Are employees right to fear artificial intelligence: People are not easily replaced

Artificial intelligence is now performing a growing number of tasks that were previously reserved exclusively for humans. From writing routine reports and translating texts to analysing data and handling customer support, AI is rapidly taking over duties that until recently required human labour. This shift is no longer merely a technological novelty. It has become a serious factor directly reshaping the global labour market.

In developed countries, this wave is already leaving tangible consequences for the structure of the workforce. Some companies are reducing staff numbers in order to replace employees with artificial intelligence, believing this to be far cheaper and more profitable for their operations.

Vuk Vuković, president of the Association of Business Informatics Professionals and an associate professor at the Faculty of Economics in Subotica, recently said that artificial intelligence in Serbia is being used at a level only slightly lower than, if not fully equal to, that in the world’s most developed countries. Given that domestic companies are not lagging behind global trends in implementation, the logical question arises as to whether employees in Serbia should fear for their jobs.

However, Vuković told NIN that fear is by no means the best way to view what is currently happening.

“Every major technological change should also be seen as an opportunity, but there is certainly reason for employees to think seriously about how their work will change in the period ahead. Above all, it is important to make a clear distinction that is often overlooked in public debate, especially outside expert circles. When we speak today about artificial intelligence, we are dealing both with specialised AI, meaning systems that can perform certain tasks very well, and with general-purpose artificial intelligence models. This second category includes large language models such as GPT, Gemini, Claude and others. Their capabilities today are broad. They can write, analyse documents, code, translate, process data, generate various types of content, and so on. Their progress from month to month is clearly impressive.

However, general-purpose artificial intelligence models are not the same as artificial general intelligence in the original meaning of that term. That would imply a system capable of learning, reasoning, adapting to different problems and successfully performing a wide range of cognitive activities at a level comparable to a human being. We have not yet reached that point, nor can we reliably say today when we will,” our interlocutor explains.

Changes in work and productivity

For that reason, Vuković believes that the most important question at this moment is not whether AI will replace humans, but which parts of our work it can already take over.

“Some of them can already be fully or partially automated by artificial intelligence, while in other cases it can significantly speed up completion. Still, certain activities continue to require experience, judgement, responsibility, communication with people and an understanding of the specific business context. At the same time, we should not ignore the fact that AI can make mistakes. That is why I expect that in a large number of professions we will see a change in job content rather than the complete disappearance of occupations. As with previous major technological changes, some professions will gradually disappear, but new ones will emerge at the same time,” says Vuković.

He adds that in a great many occupations, AI will not simply replace the employee. A far more realistic scenario is that an employee who uses the new tool well will become significantly more productive than one who does not. In other words, if someone previously needed four hours to analyse certain documentation, prepare a draft report or process a large volume of data, but can complete the same work in one or two hours with the help of artificial intelligence, it is clear that companies’ expectations regarding productivity will also change over time.

“Of course, this is only a hypothetical example, because the actual effects will depend on the type of work, the organisation, the quality of implementation and a range of other factors. That does not necessarily mean that a company will introduce artificial intelligence and immediately dismiss some of its employees. It is possible, for example, that a smaller number of staff, with the help of artificial intelligence, will do work that previously required a larger team. The extent to which this will really happen will depend from one company to another and from one occupation to another,” Vuković says.

For that reason, our interlocutor believes that in a large number of professions it will no longer be enough simply to know one’s job well. It will also be necessary to know how to do that job with the help of AI.

He stresses that the impact of artificial intelligence is most visible among programmers. Modern systems can already generate code, identify bugs, write tests, explain existing code or offer solutions to specific technical problems.

“However, I would not conclude from this that the profession of programmer will disappear. Rather, the actual content of their work will change. Programmers will spend less time on routine code writing and more time defining problems, checking the solutions proposed by AI, understanding business requirements and making more important technical decisions. That is precisely why the profile of the business informatics specialist is gaining in importance, because this is not merely a programmer but an expert who understands both technology and business, and that combination of knowledge will be even more valuable in the era of artificial intelligence. When it comes to programmers, I would draw particular attention to the position of junior developers. A large number of the tasks through which young programmers previously learned the profession can now be automated relatively easily. This presents a challenge for both companies and the education system. If artificial intelligence takes over some beginner-level tasks, then we must find a new way for juniors to gain the experience necessary for them to one day become seniors,” says Vuković.

Programmers, however, are not the only profession that will face changes because of the emergence of artificial intelligence. Vuković notes that a similar impact can be expected in administrative jobs involving data entry and processing, first-level customer support, the preparation of routine financial reports and the processing of standardised documentation, the translation of simpler and structured texts, the drafting of standard legal documents, the creation of marketing content, the analysis of large amounts of documentation, and so on.

“What all these jobs have in common is not their simplicity, but the fact that a large part of the activity is digital, repetitive and can be described with sufficient precision,” Vuković says.

Adapting education and symbiosis

Our interlocutor says that at this moment he does not expect a wave of mass layoffs in Serbia that could be attributed exclusively to the use of AI. He says a gradual process is far more likely. In some companies, staff numbers may indeed be reduced, but the effect will also be visible to a large extent through a reduced need for new hiring.

“If the use of AI enables an existing team to handle a larger volume of work, a company may, for the same growth in business, need fewer new employees than before. Of course, there is no universal coefficient for productivity gains here. The effects will vary depending on the type of work, the company, the way work is organised and the quality of the implementation of artificial intelligence. At first glance, such a change may not appear as dramatic as a major wave of layoffs, but in the long term it can have a very strong impact on the labour market,” Vuković stresses.

Our interlocutor adds that experience in developed countries so far shows that the impact of artificial intelligence on the labour market cannot be reduced solely to the question of dismissals. A large number of jobs are indeed exposed to the influence of generative AI, but that does not automatically mean their disappearance. In many cases, what will change first is the way the work is done.

At the same time, developed economies are in a certain sense even more exposed to these changes than Serbia, precisely because of the higher level of business digitalisation and the greater share of office, administrative and professional jobs. The larger the share of work that is digitalised and based on information processing, the greater the scope for AI to automate certain activities or reduce the need for human labour. That is why, says Vuković, developed economies, although better placed to take advantage of artificial intelligence and increase productivity, may at the same time be more exposed to a reduced need for certain types of jobs and new hiring.

Symbiosis between human labour and AI

“For Serbia, one of the key questions will therefore be how quickly companies, employees and the education system will be able to adapt to these changes. Technology is currently developing faster than the requirements of jobs, study programmes and professional training systems are adapting,” Vuković says.

However, this pace of development raises the question of whether, in some occupations, instead of a reduction in the workforce, there could be a symbiosis between human labour and artificial intelligence.

“In a large number of professions, I expect that this is exactly what will happen. A doctor can use artificial intelligence to help analyse scans and medical records, but decisions on treatment and responsibility remain with the doctor. A professor can use it to prepare teaching materials or analyse students’ results, but education is not merely the production of content. It involves motivating students, understanding the problems students face and, ultimately, passing on experiential, or tacit, knowledge possessed by the professor. The same applies to engineers, managers, lawyers, financial experts, programmers and many other professions,” says Vuković.

PROFIMEDIA / COLUMBO / IMAGEBROKER / PROFIMEDIA

Asked whether the application of AI in some fields could create conditions for additional hiring, our interlocutor responds positively.

“There is already growing demand for experts who develop AI systems, work on data processing and analysis, integrate artificial intelligence into business processes, verify the reliability of such systems or ensure their security. Artificial intelligence may reduce the need for employees in one part of an organisation’s business processes while simultaneously increasing the need in another. For that reason, I believe the future of the labour market should not be viewed solely through the question of how many jobs might disappear. An equally important question is what new jobs will emerge and which competencies people will need in order to perform them,” Vuković says.

An unexpected trend in large companies

Meanwhile, a considerable number of employees who fear that artificial intelligence could take over their jobs may be at least partly reassured by a new trend in the operations of major global companies. Corporations that laid off workers following the introduction of artificial intelligence are now hastily bringing people back. Examples from giants such as Ford, Commonwealth Bank of Australia and IBM clearly show that completely pushing humans out led to a decline in the quality of operations and frequent mistakes. Artificial intelligence successfully handled mechanical tasks, but completely failed where human judgement, accountability and an understanding of the specific business context were essential. In other words, this course of events in practice shows that even the most advanced technology will not be able simply to replace people in carrying out their everyday jobs.