Serbia and AI – First on paper, penultimate in reality

Since the concept of artificial intelligence gained global popularity as a technological achievement that would completely change the world, there has been constant boasting from the top ranks of the Serbian government that Serbia is one of the leaders spearheading this innovation revolution. Millions of euros are allocated for this purpose.

The most modern data centers are emerging, supercomputers and software that translate 57 languages could testify that this is indeed the case, but when looking at where these tools are in practice, the results are not so impressive. Only 1.8 percent of companies in Serbia use artificial intelligence in their operations, according to a Eurostat survey, which is not only far below the European average, but we have also been well surpassed by the countries in the region.

The result does not surprise many local experts, who warn that we have not yet fulfilled the steps that are prerequisites for the development of artificial intelligence, namely digitalization, automation, and infrastructure. Some estimates indicate that more than 50 percent of companies in Serbia are still waiting to invest in digital business, and the information systems in use are at least three generations behind what is currently relevant in the world.

For comparison, the extent of our problem in the application of artificial intelligence in business is best explained by the fact that the European average in this segment is eight percent, while some countries such as Denmark and Finland are significantly above that, with around 15 percent. Among the countries in the region, if we do not compare Slovenia, which is at the top with an above-average 11.4 percent, Croatia has the best result with 7.8 percent, followed by Montenegro with 5.6 percent, while in Bosnia and Herzegovina 5.3 percent of companies rely on smart tools in their operations.

All this contradicts the data that came from the top of the government last year, claiming that we were 57th out of 193 countries in readiness for the development of AI tools, as the undisputed leader in the region, leaving behind Croatia (70th place), Montenegro (78th place), North Macedonia (83rd place), Albania (89th place), and Bosnia and Herzegovina (117th place), as well as the praise that in 2019 we became the first country in Southeast Europe and the 26th in the world to adopt a strategy for the development of artificial intelligence.

Lack of infrastructure

According to a study conducted by Fortune Business Insights, the entire AI technology market was worth over 500 billion dollars last year, and thanks to rapid technological progress, predictions are that by 2032 this figure will exceed two trillion dollars. A global study conducted by PwC consultancy on artificial intelligence shows that almost half of the economic benefits by 2030 will come from product improvements and boosting consumer interest, where artificial intelligence could contribute to GDP growth of up to 9.9 percent in northern Europe and 11.5 percent in southern Europe.

Serbia will invest 100 million euros in the development of artificial intelligence over the next two years. By the end of 2026, plans include investing 30 million euros in another supercomputer, which will be free for researchers and startups, 20 million euros in software for the public sector, primarily in healthcare, energy, and transport, as well as 30 million euros in incentives for the development of artificial intelligence, according to an announcement from the Government.

Zoran Ševarac, a professor at the Department of Software Engineering and the Artificial Intelligence Laboratory at the Faculty of Organizational Sciences (FON), says that the application of artificial intelligence requires infrastructure, which we do not have, and that it is not enough to have resources – processes must be set up to extract value from them and serve a purpose.

“It’s not just about saying we have resources for artificial intelligence – we need to have practical application and used AI somewhere. In our case, all those information systems are three generations behind what is current in the world. Of course, this varies from company to company, but generally, the level of integration is low. Email, Excel spreadsheets, and files are still heavily used, and AI is not fully integrated and connected. If entire processes were fully supported through information systems, then you would have some kind of upgrade with artificial intelligence, but without that infrastructure, the application of artificial intelligence is almost impossible. Primarily because you cannot collect data, and even if you succeed in that, where will you place that model to work? It’s like trying to create the Internet while still not having electricity. That’s our level and that’s where I see the key problem,” the professor points out.

He explains that some companies in Serbia use AI tools in their operations, but this is at a very low level, and in this process, we cannot boast that we are leaders, as government officials often claim.

“We cannot be leaders just because we construct a building and buy equipment. That is one of the prerequisites, but it’s the wrong sequence of steps. We spent money on a data center and that’s all well and good, but the question is what it will be used for, who will use it, and for what, and what is the benefit of all that. First, digitalization must be completed so that the information structure can support the application of artificial intelligence. The bureaucratic framework around data usage needs to be less rigid. And the goals and expected results need to be defined,” the professor adds.

Dr Dejan Mirčetić, a research associate at the Institute for Artificial Intelligence in Serbia, shares a similar opinion regarding infrastructure, but also emphasizes that the level of readiness of companies to apply AI tools varies depending on the sector and individual company.

“Factors such as digital infrastructure, the availability of data experts and capital, as well as the general culture of innovation, play a crucial role in determining a company’s readiness for AI implementation. Additionally, the application of AI is not just a technological issue but also requires a change in business culture. Therefore, in the era of modern challenges, the synergy between academic and research institutions and the industrial sector is extremely important for successful technology transfer and stimulating the growth of the startup ecosystem,” concludes Dr Mirčetić.

Where is AI Used in Serbia

In Serbia, smart tools are primarily used by technology companies as well as large domestic and international chains, which even develop AI solutions tailored to their own businesses.

AI tools are used to enhance internal processes, in communication with employees, but also as customer support, for checking complaints, identifying faulty invoices, or providing information.

Although not to a large extent, artificial intelligence is also applied in healthcare, for diagnostics and analysis of medical data, and in mammography. Its application foresees significant benefits in agriculture, and today some households use AI to optimize production and predict plant diseases. In the financial sector, AI is used for risk analysis, fraud detection, and personalizing banking services, and even for client assessment.

In transport and logistics, artificial intelligence helps in route optimization and predictive vehicle maintenance, while the telecommunications sector uses AI for analyzing user data, predicting failures, and improving user experience. Perhaps most familiar to us is artificial intelligence in retail, where smart tools tailor personalized offers and send them to consumers via apps, analyze their habits, and manage inventory.

(NIN, 05.07.2024)

This post is also available in: Italiano

Share this post

Leave a Reply

Your email address will not be published. Required fields are marked *

scroll to top