AI, or Artificial Intelligence, is arguably the next frontier in human technological advancement.

Its full potential has certainly not yet been realised, with the speed and scale of advancements being reminiscent of sci-fi fiction. Elon Musk’s company, Neuralink, for example, has recently implanted a microchip into a human brain. AI will also see many breakthroughs in medical science and increases in diagnostic and treatment power over the coming decades.

It feels as if there has been an explosion of AI-related talk, and endeavours, over the last decade, with more and more people realising the ultimate potential of AI. AI has become more mainstream, and people are starting to become nervous about it. In fact, AI, which is really more akin to ‘intelligent automation’, has actually been around in various guises for some time. 

Licence-plate recognition cameras, for example, use AI to match what the camera is ‘seeing’ to a pre-loaded database, making split-second comparisons in real time, and even sending real-time alerts when a match is detected.

Apple has been playing around with AI for some time now and so have all the other major telecommunication brands. Chatbots, banking apps, facial recognition, and algorithms that make suggestions based on social media habits and music tastes  are all examples of AI automation in our everyday lives.

It is self-evident that AI automation has slowly crept into the facets of our lives that we take for granted, and it has been useful, for the most part. We feel this way because AI has not yet been a threatening force. AI automation has been a rather benign and generally helpful force and, in terms of everyday existence, not one that has been absolutely necessary. After all, it is fairly easy to search for music rather than have it automatically suggested. The problem now is that AI automation has moved from being a benign helpful force to one that is seen as more of a threat to those elements in life that really matter.

Increases in efficiency

The conversation has shifted from the grandiose foretelling of the limitless power of human existence to the very real prospect of AI automation rendering large sections of the workforce obsolete, as its power is understood. Businesses stand to gain the most out of AI automation, with significant increases in efficiency, the minimizing of mistakes and human error, and detailed data analytics presented in real time, to name a few. Businesses will also save on staffing costs, as certain elements of their workforce are no longer needed. For better or for worse, businesses have every incentive to lean heavily on AI automation.

It is important that governments take heed of the impending implications of AI automation advancement. If some jobs are rendered obsolete, the burden on state resources will likely increase, as unemployment levels rise. In a country like South Africa, state resources are already stretched. Innovative solutions will be necessary to help populations keep up with the pace of advancement and ensure an economic environment that multiplies accessible employment opportunities.

Western ‘first world’ countries are better placed to absorb the negative socioeconomic consequences of AI automation. The United States, for example, has the financial ability to mitigate many of the risks posed by AI automation. Industries in these countries are more robust and socio-economic data more favourable, allowing for a greater transferability and adaptability of skills to meet the AI challenge.

The South African economy and those in other lower to middle-income developing countries are certainly not as robust as those in the ‘first world’. For South Africa, the risks of advancing AI automation are extremely real for a population dealing with  current levels of unemployment and job shortages. Quoting percentages and using the language of economics can sometimes overshadow the very real human scale of the challenge. In real terms, out of the 62 million South Africans, upwards of 21 million people are not participating in economic activity, and only 16 million people have some form of formal employment. Couple that with other socio-economic constraints the country is facing, and it is little wonder that high levels of poverty and the associated risks persist.

South Africa’s unemployment problem will be amplified by its effect on industries with increasing levels of AI automation. The retail sector is going to be one of the biggest losers (in job-loss terms) to AI automation. The retail sector is also one that employs a significant amount of unskilled or semi-skilled labour. People in this group are often paid the minimum wage, and their wages are significant contributors to household income. This segment simply cannot afford to be sidelined by automation.

Other sectors

The manufacturing sector is another that may see AI automation replacing a large number of jobs. Extrapolate outwards, and the problem becomes even more clear. Other sectors that are due to lose out to AI automation include copywriters, graphic designers, customer service, software engineers, and the list goes on. Added to this, the African Development Bank has predicted that 100 million youth on the African continent will be without jobs due to the advancement of AI automation in the coming decade.

The counter-argument to this is that AI automation is going to generate far more jobs than it replaces and thus contribute significantly to GDP growth, creating more jobs. McKinsey & Company authored a report on AI in South Africa, entitled ‘The future of work in South Africa’. This report finds that by 2030, the retail sector, administrative support, and government sectors, manufacturing, transportation and warehousing,  agriculture, forestry and fishing, real estate, and information sectors stand to lose a combined 531,000 jobs. Sectors in line for significant growth include health care, construction and other emerging services, scientific and technical services, educational services, arts and entertainment, accommodation, food services, and wholesale trade, to the tune of 1,304,000 jobs.

While on paper these numbers are reassuring, the crux of the matter is that these numbers belie the fact that not all jobs are equal, and certainly not all circumstances are the same.  The difference is that people in skilled-labour jobs have a higher likelihood of being able to upskill and/or reskill themselves to meet the new job landscape that AI automation will create. Populations performing various forms of unskilled or semiskilled labour may not have this opportunity, and may have to rely on the ability (or will) of their employers to create such conditions. Ultimately, the South African labour force as a whole needs to upskill itself to meet the challenges resulting from AI automation.

May fall through the cracks

Should heed not be taken of this warning, it is likely that unskilled and semiskilled labour populations could become the ‘missing middle’ that may fall through the cracks of the automation revolution. This is a population that can least afford to do so. The economic impact of AI automation cannot and should not be studied in isolation from its human cost. The human cost affects the prosperity of ordinary South Africans, displaced by the wonders of human technical advancement.

So what can be done to save this ‘missing middle’? According to McKinsey & Company, a collaborative approach between Government, business, and educational institutions will be imperative to prepare for the future of tomorrow and also deal with the problems of the present. Basic education reform that reorientates itself to prepare children for the technological revolution that is on our doorstep will go a long way to shaping South Africa’s future. Children need developed skillsets in sectors that will see the most growth. Education is also the key that unlocks the cycle of perpetual unemployment that plagues South Africa today. Focusing on basic education is an investment in the future but does not solve the problems of the here-and-now.

A public/private partnership between Government and the private sector will also be necessary to spearhead this ‘population skills revolution’. While educating the young serves in planning for the future, a public/private partnership has to take a ‘no one left behind’ approach that accounts for all segments of the work force.

Upskilling

To ensure that the ‘missing middle’ is not left behind, government could look at incentivisation schemes that encourage the private sector to hold on to their employees and provide their own upskilling programmes. Government could also partner with educational institutions to develop easily accessible programs that upskill or reskill those in the ‘missing middle’, and place them on a path towards job acquisition, retention, and relevant further education and training.

The writing is on the wall for South Africa. AI automation isn’t all bad, but it is important to take into consideration not only the jobs that will be created but also the jobs that will be lost. There is a human cost in AI advancement and planning should be done accordingly.

South Africa cannot afford more people sinking below the breadline. Needless to say, the answers are complex and the exact solutions even more so. A multi-pronged approach will be needed to solve these problems, but South Africa must address them if it is to compete in the world of tomorrow.

 [Image: Gerd Altmann from Pixabay]

The views of the writer are not necessarily the views of the Daily Friend or the IRR

If you like what you have just read, support the Daily Friend


contributor

Born and raised in Cape Town, Adrian Badminton has degrees in both psychology and sociology. He has an MA in sociology from UCT.