Work is central to humanity. It governs our purpose and use of time; without it we would lose all sense of hierarchy and social solidarity. Not only is it vital to individuals and their livelihoods, it is also a central part of culture and societies internationally, with work being a staple of social order and status. Nations and communities who productively employ people are usually better off for it. It allows people to be self-sufficient, give back to their communities, and build a greater sense of a common good, allowing future generations to thrive. However, this is now under threat. As artificial intelligence and ‘smart’ automation become more common, they are starting to be real threats to employment and economies. A proper evaluation of the peril of AI must be undertaken to ensure a more inclusive future for all.
To examine the threat from automation, one must consider the impact automation had in the past. When examining the historic impact, an obvious place to start would be the Industrial Revolution. Although many acknowledge the Industrial Revolution as being an age of improvement, at the time it was not seen this way. Wage increases for workers (a side-effect of increasing productivity) were not apparent until the middle of the 19th Century. Robert Gordon suggested that technological innovations contributed modestly to growth for many decades after their inventions due to three factors: public distrust on account of glitches in the technology, a lag in workers adapting to new technology, and unemployment. However, as observed by Brookings, technology lags have decreased by 90% since 1800, but still remain a hinderance to the productivity of GPTs (General Purpose Technologies). During this lag workers who are still employed are left exposed to this raw technology and, as was the case in the Industrial Revolution, workers experience very modest wage gains in return for longer shifts in worse conditions. In times of uncertainty low-skilled and low-paid workers have little say in the technology that inevitably impacts their jobs, which in the Industrial Revolution led to exploitation of workers in factories. A representation of this was created by Michael Kremer with his O-ring production chain model. As you improve the efficiency of one link in the chain, the entire production chain improves. Factoring automation into this model, one can identify that theoretically as automation makes more routine tasks faster and cheaper, the value of human production should increase, which in turn should result in increased wages and standards of living for the people who are employed within that production chain.
A contemporary example is the introduction of ATMs in the 1970s. As Automation began to disrupt the industry one would assume the number of bank tellers would have decreased as ATMs grew more popular, this was actually not the case. They increased in the US from 500,000 in 1980 to 550,000 in 2010. This was evidence that perhaps automation was not an imminent threat to employment and complemented it. This is because as ATMs reduced the cost of each bank branch, more tellers could be employed to work in a more emotional, rather than mathematical, style of banking. This gave rise to ‘relationship banking’. As bank tellers became more like salespeople and less like machines, they became better paid and were happier at work. This is because their tasks, which were replaced by automation, created a new set of more fulfilling jobs, which were subsequently complemented by the machines that were once a threat to their positions and livelihoods.
However, although automation can complement labour it can also destroy it entirely, creating a new division of labour in which workers are uneducated. Although this may not lead to a net loss in jobs, it is still a significant social issue, as many will be left without the means to retrain themselves for more productive positions complemented by new technologies. As Yuval Noah Harari states, people will not become unemployed but unemployable. As observed in the graph below, non-routine tasks will make up a much larger share of the labour market in the future. This is because middle-skilled labour is routine and programmable and human workers will be more inefficient and expensive than robots, which will result in a massive transition. Middle-skilled workers will become unemployed and will have to take up low-skilled work or will have to retrain to receive more cognitive tasks. The latter will require a stronger government response and is less likely to happen than the former. This is because many people will need to have a steady income and cannot take a break from work to retrain, despite the long-term benefits. A possible solution to this would be a universal basic income as it would give people the economic flexibility that they need when it comes to integrating themselves in the new age of technology.
A study conducted by the Aspen Institute found that 44% of workers without a university degree will likely have jobs that can be automated in the next decade, compared to 1% of those who have obtained a university degree. This is due to less-educated peoples lack of intellectual mobility, which constricts them to one specific area of labour and will result in them having to retrain should they want to have a more productive and well-paid job in the future. As stated above, these middle to low paid skilled workers, have little economic mobility, which requires intervention from the government. Andrew Yang, a strong supporter of Universal Basic Income argues that people who have no option but to take up minimum-wage jobs when they become unemployed should receive ‘free money’ so that they are able to reskill and have more options as the age of artificial intelligence arrives. UBI would also help to decrease the exceeding inequality we will likely observe, as the middle class continues to shrink.
Since we live in unprecedented times, and predictions can only be so accurate, we must be optimistic, but not ignorant. Governments should ensure the protection of workers and a possible avenue for doing so would be a universal basic income. AI and smart automation will massively disrupt labour markets and could cause mass unemployment in the years to come, but in the long term, positive ramifications of these emerging technologies will be observed. A perfect example of this is the WEF’s prediction that AI will displace 75 million workers by 2030, but will create 133 million new jobs. Looking to the future, we must find ways to coexist with AI as it is here to stay.