How AI will conquer the job market
Artificial intelligence (AI) has been a topic of discussion for many years, with its impact on society being both positive and negative. While many have touted its ability to automate tasks and increase efficiency, one of the most significant, even though often overlooked, concerns about AI is its impact on employment.
As the world continues to advance technologically, the question of whether AI will lead to widespread unemployment looms larger. The fear is growing that AI is encroaching in areas where human abilities were once deemed indispensable, threatening to do for cognitive ability what machines did for muscle power. With studies suggesting that up to 50% of current jobs could be overtaken by AI and automation, it’s hard not to wonder whether computers will soon put a lot of people out of work. In this article, we delve into the dark side of AI’s impact on employment and examine the evidence suggesting that AI will increase unemployment.
“There’s no economic law that says ‘You will always create enough jobs or the balance will always be even’, it’s possible for a technology to dramatically favour one group and to hurt another group, and the net of that might be that you have fewer jobs.”
Erik Brynjolfsson, Director of the MIT Initiative on the Digital Economy
To destroy, transform or create
Various studies have been done about the effects of AI on employment. In a report published by The World Economic Forum, it is estimated that about 75 million jobs may be lost directly due to AI and automation by 2025. The report also states specifically industries such as manufacturing and retail are at risk of being completely automated. Moreover, a study by Forrester Research predicts that automation will eliminate 16% of jobs in the US alone by 2025, while only creating 9%; resulting in a net loss of 7%. These are predictions for the near future, but these numbers can grow quickly the more and more AI develops. Researchers at the University of Oxford, for instance, report that approximately half of all jobs may be at risk at a later stage. Jobs like clerks, telemarketers, paralegals, cooks, waiters, receptionists, bank tellers, security guards, data analysts, tax preparers, and many more.
Nobody is Safe
However, the risk of obsoleteness is not limited to these occupations. The risk that highly educated and skilled professionals in the top-income classes, such as doctors, architects and even programmers themselves, could be displaced by future AI-powered machines or programs is ever increasing too. To illustrate is in the figure on the right the estimated risk of automation visualised per job type by the researchers of Oxford.
Recently, a new artificial application shocked the world. In a matter of seconds, the artificial intelligence program ChatGPT can generate a wide variety of content, such as emails, business plans, company logos, computer codes, newspaper stories, academic studies, and even works of fiction in the vein of Joyce or Dostoyevsky. It is able to perform jobs that appeared to require the skills of humans with significant formal education. Additionally, another relatively new AI neural network tool named DALL-E is able to create and edit incredible unseen images from natural language instructions. Even the creative sector is slowly but surely being infiltrated by AI applications and creative artists already are genuinely concerned regarding their livelihoods.
“Right now when you type in my name, you see more work from the AI than work that I have done myself, which is terrifying for me. How long till the AI floods my results and is indistinguishable from my works?”
Greg Rutkowski, commercial illustrator in the gaming industry in Forbes interview
Job loss does not tell the whole story
There is more to measuring the AI takeover than just the unemployment statistics. AI won’t only replace whole jobs, but will also be used to only partially take over careers. Where a human would have done 100% of a task, it now would do a proportion while the computer would do the rest, which is then portrayed in salary too as less manual work is required.
In combination with whole replacement will this exacerbate the decline of middle-skilled jobs and rising wage inequality observed in the recent past, particularly in many developed countries. According to a new academic research study, automation technology has been the primary driver of U.S. income inequality over the past 40 years. The report, published by the National Bureau of Economic Research, claims that 50% to 70% of changes in U.S. wages, since 1980, can be attributed to wage declines among blue-collar workers who were replaced or degraded by AI or automation.
How about new jobs?
Although it is popular belief that in any upcoming industry new businesses are created that are in need of numerous employees, one might be surprised how this fails to solve the unemployment problem. As AI startups emerge, they are creating new opportunities for people to carry out new jobs that were not previously possible. Black has shown, for instance, an increasing demand for experts in AI development, data science, and machine learning. These occupations, however, are quite distant from the jobs that are in danger of replacement as discussed earlier. Those that are replaced will not be able to just switch to one of the AI-related jobs; the skill gap is simply too big. It is expected that this may lead to unemployment still, as workers are unable to apply for the jobs that are made available in the new economy.
Retraining and Exponential Growth
When the skill gap is too big for immediate transfer, one could argue that people just need to be educated or retrained to meet the requirements. Additionally, the education and retraining for these positions could even create a whole demand for a newly increased job market in education, indirectly creating more jobs that can be filled.
Although this might offset some job loss and help some people switch to AI-related jobs, it won’t be able to catch up fast enough: As predicted by Gorden Moore back in 1965, the transistors on a microchip are doubling approximately every two years and the cost of computers halving. Alongside other predictive laws like Wright’s Law (cost decline) and Kryder’s Law (storage increase), processing power for computers has seen and is expected to continue exponential growth. Since AI itself is completely powered by computers, it logically follows that advancements in AI technology follow a similar exponential trajectory; a pace that, as is supported by the research of Cukier, our education systems can simply not compete with.
What to do?
In conclusion, while AI has the potential to bring many benefits to society, its impact on employment is a cause for concern, with studies suggesting that up to 50% of current jobs may be overtaken by AI and automation. Low- and medium-skilled workers are expected to face the most pressure, while highly educated and skilled professionals, even those in the top income classes, may also be displaced by future AI-powered robots. Although AI startups are creating new job opportunities, the skill gap between jobs being replaced and new AI-related jobs may be too big for immediate transfer. While education and retraining may help offset some job loss, it may not be enough to catch up with the fast pace of technological change.
Now, as this article shows, the risk of mass unemployment by AI can not be overstated. However, how we will let this impact our society is for us to decide. It is imperative that we consider the social and economic consequences of AI and find ways to ensure that the benefits of automation are shared equitably. This should include exploring alternative models of work, income, and welfare that can support all people in an AI-driven future. As a society, it is our responsibility to work towards a future where AI creates new opportunities rather than perpetuating poverty and inequality. Let us use AI as a tool to empower and uplift humanity, not to leave people behind.