“These aren’t the jobs you’re looking for!”
Experts have already predicted that 47% of occupations in the US are at risk of being replaced by artificial intelligence in the near future. Most of these jobs will be those that can be broken down into simple, repetitive tasks – for example, workers on manufacturing lines or baggage handlers in airports. In a world where low-skilled labour is a very densely populated area of the job market, technological unemployment is likely to be at an all-time high. On the other hand, by 2019 the number of high-skilled jobs is projected to be at more than double the level in 1991. So, we can see that in years to come, labour may be concentrated into two main groups: high-paid professionals at little risk of being automated, and low-paid workers who now operate the machinery that seized their livelihoods in the first place. Income inequality is likely to rise by virtue of this economic phenomenon, known as “job polarisation”.
The critical question is, what can governments do to maintain low levels of unemployment and to achieve an even distribution of labour across the workforce? The answer here lies in education – and in fact, artificial intelligence is set to play a central role in the future of schools, colleges and universities. Software called “adaptive learning” will tailor and customise the course that a pupil takes, placing information in an order that they will understand best, maximising efficiency and overall quality of education. This approach can be implemented through vocational qualifications like BTECs or apprenticeships, which are likely to become much more popular in the future because they provide technical training and enhance practical skills that a robot cannot replicate. In doing this, the government will begin to nurture a highly-trained, highly-skilled, and highly-paid workforce, even if the job market is partially occupied by machines.
The rise of Robocop?
Yet, we must recognise that there are many situations where intelligent robots are a more suitable choice for the job than humans. Take, for example, emergency services, soldiers, security personnel and armed police. They risk their lives on a daily basis, and are sometimes unsuccessful in their operations. Automated, error-free technologies could ensure that our everyday lives are safer, by performing these tasks without endangering human life. In fact, recent weeks witnessed a horrific terrorist attack in Nice, seemingly only made worse by a slow police response. Here, hi-tech robotic law-enforcement technology would certainly have brought an improvement in efficiency. But this by no means calls for machine-gun-wielding robot armies – even an instant communication network between patrols would have quickened the reaction of security forces, and ultimately would have saved lives.
In addition to providing an increase in safety, automated machines can also work around the clock without the risk of tiredness, stress or illness. This presents numerous advantages for science, as learning software can repeatedly test different criteria and outcomes much more quickly and accurately than a human can. Therefore, in the future it is likely that global mortality rates will fall and life expectancy will rise, as a result of the role of artificial intelligence in discovering new medicines and research methods.
Despite these benefits, putting artificial intelligence in positions of responsibility may be problematic. In the event of something going wrong, do we blame the machine, the operators, the inventors, or the manufacturers? This issue of accountability will undoubtedly be an interesting topic for debate in the near future – but so too will the drastic changes in social structure that we are likely to see.
The study of economics is based on the assumption that there are infinite wants, and only limited resources – but, as a result of future innovation, we are likely to be able to fulfil all our needs at the touch of a button. In his book ‘Capitalism, Socialism and Democracy’, Joseph Schumpeter, a somewhat unconventional economist, explores this idea in great detail. He declares that the key to capitalist economic growth is competition for the market, rather than in the market – that is, innovation rather than market competition. Therefore, when the technology of the day is so advanced that all economic wants can be satiated, the capitalist purpose disappears, and there is no other choice but to march into socialism. Other economists have called this “fully automated luxury communism”, which sees a wealthy utopian society without any need for labour, as all tasks can be performed more efficiently by government-owned machines.
But there are some dangers to this “socialist blueprint”, as Schumpeter calls it. With all labour being carried out by artificial intelligence, the human race lives in a post-work society, and becomes inactive and obsolete. This potential future is imagined vividly in the 2008 film “Wall-E”, in which the remaining residents of Earth, living on a spaceship in a state of extreme lethargy, depend on robots to do everything for them. In fact, we are already seeing glimpses of this today with iPhone apps, games and smartwatches that offer instant information and gratification, in return for minimal effort.
Potential flaws of the socialist argument
As an economist, Schumpeter seems to focus too narrowly on the role of technology and innovation in an economy, when there are many others that should be taken into account. For example, he neglects the issues of international relations and macroeconomic policy, while continuing to analyse and criticise only a very small part of the capitalist system.
“Fully automated luxury communism” also has its problems. The proposition that all labour can be automated may seem reasonable, but there are some emotional tasks requiring human interaction that even the most intelligent robot simply cannot undertake. Thus, as technology reaches its peak in popularity, we should expect to see increased employment in consultancy, therapy and similar occupations, as no software can be programmed to imitate their behaviour. Because of this, Schumpeter’s prediction of a post-capitalist society is very unlikely to materialise in the real world, as there will always be a demand for labour. But, let us entertain the possibility that artificial intelligence can advance to a near-human level of empathy – what would happen if the boundaries between man and machine became much more obscure?
Rights for robots?
Today, robots displaying the characteristics of rationality are often portrayed in Hollywood dystopian horror movies. But in the currently unimaginable distant future, will artificial intelligence have developed to such an extent that these moral machines are commonplace in reality? This would present a number of huge dilemmas for the global political economy.
Before the suffragette movement in the early 20th century, women were not allowed the vote. It was only in 1928 when the Conservative government passed the ‘Representation of the People Act’ that women over the age of 21 were allowed a say in political elections. To a Victorian or Edwardian gentleman this would have been outrageous – almost as outrageous as giving the same right to a robot today. Due to the rapidity of social change, it is impossible to predict what the world will find acceptable in coming decades and centuries. And provided that some countries in 2150 still live in a democracy, machines may well be given electoral power because of their intellectual capabilities. In fact, even though the word “democracy” finds its root in the Greek “demos”, meaning “people”, artificial intelligence might begin to play a key role in global politics, whether we like it or not.
The internet is already plagued with behaviour-analysing algorithms, a form of artificial intelligence called “cookies”. Using this data, companies can create products and promotions tailored to our individual desires. Now, imagine if it was more than just our search history that was monitored – if every purchase, every move, and every conversation was recorded by intelligent software and stored in the growing number of databases across the world. With access to such information, governments would be able to create seemingly irresistible new policies to win political favour, forged from the personal preferences of entire nations. This data could also be used to shield a government’s plans from the public eye – even if they brought about austerity, through appealing to their citizens’ wants they would still be able to maintain their popularity. In some countries, this dubious policymaking might result in the level of corruption soaring, while their populations, happy as ever with the decisions made, would be none the wiser.
A socio-economic and political reboot
Considering the potential possibilities, the emergence of artificial intelligence will undoubtedly be revolutionary. It will undermine the very structure of society, and force a new one to be born. Textbooks will have to be rewritten, and policies renegotiated. The study of conventional economics will need to be completely redefined, as signs of rationality are exhibited by machines as well as human consumers. This wave of changes is likely to deliver an act of “creative destruction”, radically altering the ways in which we think about the global political economy.
However, the problems presented by artificial intelligence cannot be solved by economics alone. If these problems are going to have such a profound impact on the global economy, then we need to think about ethics, politics and any social impact before coming to any conclusions about future plans. In fact, if machines do begin to bear an uncanny resemblance to humans, this will be more of a moral issue than an economic one. Most of all, there are important questions that we must ask ourselves – how will the human race coexist successfully and peacefully with such a formidable cognitive force? And if something goes wrong, who will be held accountable for the consequences that might follow?
Contributed by Greg Tucker, Economics Editor