A major technology shift is happening in the workplace. What is the impact on the work environment when artificial intelligence (AI) and robotics are introduced on a large scale? And what happens in the labour market when certain skills are no longer required? We asked two researchers in IT and economics.
AI and machine learning are general technologies with a broad range of applications – from customer service in a company, to the assessment of X-ray images or legal advice. Despite rapid development, there is little research into how the introduction of AI and machine learning affects the work environment.
"This is worrying, although it is often the case with technological developments that the technology comes first and researching the consequences comes later. It is only after something has been introduced that we are able to know the effects it may have," says Åsa Cajander.
Cajander is a professor at the Department of Information Technology at Uppsala University specialising in human-computer interaction.
Use AI as a tool
Information Technology. Photo: Mikael Wallerstedt
Current research indicates that those with higher education are often given the option to use AI as a tool in their work. When it comes to simpler work, however, you hear about people being replaced by AI instead.
Research also shows that we find it useful to hand over certain tasks to AI, such as safety-critical systems.
"Sending a robot into a fire instead of a human being is positive for all involved. In these instances, we can be grateful for robots, AI and automation. Carrying out simple administrative tasks does not make many people happy either. However, from a user perspective, work tasks that are part of a core business that you are passionate about are not as suitable for automation," says Cajander.
Focuses on technical development
AI research often focuses on technical development, e.g., how to design a successful machine learning project.
"There is very little research into how to introduce AI into an organisation. What sort of leadership is required? How should it be implemented?" says Cajander.
She has been researching digitalisation and the work environment for twenty years and believes a lot of this knowledge can also be applied to the implementation of AI systems.
For example, the systems need to be transparent if they are to be used as decision-making tools. Cajander has studied a computer tool that collects information about people who contact the healthcare system and then recommends nurses to take a particular course of action.
"What was missing was an explanation for the recommendations. This is known as 'explainable AI' and is of vital importance. People have to understand why the system thinks as it does."
Your professional role changes
In the long term, your entire professional role changes when you work closely with an AI system. We tend to be observers of a process, rather than being actively involved. However, we need to monitor the process and intervene when things go wrong.
"We are transitioning from something very active and quite creative to something quite passive. Will it be a work environment that we enjoy, where we feel we can use our expertise?"
There are also major problems with a lack of work engagement in Europe, says Cajander.
"Many people feel incredibly disconnected and unmotivated by their work. How do we counteract this? How do we create meaningful work in the future with AI and automation? What if AI could eliminate the boring, repetitive and simple tasks that nobody is motivated by?"
Positive effects on the labour market
Economics. Photo: Mikael Wallerstedt
From an economic perspective, new technologies often have positive effects on the labour market, according to research. For example, the average wage increases, says Georg Graetz, researcher at the Department of Economics at Uppsala University.
"In Sweden, the average wage has increased by approximately two per cent per year, disregarding inflation, mainly thanks to new technology. However, there are also people who are adversely affected by new technologies and who may lose their jobs."
New technologies also risk creating inequality, according to research in the US and the UK. Wage disparities between low-skilled and high-skilled workers have increased since 1980 and new technology is the most likely explanation. However, this development has not taken place in Sweden.
"It is difficult to know for certain what this may be due to, but the most likely explanation is our collective agreements, which have far less importance in the US and the UK. In Sweden we have strong standards, and the wage spread is compressed."
The number of jobs has risen
In the debate on automation there is a perception that new technologies are taking jobs away, but in fact the number of jobs has risen continuously throughout history, throughout the world. Graetz shows graphs of the number of employees in Sweden divided into different age groups and gender. In most groups, the number of employees is stable or increasing. The exception is found in the younger population, where fewer are employed and more are studying.
This is not the case in all countries. In the US, there is a steadily declining employment rate.
"There are, of course, places around the world where there is concern that structures are changing in a way that is not positive for everyone. However, it is a mistake to think that there are a set number of jobs, and that automation leads to fewer jobs. This is simply not true."
A demand for different forms of expertise
When someone works, produces something that creates value and generates income, they consume more, which creates more jobs. Therefore, all major changes in the labour market, such as the transition from an agricultural society to an industrial society, have led to the creation of new jobs.
However, changes in technology contribute to a demand for different forms of expertise.
"Sometimes, it may be difficult to adapt in the short term, but in the long term this has never been a problem. Although the big concern is, of course, that there may be a situation in which technology changes so quickly that we cannot keep up with it," says Graetz.
Annica Hulth
