In the year 1900 forty percent of the total United States labor force was farming. Today, this number has dropped to less than two percent, yet everyone is still fed and a greater fraction of U.S. adults are in the workplace now than in 1900. This massive increase in productivity was made possible via the invention of tractors and other technologies that mechanized many facets of the farming process reducing the human labor necessary to accomplish the same, and even more, food outage. Many of the greatest human inventions have been created to reduce human labor, from the printing press and tractors to self-checkout-machines and drones. This process is known as automation, and, like farming, it has transformed countless industries and replaced millions of workers throughout history. Automation is the essence of technology. Currently, it is estimated that up to 300 million jobs worldwide will be automated by 2030 with approximately one-third of U.S jobs at risk of being made obsolete. What will happen to the one-third of the U.S working class population when robots and software being to replace them? Many economists are beginning to raise red flags, warning that near-future automation will cause mass unemployment and a potentially radical socio-economic divide.
This is not the first time this concern has been raised, however. Articles can be found in the New York Times from the late 1920s and the early 30s titled like "March of the Machine Makes Idle Hands." There was another surge in automation anxiety in the late 1950s with President Kennedy reporting automation as being the #1 "Job challenge." Historically the process of automating mundane and often times dangerous human work has been beneficial to society by raising living standards, creating better jobs, and giving people greater freedom to be creative and therefore further improve life for themselves and others. This is explained by the Harvard Business review "We have found that automation often results in higher compensation for workers, with those costs often being offset by greater productivity." For example, the invention of the personal computer made millions of jobs obsolete; however, it created entirely new and previously unimagined industries. Ultimately this generated roughly 18.5 million new jobs, even after accounting for jobs lost.
The same story applies to railroad workers, factory workers, telephone operators, gas pumpers, elevator attendees, travel agents, milk-men, and the list goes on and on. What becomes clear by looking at these numerous previous examples of automation is that it much easier to see the jobs being replaced by robots than it is to predicate and imagine the jobs that will be created afterward. As was almost comically phrased by economist David Autor "These predictions strike me as arrogant. These self-proclaimed oracles are in effect saying, 'If I can't think of what people will do for work in the future, then you, me and our kids aren't going to think of it either.' I don't have the guts to take that bet against human ingenuity. I can't tell you what people are going to do for work a hundred years from now, but the future doesn't hinge on my imagination."
Many economists persist that historical evidence is not applicable though because "things are different" this time. This is explained in the Kurzgesagt video Why Automation is Different this Time when it says "While even our smartest machines are bad at doing complicated jobs, they are extremely good at doing narrowly defined and predictable tasks" and this they claim is what has previously limited the reach of automation and has allowed many humans to work in more creative and cognitively demanding jobs. There is, however, a new type of technology emerging that allows machines to do more complex and "cognitively demanding" jobs: artificial intelligence and machine learning. Many are unaware of, or are unwilling to, understand the incredible scope of this software and the level of automation that it can bring. AI and machine learning technology are capable of predicting the revenue of a movie using nothing but its script, best humans in the game of GO or Jeopardy. The "new" automation that can be created using these complex softwares, some claim, will create additional productivity but will fail to create new jobs and keep up with population growth. This new type of automation, some fear, could also create a vast socio-economic divide where a small rich upper-class control vast swaths of these robots and softwares.
There are two fundamental economic principles that refute this idea, or at the very least place it into perspective. The first was coined by Harvard economist Michael Kremer as the O-ring principle. The O-ring principle "presents a production function in which production consists of many tasks, all of which must be successfully completed for the product to have full value" and that if the reliability of one task increases than the value of the reliability of all the other tasks goes up. When these new types of software and robots begin to automate many of today's jobs or facets of them, the roles that humans play in creating valuable companies will increase. One small-scale example of this happening today is tellers. The ATM (Automated Teller Machine) was introduced 45 years ago and yet today there are twice as many tellers as there were at its introduction. The role of the teller "became less like checkout clerks and more like salespeople, forging relationships with customers, solving problems and introducing them to new products like credit cards, loans and investments: more tellers doing a more cognitively demanding job" as is explained by Dr. Autor. This principle, the O-ring principle, ensures that humans will always be needed to create great and functioning businesses and that the more other jobs are automated the more valuable human creativity and ingenuity will be needed.
Clearly, however, automation is capable of eliminating some types of jobs. Few people today have heard of long-shore-men. Why? Because for the most part they no longer exist. New forms of technology and automation destroyed that job. What happens when nearly all of the current jobs have been automated? The second economic principle that further refutes this conjecture is that material abundance has never stymied perceived scarcity. In order for those living today to achieve the same average living standards of 1915, they would need to work just 17 weeks a year. Most people, however, do not choose to do this, most people choose to work harder than this to leverage the enjoyments of the modern world. The majority of the United States' economy is driven by non-necessary, luxury, items. Basic human needs in the United States, with the rest of the less-developed world rapidly catching up, are taken care of for the majority of people. As technology and automation further mature they free up our time, further the realm of possibilities and "spur consumption." Extreme sports, yoga, travel and tourism, podcasts etc. are all examples of trivial activities, however, because of perceived scarcity, people are willing to spend money on these things. Autor David showcases the reality of this principle when he said "If I were a farmer in Iowa in the year 1900, and an economist from the 21st century teleported down to my field and said, 'Hey, guess what, farmer Autor, in the next hundred years, agricultural employment is going to fall from 40 percent of all jobs to two percent purely due to rising productivity. What do you think the other 38 percent of workers are going to do?" I would not have said, 'Oh, we've got this. We'll do app development, radiological medicine, yoga instruction, Bitmoji.'"
Regardless of the large swaths of automation that have already occurred there are currently more jobs available in the U.S than there are unemployed. * Automation will not create a world where there are no jobs. There will be jobs. But not all jobs are created equal. Due to an economic phenomenon known as employment polarization, the number of middle-class jobs is shrinking and will most likely continue to decrease. This is because these are the jobs most vulnerable to automation. These are the jobs that have the most easily interpreted rules, procedures, and repetitive tasks vulnerable to the increasingly advanced machine learning software. This poses a problem for the incoming working class and calls for serious consideration of the nature of future jobs.
Automation has been, with technology, developing for thousands of years, making our lives easier every step of the way. The would-be-farmers of the 1930s needed more education to work in the new industrial world of the 1920s. In the same way, today's workers will need to have more education to work in tomorrow's, more automated, world. The United States heavily invested in High School education in the 1920s to create a class of more specialized and cognitively capable workers that would be necessary to maintain a healthy and employable economy. As is explained by The Verge, "In 1910, only 18 percent of children aged 14 to 17 went to high school; by 1940 this figure was 73 percent." Today High School attendance is essentially a given, but at the time it was a major economic transition for the United States. This meant taking young working children out of the field and into classrooms around the United States.
A similar economic transition will be happening in the United States within the next two decades, however, this time technology is maturing much faster. A chart showing the increase in the number of transistors that computer engineers have managed to squeeze onto a computer chip over time, and therefore the increase in computing power over time, grows exponentially. So it will be with automation. According to a study conducted by McKinsey Global Institute, 60% of all current jobs will be at least one-third automated by 2030 with potentially one-third of hours worked globally being automated by the same year. Technology and automation have never advanced this far this quickly before, however, the potentially negative effects of future automation happening this immediately can be mitigated if governments have an active part in the process. The paper continues by comparing the level of action needed to ensure we are prepared for this impending economic transition to the Marshall Plan and saying that those who will most likely need job re-training will not be "the young, but middle-aged professionals."
In the same way, the United States invested heavily in High School education in the early 1900's we need to be prepared to invest in higher education. College students, and more importantly their universities, need to be made aware of the near future automation. Universities need to begin shifting their majors, degrees, and curriculums to focus on skills that are immune to this coming revolution and focus on arming their students with skills that will make them valuable in a post-AI-automation world. Doing this will ensure that their alumni are always employable. Universities need to refrain from teaching their students anything that can be performed by a robot or AI-computer. One example of this is the changes that are currently taking place in the engineering space. As Maurice Conti explains in his TED talk, AI computer programs have been created that design 3D models for anything from drone and car chassis to bridges and plane parts. The computer is simply given the constraints for the model and objectives, such as lateral strength and lightweight or heat dispersing and aerodynamic; then computer uses a process modeling natural selection and evolution to come up with an often strange looking, yet effective design, that never could have been imagined by humans. So, for example, many universities may opt to spend less time teaching engineers how to create 3D models, a skill that will likely be automized and improved, and more time teaching how to communicate effectively, problem solves, and prototype. Similar instances can be found in many other fields of study. Colleges need to be proactive in addressing these changes so that the next generation of people in the professional workplace are prepared to contribute to their companies and communities.
A minority think that a new type of economy never before seen may emerge in the wake of automation. An economy based on arts. I believe that this is misguided. At art's economic roots artists make money because their creations are popular, more popular than other artists work. There is intrinsically only so much human attention to go around, we can only have so many paintings and listen to so much music, and therefore this type of economy can never exist. In contrast material economies and services do not require human attention. The belief in a future art-based-economy is passive and will be detrimental to efforts preparing for the future automized world.
Others believe that we should try and inhibit the advance of automation by placing regulations on particular technologies: such as banning robots. Evolution gambled on humans and their ability to create tools and technology. That gamble is what eventually set us apart from all other species. The future of automation has many potentially great benefits, just like tractor automation gave many people the freedom to specialize in their new fields and discover new beneficial technologies. One great example of this is the potential benefit of automation in the medical field. The primary purpose of IBM's Watson is to become the greatest medical researcher and diagnoser in the world. Watson could do this by keeping track of every patient's complex medical history, understanding every drug's reaction with every other drug, learning from medical mistakes all over the world, constantly updating its knowledge with the most current medical research, and finding hidden correlations by tracking what happens to all of its patients all over the world. Many of these things are beyond human knowability and would be impossible for a human to do. Watson could potentially be a life-saving technology considering that an estimated 98,000 Americans die every year due to medical errors according to Maggie Fox at Reuters. IBM believes that many more could be saved because of correlations that Watson could find and advances it could bring to the medical industry. This technology may seem far-fetched, but Watson is already being tested at Slone-Kettering giving guidance on lung cancer treatments as described by a CGP-Grey video titled "Humans Need Not Apply." This is ultimately the goal of technology: to create better lives by reducing the amount of human work that is necessary to sustain life, and, in the near future, doing work that humans cannot do. Limiting automation is synonymous with limiting technological advancement.
When economists are asked what they believe the impact of automation will be on human-job-prospects over the next 15 years their responses are typically binary, but there is one thing that everyone agrees on: automation is coming, just like tractors and the internet, and it will change everything. In the same way that it is difficult to predict the nature of future jobs, it is difficult to predict all of the economic implications that automation will have, but a few things are clear. We should prepare to learn how to work in a further automized world, among the software and hardware robots of the future. For millions of years, we have had to force our will into our tools. Even the most complicated tools have required manual human input every step of the way. In the near future computers will give us feedback and ideas, allowing us to work with them to create things we otherwise couldn't. Furthermore, we need to be vigilant and plan to invent ways that ensure automation will benefit everyone. The impending automation revolution can be an incredible boon to the furthering of human health, and comfort, just like the agricultural, industrial, information, and technological revolutions before it, so long as we prepare for and treat it correctly.