The Future of Work and Artificial Intelligence

Some thinking on the impact new technologies will have on work and labour markets globally

In August 2018 I had the privilege to participate at an event organized by CUTS International in Bangkok on the promises and opportunities of the digital economy. I gave a presentation on the above topic and am sharing it here so others may also read it and offer their comments.

This is a topic that many people – particularly governments – are thinking about very seriously now. And the direction most observers are coming from is a lot of nail biting and worry about the inevitable elimination of jobs that technological trends such as big data analytics, automation, advanced robotics and other technologies are bringing about and will continue to do so.

One demographic where this is going to have a particularly big impact is on the hundreds of thousands of people in developing and developed countries alike who earn a living from driving. Autonomous vehicles promise a whole range of tangible benefits in areas such as road safety, traffic congestion, air quality (because we will need less cars), but they also will put a lot of people out of the only livelihood they have.

Of course, this won’t happen overnight. Even though we are largely there technologically, the adoption of purely autonomous vehicles is some way off, since we need to adopt supporting laws and regulatory frameworks, figure out issues such as accident liability, and build the enabling infrastructure particularly communication networks that can support all the data that will need to flow to make this technology work, namely 5G. So the first commercially viable 5G networks accessible to the public are only being tentatively rolled out starting this year, and it will probably be 5 years until we have anything resembling the comprehensive 5G coverage that would make autonomous vehicles possible.

Despite what many think, this is of course not the first time in history that large civilian populations have found themselves faced with the consequences of their own obsolescence in the workplace. Some American economists have attributed many of the long-run labor market effects of the Great Depression to the fact that the U.S. economy at the time was in the process of transitioning from a largely agrarian-based production model to a primarily industrial economy. And in our own recent history we’ve seen jobs eliminated in sectors such as heavy manufacturing for the car industry, where significantly smaller numbers of workers are now needed to man a car-making factory. Those jobs are now gone and are not likely to come back anytime soon. So that begs the question what human labour is still required in the car factory, and the answer is those workers who program and operate the robots that do all the hard, dangerous and heavy assembly work, welding and other tasks. And those workers are much more highly paid than the ones who used to work on the production lines, and also enjoy working conditions that are much safer and more amenable. So that’s really a win for mankind generally, but a loss for those who used to earn a living on the production line.

In trade policy we talk about the distributional effects of trade liberalization, meaning that generally trade liberalization has positive and net-welfare enhancing effects for economies as a whole, but that the gains from trade are very unevenly distributed. And the same thing happens as productivity enhancing technologies become adopted in the work place. The losers are always those with the lowest level of education and skills. In fact more it is more likely that, throughout the course of human history, more jobs have been lost to technological advancements than to trade.

The track record of what we call trade adjustment assistance is unfortunately not great. And generally these programs tend to fail, since retraining or re-skilling workers is notoriously difficult. But the failures of the past don’t need to be repeated if we can learn a few valuable lessons.

Firstly I think that people need to take responsibility for their own skill levels. People are in the habit of relying on institutions to teach them what they learn, whether its schools, universities, or other formal teaching institutions. But in the internet age there is so much information freely available that it’s possible to learn just about any new skill if one is prepared to put in the time.

Secondly, I think that we as a society need to abandon the idea that you only study and learn for one career. People need to confront the reality that they may need to retrain themselves for a new job or career path several times in their lives, and they need to have a mindset that embraces this change. Unfortunately humans generally tend to abhor change, and resist it as part of their default settings. So again, the onus is on us as individuals to embrace the reality that the only constant is change. And these changes are becoming much more rapid thanks to the increasingly fast paced world we live in.

Some governments are considering and already experimenting with the idea of giving everyone a minimum basic income as a form of insurance against the effects of a sudden loss or reduction in income. Despite what one thinks of the merits of this idea from a humanitarian perspective, it is probably beyond the fiscal capabilities of the countries with the biggest populations and that are confronted with the biggest skills gaps. So again, we’re back to the reality that everyone needs to assume personal responsibility for the economic value of their own skillsets. This is a tough reality to face, but can also be a very liberating one once it’s actually confronted.

So what jobs are those that are likely to survive and pay the best in future? So firstly any skillset that enables a person to interact with and exploit the computational and labour-saving power of machines is likely to be well paid. Think of the people in my example of before in the car factory, the people who program and operate the robots who now physically assemble car parts into finished vehicles. This explains why many governments are putting a heavy emphasis on STEM (Science, Technology, Engineering and Mathematics).

One area of skills development that policymakers tend to overlook, perhaps because their productivity-enhancing impacts are more difficult to quantify, is of course in the creative spaces, like design, and in the performing arts. Although AI is now capable of composing music, there is no reason to think future fans will want to show up at a concert to watch a machine play for their entertainment. Acting, singing, dancing, stand-up comedy, these are activities that cannot readily be imitated by computers in a way that would render humans completely obsolete, even if the industries that produce these goods have also suffered their fair share of digital disruption.

So in conclusion, I think I would like to leave us with the thought that although the future of work is uncertain for some, the benefits for humanity generally of increased automation, digitization and other technologies are very promising. Technology has always allowed us to do more, be more, experience more, and will continue to do so in future.

Originally published by the author on LinkedIn.