Tectonic Shifts

Can Intelligence Be Artificial?

The IT industry has a habit of producing big buzzwords. They are used to announce new products and technologies, often with the promise that “nothing will ever be the same again.” To be fair, every industry does it. But technological progress often triggers strong emotional reactions, so perhaps we simply notice it more when a major change is announced by the IT sector.

Every now and then, however, a buzzword turns out not to be just a marketing gimmick.

By now, probably no one needs an explanation of how the widespread adoption of artificial intelligence has changed the way we live and work. The discussion today is about where it is acceptable to use AI, how to protect individual rights and freedoms, and how to regulate a field that is developing faster than society can fully understand it. We are no longer debating whether AI matters. We are discussing what to do with it.

Every now and then, someone points out that “this isn't intelligence at all.” In a strict sense, that may be true, but it is not particularly important. A technology does not have to be intelligent in the same way humans are in order to produce results that, until recently, we associated with human intelligence.

Perhaps the more interesting question is what we actually mean by intelligence. Our formally organized civilization has long been trying to turn human knowledge and abilities into rules, procedures, and systems. Artificial intelligence is now doing something somewhat opposite: extracting patterns from vast amounts of data and producing results that look to us like intelligent behavior.

Whatever we choose to call it, the change is real.

A struggle for dominance and influence over the entire field of artificial intelligence is already underway. From the geopolitical rivalry between China and the United States and their fundamentally different approaches to AI development, to the battle among the largest technology companies for market dominance, to questions of regulation, energy consumption, and the environmental footprint of large data centers. There will be many more topics, debates, and turning points, but one thing is clear: AI has taken its place and it is not going away.

What interests us much more, however, is what happens when this technology stops being something we read about and becomes a tool we use every day.

Slobodna domena develops software, so the changes in the world of programming have directly affected us as well. It is a major shift. Some of the programming work we had considered fundamental for years has become unnecessary, or is now done in a completely different way. The changes are profound and they are happening fast – not just year by year, but month by month.

Programming has not disappeared. The way we program has changed.

As individuals, but also as an organization, we have to adapt, adopt new technologies, and develop new ways of working. Our team uses AI extensively in our everyday work. For programming, like many of our colleagues in other companies, we use subscriptions to external providers. At the same time, we also use local AI resources based on open-source technologies.

This led to the creation of SD AI Lab, a project for which we received support from the Ministry of Economy through a funding programme for the development of cooperative entrepreneurship. The project gave us access to hardware on which we can develop our own AI projects, learn, test, and play. It also showed us that there is a big difference between using AI as a ready-made service and gaining a deeper understanding of the technology. A subscription to a tool allows us to get something done. Having our own resources allows us to explore how and why it works, where its limits are, and what happens when we move beyond the expected scenario.

Another experience pushed us further in the same direction. As part of the City of Zagreb's 2025 public call for co-financing projects that make use of open data, we applied for and developed Puls grada. Using artificial intelligence, we analyzed the names of Zagreb's streets and tried to determine where their names came from. Who was the person after whom a street was named? Where is a particular place-name located? Is it an animal, a plant, a historical event, or something else?

A deeper analysis of street names reveals a great deal about how a city relates to its space and its own history. But the project brought us something else that we had not expected.

Our research method combined human work with artificial intelligence. The flexible structure of the project allowed us to experiment and assess, for each individual problem, when AI was a good tool and when it was not.

We encountered model bias, hallucinations, and problems with understanding the Croatian language. None of this was particularly surprising. What was more interesting was how much the way we work changes when a problem is no longer purely technical. That is when working with AI becomes a space for creativity.

In everyday programming with the help of AI, many things are relatively straightforward. You define a problem, get a suggestion, check the result, and move on. With Puls grada, we had to solve problems that had no single obvious technical solution. We had to assess context, the credibility of sources, and the results produced by the model. Sometimes we needed to question the model, sometimes to guide it, and sometimes simply to conclude that its answer was not on the right track.

This taught us something that now seems more important than simply knowing how to use the tool.

AI can be extremely useful in making a decision, but it should not be the one making the final decision.

This is not a rejection of technology. Quite the opposite. It is an attempt to understand its place in our work.

Predicting the future is not easy, but it is important to look a few steps ahead. Everything is changing; some things remain the same.

AI has stirred the waters, but it is still a technology. We can implement it, adopt it, and master it. We can use it well or badly. We can trust it too much or too little.

As a technically oriented cooperative, and as people who enjoy technology, we have set ourselves the task of using it appropriately – in the spirit of open source and with respect for users' privacy and rights. That means investing in our own capabilities, knowledge, and skills. Not just learning how to use new tools, but understanding their possibilities and limitations.

Perhaps that is why, in the end, the most important question is not whether intelligence is artificial. What matters more is what we do with it. If it helps us better understand a problem or create something we could not create before, then it matters less whose intelligence brought us to the result.