Are AI and Automation Good for Jobs?

One of my colleagues is about to start some work with the local Department of Trade and Industry (DTI) on the potential impacts of the 4th Industrial Revolution (4IR) concerning the future of jobs. They are, as most governments around the world, concerned about how technologies like Artificial Intelligence (#AI) and wide scale automation using technologies such as Robotic Process Automation (#RPA) will impact the labour market.

Fears and Predictions of Job Losses

In recent years there have been an endless stream of studies, reports, predictions and articles written on just how fundamentally AI and automation will change the jobs landscape, with some predicting that there will be a catastrophic loss of jobs in the near future.

Pioneering work from Carl Frey and Michael Osborne, based at the University of Oxford, first projected job losses due to automation in late 2013. The Oxford Study, as it is sometimes called, predicted that 47 percent of U.S. jobs could be lost due to automation. Kai-Fu Lee, who has worked for Apple, Google and Microsoft and who is the founder of venture capital firm Sinovation Ventures, has claimed multiple times that robots are likely to take some 50 percent of jobs in the next decade. Others like Vivek Wadhwa, a tech entrepreneur and faculty member at Carnegie Mellon University in Silicon Valley, are even more pessimistic and predict that current studies are underestimating the impact of technology, with up to 80 to 90 percent of jobs being eliminated in the next 10 to 15 years.

Quick Reality Check

But what does history and the reality on the ground tell us about the impact of technology on the labour market. Is there a direct correlation between the introduction of 'human replacement' technology and job losses?

A great early example is the introduction of the Automated Teller Machine (ATM) from way back in the late 1960s. It was widely predicted that the introduction of the ATM would put bank tellers out of work, but the reality was the opposite. James Bessen has discussed this phenomenon, detailing the actual outcome;

The average bank branch in an urban area required about 21 tellers. That was cut because of the ATM machine to about 13 tellers. But that meant it was cheaper to operate a branch, and when it became cheaper to do so, demand for branch offices increased. And ass a direct result of the expansion, demand for bank tellers actually increased.

Research from the Fox School of Business, Temple University, presented at the CIST 2015 conference shows an interesting example for nursing homes, where the implementation of automation technologies decreased the staffing levels by 5.8 percent in high-end nursing homes, while low-end homes saw an increase in staffing by 7.6 percent. This study showed that automation does not affect jobs uniformly throughout an industry or market, with multiple factors determining the impact of automation.

A study of Spanish manufacturing firms found that more productive firms are more likely to adopt robots, but this does not necessarily lead to job losses. The study found that RPA adoption generated substantial output gains in the vicinity of 20-25 percent within four years, reduced the labour cost share by 5-7 percent, but also lead to net job creation at a rate of 10 percent. Amazingly, the study also showed that manufactures that did not adopt RPA were in fact more likely to shed jobs!

Telefonica S.A., a leading telecommunications provider, is one of the global leaders in RPA and has for the last 5 years been hard at work automating large parts of their back-office operations. According to their annual financial reports, they have seen a steady decline in their headcount of almost 4 percent annually, and despite having declining revenues have been able to increase their overall profits on average almost 30 percent per annum over the past 3 years.

Finally, let's take a look at the hard employment numbers, specifically in the US. According to the Bureau of Labor Statistics in the decade between 2006 and 2016 over 51 million jobs were destroyed, while 179 million jobs were created.

Similarly, the official unemployment rate in the US has been steadily declining and is at a historic low of just 3.5 percent.

State of AI and Automation

Despite the hype and fears, many companies have been investing heavily into AI and it is seen as the major driver, together with RPA, for delivering on the digital transformation aspirations of many companies. However, even if a company pumps money into AI, it doesn’t mean the end result will be more productive.

A survey by Protiviti, a consulting firm on AI projects, found that only 16 percent of business leaders say they are getting significant value from AI. However, they also note that this figure could jump to almost 50% in just 2 years. Furthermore, while businesses are aware of the importance of AI, the process of adopting such technology has been slow, particularly among small and medium enterprises. A report from Bluewolf found that only one in five (21 percent) small businesses have implemented AI-based solutions. They note that in the process of adopting AI, enterprise organizations face a dual challenge of balancing the improvements in productivity and satisfaction while they shift the digital mindset of the organization.

Is it a situation of "much ado about nothing"?

So, it seems at least currently, that the fears of major job losses are not yet completely founded and the predictions may be a little one sided. One reason for this is that each of the studies assumes a unidirectional impact, which simply doesn’t reflect the reality. This effect is described by economists who talk about technologies having a complementary or substitutionary effect on jobs. That is, they can either complement people by making their work more productive, which typically leads to higher wages, or technology can replace people. But automation technologies will impact specific tasks within specific jobs within specific firms within specific industries, all at different rates and directions. Which means there is no simple or direct correlation between companies implementing AI and automation and a negative impact on the labour market.

Another factor is that some companies like Microsoft are actively seeking for synergies between AI and humans, where AI can be used to amplify human ingenuity not destroy it. Besides global platform and software companies like Microsoft, there are many think-tanks, NGOs and other organizations that are actively working to ensure that AI has a positive impact on humanity.

What does this mean for governments and entities like the DTI and for professionals such as us? Well, the reality is that there will definitely be some impact on jobs due to the ongoing adoption of AI and automation, and if you are caught on the wrong side of it, then it may have dire consequences. Governments should ensure that they are building a resilient labour environment, while at the same time ensuring that the correct digital skills are being infused into curriculum. No-one has a silver ball to predict the impact to our careers. However, it's probably prudent to rather be on the safe side and ensure that you are applying a growth mindset to your career and ensuring that you are building relevant digital skills for the future.

Main source: https://www.americanactionforum.org/insight/understanding-job-loss-predictions-from-artificial-intelligence/