that seems true. Most automation is done by at best Weak Artificial Intelligence: very “smart” at a specific task and completely incompetent at anything else. The [ulr=http://en.wikipedia.org/wiki/Watson_%28computer%29]IBM's Watson supercomputer and DeepQA program[/url] can and has trumped the best humans at answering even cryptic questions from memorized knowledge, it certainly smart at that and will likely replaced tech support workers and research aids, but it can't create new know data, or assemble a product or drive a car. Other WAI though can do each one of those task from evolving software that discovers new technologies, to image recognizing factor robots to self driving cars, but WAI can only do those tasks they are specifically programmed and built for: they can't think about or are even aware of "anything" else, they have no consciousness. A Strong Artificial Intelligence that could solve or adapt to any problem a human can and/or could qualify as conscious is likely decades away, and some theologians and philosophers say its impossible because of some supposed mystical and supernatural property of the human brain. But all that is irrelevant: machines don't need to take all jobs, just enough, fast enough, its simply a matter of automation taking jobs and incomes faster then they can be replaced.
Imagine machines will soon replace most service jobs, what jobs are left? Maintaining the machines? Sure but only a few people are needed to do that verse all the number of people who did those jobs the machines replaced. Ok how about jobs as engineers, biochemist, inventors... but those require extensive educations, above average intelligence and creativity. Ok how about artist and craftmen? Aside for the need for creativity and skill again, not enough people will be around to buy all those arts and crafts. There won't be more people having plumbing problems and needing more plumbers, having electrical problems and needing more electricians, heck if you have a computer problem these day you usually contact a machine first that can answer some of your problems rather then a paid specialist.
If automation enters an industry but the industry can't grow to match (because of physical limitation at most), then its work force must shrink, we have the historical drop of agriculture and manufacture jobs as proof of that. The profits that was taken to pay wages now goes up to the owners of the businesses, and the unemployed workers must compete in a shrinking job market, this would cause incomes to drop, middle class to shrink and become poor, the poor grow poorer and the rich to grow richer. That is exactly what is happening today! Stock markets and capital gains are higher then they have ever been before, yet unemployment is up, people are getting lower and lower paying part time jobs. The middle class is shrinking into poverty.
Much of this is blamed on outsourcing, a problem only in that its giving someone else “your” job. Outsourcing is a problem for the working population of developed nations, it drives salaries down to the lowest common denominators of the world. But truth be known machines are now competing with developing country's cheap practically slave labor! For example Foxcon the Taiwanese electronic giant that assembles all Apple products of all things is beginning to to automate. It all a matter of when will machines cost less then a practical slave in a undeveloped country. Also unskilled impoverished labors can't compete in quality verse the precision of automatons, nor will automatons commit suicide or demand pay raises.
This can't continue for long. If more products could be produced maybe their price would drop and the ever poorer people could still buy them but as I covered before there are fundamental limits in energy and materials that we appear to already be hitting that limit production. As the people get poorer they can't buy the products anymore. So demand will drop, and that means the rich will start to see profits drop. In fact the economy of individual nations and thus every nation because of globalization will become increasingly unstable and fall into recessions one after another. Oscillating wildly or possibly even entering depressions until solutions are implement.
and I will talk about the solutions soon enough, some of which could lead to a future as wonderful as today is to a black slave of 1800 America, but for now, does anyone disagree with the automation being the fundamental cause of our economic woes?
Last edited:
