Artificial intelligence can now analyze vast amounts of data, find patterns, and suggest solutions faster than humans. The question of whether it can replace human decision-making is increasingly being asked. But practice shows that even where AI can significantly help, the responsibility – and the final decision – remains with humans.
Today, we at GRiT are addressing the topic of artificial intelligence within corporate processes. AI is becoming part of the tools that companies use every day – including in the area of processing received invoices, where it is also newly used by iNVOiCE FLOW .
Andrej Gono , founder of the startup Citymind , has also been working on this topic for a long time. In his work, he connects the development of artificial intelligence, its practical application, and academic research.
Citymind focuses on changing the way businesses, institutions and communities operate through artificial intelligence. Their goal is to make advanced AI technologies accessible in a way that is understandable, intuitive and usable across various sectors – from public administration to universities to businesses.
Instead of experimenting with technologies, they design smart AI solutions that simplify communication, automate routine activities, and give organizations better insight into data and processes.
In addition to managing the company, Andrej is studying for a doctorate at Mendel University in Brno , where he focuses on artificial intelligence, and collaborates on research and teaching in the field of digitalization. Thanks to this, he has direct contact with how AI is developing in the academic environment and at the same time with where its possibilities are hitting limits in practice.
AI helps structure options. But it doesn't make decisions
" At any given moment, I have a hundred different options for how to take the company forward, " Andrej Gono describes his own business experience. " I can expand to Poland, Austria or Germany. I can stay in the Czech Republic. I can go for a cheap solution or try to differentiate significantly. "
It is in such situations that artificial intelligence can be very useful. It can name these options, analyze their benefits and risks, compare them in terms of data. “ It can tell me that there are five options that make economic sense, but each one leads in a different direction. ” But that is where its role ends. “ The final decision – and especially the responsibility for it – remains with me, ” she says. An algorithm can show the path that works best in numbers, but it cannot take into account personal values, long-term relationships or the boundaries that a person sets for themselves.
“ AI can easily recommend that I expand to China because that’s where the biggest profit potential lies. But I can decide not to go that route. And that’s something that an algorithm can never fully replace. ”
A tool that counts – not an authority that decides
According to Andrej Gon, it is essential to understand artificial intelligence as a tool , not as an authority. " I often compare it to a calculator. It will calculate numbers for you correctly and quickly. But it will not decide for you whether to use the result and how ." Decision-making is not a purely rational process. It involves intuition, experience, responsibility towards the people around you, and personal values. " People often make irrational decisions. And this combination is, in my opinion, irreplaceable. "
He therefore considers efforts to position AI as a replacement for human judgment a dead end. A meaningfully designed system should help a person navigate the options, not take control of the decision.
Where are the limits of artificial intelligence?
One of the places where the limits of artificial intelligence are most evident is when a person consciously decides not to do something, even though it would make sense from a data or economic perspective.
“ There may be a decision that is optimal from a company perspective, but I know I don’t want to make it because it would break my relationships or go against my values, ” says Gono. That kind of context, he says, can’t be easily transferred to an algorithm. “ Our life situations are extremely complex. It’s influenced by our past, culture, relationships, personal responsibilities. You just can’t get all that into a model. ”
He himself illustrates this by comparing it to personal decisions: AI can help clarify pros and cons, but the decision itself always remains with the person. Not because artificial intelligence is not capable enough, but because human decision-making is based on a combination of experience, values, and personal context that cannot be fully described or translated into data.
That is why there are efforts to find completely different paths. In this context, Andrej Gono mentions, for example, Elon Musk's attempts to connect artificial intelligence directly to the human brain via neurological receptors in the brain.
“ In theory, this could give AI much deeper context – directly from human thinking and emotions, ” he says, but he adds that this is still a distant vision.
The view of a person standing between research and practice
The fact that Andrej Gono so consistently emphasizes these boundaries is also related to his work in the academic environment. As a doctoral student at Mendel University in Brno, he focuses on artificial intelligence from a research perspective and is also involved in teaching and research in the field of digitalization.
This allows them to see the difference between what is theoretically possible and what is sustainable in real systems in the long term. At the same time, they observe how easily critical thinking can be lost.
" I see it in students, for example. They even have emails generated that they could handle themselves. They use AI unnecessarily and often do not check the outputs, " he describes. He observes similar behavior in programming, where people take over the results of an algorithm without verification. " It is important not to lose the ability to think, make decisions and take responsibility. Everyone has to set that limit for themselves. "
People's roles are not changing - they are becoming more precise
According to Andrej Gon, automation does not mean the end of the human role, but rather its shift. Technical skills may be partially replaced, but the importance of human judgment increases.
" A person should have hard skills, but also be able to understand context, communicate and defend an opinion, " he says. It is the ability to make decisions in ambiguous situations, work with context and take responsibility that remains the domain of humans.
“ We should not be afraid, but learn to work with artificial intelligence,” he concludes. “If we use it correctly, it can be a very good helper. Not a replacement for human decision-making, but a support for it. ”
Are you dealing with invoice approval?
30 minutes, no obligation. We will show you how iNVOiCE FLOW fits into your ERP.
