Will AI Take Your Job? You Better Believe It

When it comes to the future of work, there is only one certainty. Anyone who tells you that they know what is going to happen is either lying to you or trying to sell you something. We are in uncharted territory. And there is a lot of uncertainty. What we do know is that a lot is going to change.

When I first started working on AI, there were three major camps when it came to thinking about jobs:

Before ChatGPT, most people didn’t actually know much about how quickly deep learning was progressing. So position three was a pretty popular stance, and those who believed in option one were considered insane. But things have changed a lot in the past few years.

Position three is now rarely discussed. It is just assumed that AI will have a massive impact on the labour market. And although position one is still quite taboo in academia, it is generally accepted as a possibility. But most researchers, myself included, fall into the category of position two.

AI, in combination with advanced robotics research, has put a lot of jobs under the pressure of potential automation. Already, AI is being used in manufacturing, health care, therapy, management, programming, the arts, and construction, to name a few—it has basically penetrated every industry. In 2013, a now infamous Oxford University study predicted that approximately 47 percent of American jobs would be at risk of automation in the near future. The researchers may have been a little ambitious, but right now it’s not feeling like they were that far off.

The reality is this: a lot of tasks that were traditionally considered to be safe from automation are no longer safe. Jobs that require artistic abilities or cognitive skills are now being automated. At the same time, the ability of machines to perform physical tasks is increasing as robots are fitted with AI systems that allow them to better sense and react to the world around them. AI’s automating powers are indiscriminate. They are affecting blue-collar manufacturing jobs and white-collar office jobs. Many who spent years, and thousands of dollars, developing specialized skills now need to live with the fact that AI can do their job faster and often better. It is a terrifying reality.

What makes it even scarier is that we have no clue exactly how it is going to play out. There are just too many variables.

For example, it is possible that a lot of companies will put off automating their staff and instead opt for retraining their current employees for new positions or finding ways to have them work with AI. This approach could spread out the potential economic disaster that would occur if there was suddenly mass unemployment due to rapid automation. Mass unemployment is bad (obviously) for a number of reasons. Economically, it presents a gigantic threat. If companies are producing more goods than ever due to automation, who’s going to buy them? There’s mass unemployment, and the people don’t have money to support businesses.

Whether or not companies take this approach is not really in our hands. Even if they do, it’s possible that an unrelated financial recession will occur, forcing companies to lay off their employees, as is normal during a recession. However, this time, it may be different as they offset these layoffs by bringing in automation technology. Now, when the recession ends, the companies have no incentive to rehire their old employees.

The prefabricated argument often used to confront fears of automation is that innovation inevitably creates new jobs. Although some jobs may be eliminated, the economic benefits of automation will create incentives for companies to create new jobs elsewhere in the economy; therefore, there is not actually anything to worry about!

As much as I would love to believe in this take, as it is so comforting, it unfortunately does not hold up when put under the microscope. There are several major problems that emerge under scrutiny, even if we assume that this statement is partially true and that new jobs will materialize that we cannot foresee right now. The first problems are to do with speed and scale.

When we typically talk about automation or innovation, it is in a very narrow sense. There may be a new innovation that automates a specific job in the dairy industry or in stock trading. But these instances are isolated. The automation in the dairy industry is not directly transferable to other industries. An automatic cow-milking system doesn’t have disruptive powers in other industries. However, the cost savings produced by the innovation could affect other parts of the industry. As production rises, the need for new jobs in other areas of the dairy industry may be created, offsetting the lost jobs milking cows.

But AI is not a one-dimensional technology. It isn’t even a single piece of technology. It is a collection of technologies and methods for collecting and processing data that can be applied in numerous ways. It is not contained to a sole application. It is what’s known as a general-purpose technology. Other examples of general-purpose technologies are the internet, computers, and, most notably, the steam engine.

General-purpose technologies cut across industries. They do not have a sole purpose but instead can be applied to an infinite number of tasks. They become productive forces that humans interact with, and leverage, to create and innovate.

As it is a general-purpose technology, AI has the ability to disrupt countless industries at once. Where efficiencies in one area may have, at one time, resulted in new jobs in another, it is now possible that the other area will be experiencing the same efficiencies due to AI. Therefore, they may be creating some new jobs, but they will also be eliminating others. Although new jobs will be created to deal with AI, there is no certainty that they will replace all the jobs that are lost from the potential coming wave of automation. But scale is not the only problem with AI’s automation effects. Speed is also a major fear.

AI developed in the shadows for a substantial period. Now that it has arrived in the mainstream, it seems to have entered with significant momentum, which means that it may hit countless segments of the economy fast and hard. Lots of people across industries may lose their jobs in a very short period of time. The creation of new jobs for the now unemployed workers will not happen as fast as the workers lose their jobs. In other words, if AI automation hits the job market hard and fast, we need to be prepared for a prolonged period between the layoffs and the emergence of new jobs. This period is likely to be both economically and mentally scarring for those who get caught in the middle. Sadly, this situation is not where the problems end.

Again, even if we accept that new jobs will appear, and even if a miracle occurs and they appear jaw-droppingly fast, there is still a problem. The idea that new jobs will emerge brushes over the question of whether the new jobs will be appropriate for those who have lost their jobs. Not all jobs are the same.

Due to the nature of AI, it is likely that it will heavily affect workers who are deemed to be “skilled workers.” This does not only mean people with a university education. It also includes those who perform tasks that traditionally require a large amount of acquired knowledge that is not common. This knowledge can be acquired through education or on-site experience. Workers are skilled because they have invested in themselves, acquiring this specialized knowledge. But now AI is able to acquire and act on the same knowledge.

For many, massive investments they have made into acquiring knowledge and skills may be wiped out overnight. That doesn’t just mean financial investments in formal education but also years of practical work and experience that previously led to them being deemed skilled or valuable in the job market. The skills that will be needed for many of the new jobs that are created in a post–AI automation world will require fundamentally different skills to those which multiple generations have trained for.

For many, massive investments they have made into acquiring knowledge and skills may be wiped out overnight.

The skills gap that may be created could produce gigantic economic and social problems. Those who have their skills completely devalued by the economy will either need to retool or take low-skilled jobs. Low-skilled jobs will not be able to provide a standard of living or social status similar to what the previous jobs did. This downgrading will almost certainly lead to adverse mental effects and resentment of the system that betrayed these workers.

That is doubly true for those who feel as if they did everything right. They did everything they were told to do to gain financial security, only to have it ripped out from under their feet. Imagine being forty-five years old with a young family and twenty-two years of work experience. Then, suddenly, you are told you need to start from the bottom. Your skills are no longer valuable. The further along you are in your career, the more devastating needing to retrain can be.

However, it is not just those who are far along in their careers that are at risk. University education has long been positioned as the fastest and most efficient way to gain knowledge and skills that can set you apart. In countries such as the United States and Canada, students often take on substantial debt to fund their education. The rise of AI means that many skills taught at universities, especially in professional programs such as business, are being automated. This means that recent graduates may be saddled with massive amounts of debt for training that is now practically useless.

For many, the option of retooling without specialized aid from the government will be impossible. Since they still need to eat, they will be forced to take a low-skilled job. Maybe they will be supervising a team of packing robots at an Amazon warehouse or cleaning rusty gears. Whatever they end up doing, the economic numbers will show them as being employed. But that is not the full story.

Economic problems are often discussed in sterile terms. We talk about unemployment numbers and labour productivity. But it is so important to remember that behind each number is a person. That those numbers have souls, and that they love and are loved. How we as people experience the phenomenon of AI automation will affect us. It will shape our views and our values. The effects of automation can traumatize, demoralize, and even radicalize.

And it is not just those who are economically displaced that are affected. Wealth inequality and economic despair affects us all. We do not live in an isolated society. What happens to your neighbours affects the world you live in too.

Imagine being forty-five years old with twenty-two years of work experience. Then, suddenly, you need to start from the bottom.

Moreover, financial struggles have massive effects on families. I was eleven years old when the great recession affected my family. For their privacy, I will shy away from the specific details. Yet I will be clear: I was deeply changed by what I experienced during that time. Even as my parents did their best to shield me, their youngest child, from what was happening, the effects were all too real.

To be honest, that is one of the reasons I started researching AI. Although my work eventually strayed from examining the effects of AI on jobs, it is a topic that has always made me nervous. A storm is coming, and we need to prepare. We need to ask ourselves hard questions about what we want this transition to look like and make difficult choices about how we achieve it. And we need to do it fast.

• Big Tech Is Hiding the Environmental Cost of Chatbots
• In the Age of AI, Are Universities Doomed?
• The Fastest Way to Lose a Court Case? Use ChatGPT

If there is one takeaway here, I hope it is this: even in a world dominated by machines, humans, and our decisions, are at the centre. Just as we build the machines and make decisions about how they are built, we build the systems that they are introduced into. We are not helpless bystanders when it comes to AI automation. We get to decide what the future of work looks like. We get to decide how we handle the great transition.

A group who once understood this idea was the Luddites. They were skilled workers who became famous for destroying the new factory machinery that was automating their work. But the story of the Luddites is much more complicated than it is often made out to be.

The Luddites were not anti-technology. In fact, many Luddites embraced the new technologies and were thrilled to work alongside them. What the Luddites feared was that their bosses would not use this technology to improve the products their customers received but to cheapen them, and that, at the same time, they’d gain more control over their workers.

In the labour market, skilled workers have a lot of power. They can speak up for their rights and demand better conditions, because they are harder to replace. By contrast, unskilled workers are easily replaced or controlled. If the company’s desire is just to make money, unskilled workers are typically their preference.

The factories that employed the Luddites could have functioned, and continued to make profits, without completely replacing their workers. Although they may not have needed every skilled worker they had before the automation, there were still opportunities to use the technology and pursue projects that required the skilled workers.

At the end of the day, the battle the Luddites fought was one over economic and personal power. They were living through a transition similar to the one we are living through today, and they didn’t like the way it was being handled. They didn’t like the way technology was being used to cheapen production or the way it was devaluing those who worked with it and robbing them of the potential to take pride in their work by subjecting them to an existence as just another cog in a machine—a cog that could be thrown away by the powers that be if necessary.

And they especially didn’t like the fact that the way these machines were being implemented was vesting massive amounts of social, economic, and political power into the hands of those who owned them. The Luddites were living through one of the last great societal transformations, the Industrial Revolution.

There is no definitive start or end date for the Industrial Revolution. But it marks the period of human history when we completely transformed the ways we made things. Before the Industrial Revolution, we mostly made stuff by hand. Tools were used, but they were simple and needed to be wielded by someone who understood them. A large majority of the population worked as farmers, and they used their physical brawn and knowledge of the land to till the fields and feed the community.

But then the Industrial Revolution began. New technologies such as the steam engine and water wheels made it possible to mechanize production. But the way production was mechanized was not set in stone. It was driven by human ideologies about how work should be done and what kind of work would lead to a better world.

Ideas about cutting costs and increasing production led to increased de-skilling of work. As more machines were introduced, humans began to be treated more like machines. Jobs were divided into small repetitive tasks. Instead of constructing a whole shoe, a worker would spend their day making the same boot part hundreds of times, every day.

The Industrial Revolution can be remembered through two competing images. On one side, it was the beginning of a gilded age. The machines of industry harnessed the productive force of society, and we began producing more goods than ever. We unlocked the potential of man and machine and built unthinkable wealth. It gave us the ability to provide more riches for everyone.

But it was also a time of great suffering, when thousands were forced off their farms or pushed out of professions in which they had found deep meaning. Communities were shattered, and rural populations were forced into the slums of the cities to find work. With no skills, workers had little power and worked for pennies. Those who owned the factories hoarded vast wealth and understood the power their positions gave them. Corruption was rampant, and they worked hard to maintain their place at the top of the hierarchy.

Children worked in factories and starved in the streets. But they were employed. Most members of society had jobs, but were they really living? It wasn’t uncommon to work ten to sixteen hours a day, six days a week, just to make ends meet. But it didn’t need to be this way.

The future of work is not predetermined. It is up to us to decide how we navigate this uncharted territory.

We as a society made choices about how the technology of the Industrial Revolution was to be used. We could have chosen a path that valued the worker and aimed to reduce suffering and distribute the benefits of increased production more evenly. Instead, we often chose efficiency and profit over human dignity and community.

The story of the Industrial Revolution, like the story of the Luddites, is a reminder of the power of human agency in the face of technological change. It’s a cautionary tale that highlights the importance of making conscious choices about how we integrate new technologies into our society, ensuring that they serve to enhance human well-being rather than diminish it.

As we stand on the brink of another transformative period, driven by AI, we are faced with similar choices. But this time, we have the past to learn from. We can see how the Industrial Revolution played out and ask, “Do we want to do that again?” Will we repeat the mistakes of the past, allowing technological progress to exacerbate inequality, displace workers without offering viable alternatives, and concentrate power and wealth in the hands of a few? Or will we learn from history and steer this technological revolution in a direction that benefits all of society?

The great transition we are undertaking is not just about the role of AI in the economy. It is not just about our jobs. It’s about who we decide we want to be. It’s about how we, as a society, choose to respond to the challenges and opportunities AI presents. It’s about the kind of world we want to live in and the values we want to guide us.

The future of work is not predetermined. How we will live with AI is not etched in stone. It is up to us to decide how we navigate this uncharted territory, ensuring that the journey leads to a destination where technology serves humanity, and not the other way around.

Excerpted, with permission, from Artificially Intelligent: The Very Human Story of AI by David Eliot, published by the University of Toronto Press, 2025. All rights reserved.