Igor Kranjčec: Artificial intelligence doesn't forgive wrong data, but it frees up time for what's important

Igor Kranjčec, founder of the Croatian tech startup Farseer, spoke at a recent event hosted by Hrvoje Krešić in Novi Dan about the impact of artificial intelligence on the labor market. His remarks highlight a shift in how experts and industry leaders view AI’s role, moving away from fears of mass job losses toward a more nuanced understanding of its transformative potential. While AI does not absolve users of incorrect data, it offers time savings for critical tasks, according to Kranjec. In the past few months, discussions around AI's influence have evolved significantly. Initially, predictions suggested widespread layoffs and the disappearance of certain professions. However, leading experts and industry leaders now emphasize that AI brings dramatic adaptation rather than necessarily ending jobs. The conversation has shifted toward preparation for this new era, including the reasons behind delayed financial sector integration and how Farseer, a promising Croatian startup, views business evolution. These insights were shared during a study session with N1 Television. Drawing parallels with earlier industrial revolutions, Kranjec notes that similar concerns arose with the advent of steam engines, assembly lines, and the internet. He explains that while industries inevitably evolve, some roles will disappear, but new opportunities will emerge. For survival in today’s market, he stresses the importance of adapting one’s approach and acquiring specific skills. Kranjec highlights critical thinking as one of the most essential skills currently. He points out that everyone now has access to tools that allow them to generate prompts and develop applications using different platforms. This democratization of technology means that individuals must critically evaluate outputs to ensure accuracy and usability. He emphasizes that knowledge of whether the results obtained are correct is vital. When asked for advice to those preparing for the future, Kranjec cautions against complacency. He warns that no one should feel satisfied immediately and that systems of artificial intelligence should not be approached carelessly. He suggests that conversations with these systems should be treated as dialogues meant to confirm existing beliefs rather than challenge them. Instead, users should verify information through traditional methods such as searching online or consulting books. He adds that although this process initially creates more work, it ultimately enables secure automation over time. The application of AI in finance presents particular challenges, as inaccurate numbers can lead to massive losses. Kranjec explains that AI in the financial sector arrived relatively late due to its sensitivity. At the core of Farseer’s operations lies the precise alignment of input data, which is crucial for accurate outcomes. He underscores that if the input is incorrect, the output will certainly be unreliable. Therefore, Farseer places great emphasis on ensuring the accuracy of initial data. “Once AI models are fed accurate data, they operate exclusively based on that information, guaranteeing no hallucinations,” Kranjec says. “This allows analysis to be conducted faster and better than any human. Yet, it is essential to know that someone always stands behind the number. No one in management will accept an excuse that an error was made by AI, the person who signs off on the figure is responsible.” Although 80% of large companies still perform calculations using Excel, this method becomes increasingly inadequate when planning operations involving dozens of people or making complex forecasts. As a result, trends in the global fintech sector are moving towards integrating external parameters into models. An example of this shift is evident in the growing reliance on AI-driven analytics to support decision-making processes.