Unemployment with Artificial Intelligence and Automation

Automation and Artificial Intelligence (AI) is a big discussion topic at the World Economic Forum meeting at Davos. Many prominent speakers from politicians to business leaders spoke of both the pros and cons of it. Even in Singapore, our local media has recently published an article that half of the world’s jobs can be gone if businesses just adopt the existing technologies. 30% of what most of us do on a daily basis can be automated away. This means that we could be clocking 30% less hours working or we could see more unemployment.

In 1930, Economist John Maynard Keynes published an essay stating that in the future, people will work for only 15 hours a week and spend the rest of our time enjoying life. Fast forward to now, we are far from that 15-hour workweek. The shortest workweek in the world is about double that 15-hour (Europe), Asia and America is about three times that (averaging 45 hours).

An MIT professor, Professor Erik Brynjolfsson, argued for several years that technological advancement (e.g.; Artificial Intelligence and Automation) has destroyed more jobs than it has created since the start of the 21st century. He foresees a future where even more jobs would be lost to machines, not just in the manufacturing sector, but also in professions like law, finance, and medicine.

Professor Brynjolfsson’s research showed that between WWII and the 21st century, productivity gains have coincided with jobs growth. But since the beginning of 2000, the 2 lines have diverged, with productivity continue rising but jobs growth stagnating.

Weak job growth is not the only problem caused by technological advancement. Another problem caused by technological advancement is a stagnating median household income. Median household income since the 1980s has not risen as much as the GDP per capita. GDP per capita in the 2000s is almost double that of 1975. Median household income, however, is barely 30% higher than 1975’s level. The internet phenomenon begun only in the 2000s, this means that technological advancement (not just the internet) have not benefited the average household by raising their income at a rate comparable to GDP growth.

Solution to Jobs Growth?
The global population is expected to reach 9 billion in 2050 (a 20% growth from current 7.5 billion). With jobs growth stagnating or, in some cases even declining, maybe the way to re-creating jobs is to reduce the working hours set by the law closer towards the “goal” set by Keynes. Productivity has increased multifold over the years, one man is now able to do the work that required several men to perform in the past – but has his/her salary went up by that many times?

It is simple mathematics:

Quick Maths-02
With the introduction of machines in the workforce, one man can produce what would have previously took four to complete. By reducing work hours, half the jobs can be saved while at the same time increasing individual leisure. If we were never created to sloth our whole life, maybe the working hours should shift a little?

If we have indeed become more productive, maybe we should share the gains of that productivity by working fewer hours. In fact, there has been research and studies that show humans work best when we clock approximately 30 hours working. But how plausible is such a solution?

Reducing work hours would most certainly be a drastic measure that is too shocking and damaging to our economy if it were to be implemented immediately. Reducing working hours and having more employers translates into more administration work for HR, which can potentially create more issues. Assuming that workers are paid hourly, the incentive to work decreases drastically when they are expected to travel to and fro work for a mere few hours daily. Would companies then be willing to increase wages to incentivize workers? The solution to maintaining employment in the generation of AI is more than just collectively reducing work hours to keep everyone in a job.

Opposition to the Idea that Technology Kills Jobs:
Of course, there are economists who claim that the evidence provided does not fully support the idea of technological advancement. The period of study, particularly from the beginning of this century, happens to coincide with the Dot-Com bubble and the Great Recession, which economists claim could be distorted as data from the 2000s are not compared against post-Great-Depression period.