What Artificial Intelligence Brings to the Job Market?

Within the next 15 years, almost half of all American jobs could be automated. What makes this new wave of computerization different from the past is that before the industrial revolution, innovation was done to increase productivity and made human life easier. Goods and services’ production rate increased as more innovation came about using the same amount of human work. Innovation would eliminate jobs, but also create better jobs for the market.

A clear progression can be seen; initially, we started off as a mostly agricultural society, to one of production after the industrial revolution, then to service jobs as automation became more widespread. Now, we have the rise of the information age. We look around to see jobs being taken over by machines at an even faster rate than before; the information age is booming, but seemingly creating fewer jobs. For example, in 1979 General Motors employed roughly 800 thousand workers and made 11 billion USD in revenue. Fast forward to 2012 and Google made 14 billion USD with 58 thousand. This may seem like comparing apples to oranges, but Google is an example of what created new jobs in the past – innovative new industries.

Old industries are not providing as much to the market as they used to when first created. Cars for example shaped the America we know today by creating infrastructure and cities to accommodate them. Millions found jobs either directly or indirectly because of cars, and decades of investment enabled this. Today, innovations within the car industry do not create as many jobs as it used to. There is much work to be done concerning these innovations, with AI involved. Self-driving cars and new detection technologies would enhance the quality of life by overall reducing accidents and loss of life, but they won’t create the millions of jobs as the original car did. The internet on the other hand, has created new industries; however, they are not creating as many jobs accounting for population growth or to accommodate for the industries that are suffering because of the internet. YouTube is a fine example of this with independent news organizations reaching millions of viewers, using only a fraction of the number of people it would take for a standard news studio on television to reach the same amount of people.

Human progress is based on the division of labor, as we have advanced, our jobs have become more specialized. Automation is good at narrowly defined and predictable jobs, this is what has destroyed factory jobs. But now even some complicated jobs can be broken down to their bare components that could be automated by several machines, leaving no room for people to specialize. Digital machines are able to learn and teach themselves by analyzing data several times over, such as algorithms. And in recent years, humans have started to gather data about everything; behavior, weather patterns, medical records, communication systems, travel data, and even personal data. This has created an archive of information of how humans do things, and machines can take this information to do these things even better. We can witness this happening even today.

A company based in San Francisco started using a project managing software to eliminate middle management positions. The software delegates projects that could be automated and decides where it would need professional human intervention. This caused a hiring of freelancers to work on these projects, tracking individual performance until completion of the project. So, while getting rid of middle management, it  created jobs for the freelancers. However, these jobs were tracked, allowing learning algorithms to gather data about the work and which tasks they consist of, allowing the software to eventually not need to hire the freelancers and automate the work. This software allows many costs to be cut as well, from 50% during its first year, then another 25% in its second.

While this is only one example of many, it is evident that, slowly over time, fewer humans will be doing jobs like this and others requiring less skill. Job replacement is not the only thing that needs to happen, but job growth as well. A growing population is dependent on job creation, in the past this was solved through innovation. But since 1973, the generation of new jobs in the US has begun to shrink and from 2000-2010 was the first decade in which the total amount of jobs in the US did not grow. In order to keep up with population growth, 150 thousand jobs must be created per month, meaning this bad news. This depletion of jobs has also started affecting the standard of living. In the past, it was believed that more productivity would lead to more jobs and better standards of living. However, in 1998, US workers worked a total of 194 billion hours and in the next 15 years, output increased by 42%, but with the same 194 billion hours of work in 2013.

Productivity, business, and the population had all increased; but the total number of hours worked, had not. While at the same time, wages for new graduates had steadily decreased and tuition increased – forcing many to take on jobs that did not require degrees. Productivity is separating from human labor. Innovation in this information age is different from everything we’ve encountered in the past. This process started years ago, but it seems as if automation is different this time.

Our economy is based on consumers, but if fewer have decent work, who will be doing the consuming? Are we are producing more things ever cheaply for only too few people that can buy the goods and services? The information age and modern automation be an opportunity to change human society and shape the future for the better. Reducing poverty and pollution are things that could be done with the proper automation and policy, and one thing is for sure; the machines are not coming, they are already here.