Of all recent technological advances, Artificial Intelligence is surely the area with the greatest hype and speculation. If you believe AI enthusiasts like futurist Ray Kurzweil, smart machines will have surpassed human brains by the year 2025 – and be smarter than all human brains combined by 2050. From this point on, which Kurzweil calls the Singularity, it would seem to be only a matter of a few years until machines replace humans completely, taking over even the most complex tasks and jobs.
No wonder the advance of robots and AI is a huge cause of concern for businesses, individuals and governments! According to a 2015 survey by Chapman University, Americans fear robots replacing them more than they fear death.
But how likely is this dystopian vision of the AI revolution really? Will AI really spell the end of all work, maybe even the end of humanity? In The AI Economy: Work, Wealth and Welfare in the Robot Age, author Roger Bootle tackles some of our biggest questions, hopes and concerns surrounding the age of AI from a pragmatic perspective.
How can we predict the future of the global economy? Well, in order to understand the present and speculate on what’s next, we might look to the past for answers.
For economists, the most significant historical process to study is the Industrial Revolution. But according to American economist Robert Gordon, there has not just been one but three separate industrial revolutions. The first was the process of technological innovation, social and political change, which began in Great Britain in the late eighteenth century with the invention of steam engines and railroads.
The second one began at the end of the nineteenth century, with the invention of electricity, combustion engines and telephones. The third one happened in the 1960s, after the invention of the computer. And now, with the technological advancements of robots and AI, we are about to enter a fourth industrial revolution.
However, AI and robotics have been so overhyped that it’s tempting to think that they will usher in a much more radical transformation of the world, rather than “just” another industrial revolution. Now, it’s true that robots and AI have made significant strides in recent years – in processing power, algorithmic decision-making and text and image recognition. In 2016, for example, Google’s DeepMind AI beat the reigning human champion of Go, a complicated Chinese board game.
But even here, progress has been limited. For example, Google recently tried to train AI to recognize images of cats on YouTube with much less success: it took the power of 16,000 computers to identify a single one. And then there are still plenty of skills, including creative thinking, emotional intelligence and manual dexterity, that robots and AI have difficulty mastering, and there is no obvious solution in sight. So despite what AI enthusiasts promise, there is little reason to believe that machines will replace humans in all but a very specific set of jobs in the near future.
So what can we expect from the fourth industrial revolution of robots and AI? Well, even though the previous revolutions did not lead to an immediate improvement in living conditions for workers, over time their wages and quality of life increased. In the year 2000, the global per capita GDP was over thirty times more than it was in 1800. Similarly, the AI revolution might be more of a process than a single, dramatic event. But if it follows the same pattern, it will boost productivity and economic growth, making everyone better off in the long run.
The robots are coming to take our jobs! That’s one of the main fears surrounding the rise of AI – but it’s largely unfounded. The McKinsey Institute estimates that, in richer countries, only 14 percent of jobs are “highly automatable,” while only 5 percent of jobs are “entirely automatable.” This does not warrant dystopian visions of mass unemployment. But it still means that, by 2030, 375 to 700 million jobs could be made redundant by robots.
Some of the jobs likely to disappear are cashiers, grocery baggers, check-in assistants and other repetitive, low-skill work. Even routine legal work, accounting, analyzing and simple translation jobs can now be done by AI.
But such change isn’t necessarily a disaster for the economy. As some jobs are taken over by machines, new ones appear. In the 1900s, agriculture accounted for 40 percent of employment in the United States. Today, it only accounts for 2 percent, and yet overall employment has not diminished.
The World Economic Forum estimates that, by 2026, 12.4 million new jobs will be created in the USA. Some of these will directly relate to developing, building and maintaining robots. Others will be an indirect result of the AI revolution. For example, the spread of robots could free up people to take on more “human” positions, such as providing customers with personal guidance and advice.
In many cases, robots and AI are simply not what they’re cracked up to be. For example, the replacement of drivers by self-driving cars is going much slower than expected. For technical and legal reasons, none of the models available today are fully autonomous. Even when the vehicles steer themselves, drivers must stay alert and be ready to intervene. Fully self-driving vehicles function on only very restricted routes, such as between airport terminals.
Humans are simply better than machines at creative work like art, design or journalism – but they also come out on top in jobs that require flexible thinking and manual dexterity, such as plumbing, gardening and electrical work. A robot might be able to build your car, but if it breaks, you still have to go to a human technician.
So, in many areas, machines will simply work alongside their human counterparts, boosting their productivity. For example, many doctors are already using surgical robots to help them perform complicated procedures.
Another common misconception is that robots “work for free.” But robots and AI are very expensive to build, develop and maintain, and always at risk of obsolescence. The average industrial robot costs around $100,000 to buy, and up to four times as much to maintain over its lifespan. This means that, for some companies, human labor might simply be the cheaper option.
According to the Deloitte’s Shift Index, 80 percent of people hate their jobs. Yet the average adult today spends most of their life in full-time employment, working around 30 to 40 hours a week. And many office workers, lawyers and bankers work even more than that. If we hate our jobs, why do we still work so much?
The answer is not so straightforward. Beyond just paying the bills, a job is for many people a source of value and purpose. And, while work can be a source of stress, unemployment can be an even greater source of angst. Moreover, society values economic success, and rewards people’s competitive streak. And as inequality widens, people on the lower side of the income spectrum need to work more to make ends meet.
But the AI revolution might finally free us to work less, and there is a strong case for doing so. Studies show that nations who work a lot, such as South Korea, consistently report lower levels of happiness than nations who work less, such as Denmark. Nations in the latter group also have a higher number of volunteers.
American founding father Benjamin Franklin was one of the first people to suggest that, in the near future, people should not need to work more than four hours a week. Since then, the end of work has been proclaimed multiple times but nothing of the sort has come to pass.
But as robots and AI gradually infiltrate the working world, they might free humans to engage in the more meaningful parts of their jobs and work less in general. The boost in productivity, GDP and material wealth that the machines could bring means that more and more people will be able to opt for more leisure time over more work. This could be reflected in shorter workdays, a shorter workweek, longer vacations or earlier retirement.
In some places, the shift is already happening. The IG Metall, the national German union of metal workers, recently succeeded in reducing the work week to 28 hours for around 900,000 employees.
It’s also a question of choice and preference as to whether people will work more or less in the new economy. These, in turn, hinge on larger social and cultural factors. A culture that places more value on hobbies, community work and personal development and less value on economic success will encourage people to make use of their new freedom.
Check out my related post: What is the future of Artificial Intelligence?