Will intelligent systems and robots soon replace people who work as engineers, programmers or mathematicians? The statement by labor market economist and Nobel Prize winner Christopher Pissarides caused a stir a few months ago when he warned of major upheavals in the labor market due to artificial intelligence (AI) in the field of MINT.

Anne Kokenbrink

Editor in business

  • Follow I follow

    “The skills that are needed now – collecting, merging, developing and using data to make AI usable for jobs – will make these very skills unnecessary later because AI will then do the work,” he said in a lecture warned against taking up a STEM degree course. The abbreviation MINT stands for mathematics, computer science, natural sciences and technology.

    Such statements, which were made similarly by the Swiss bank UBS, seem unsurprising. After all, AI is currently making huge leaps. AI expert Holger Hoos also says this. He is director of the AI ​​Center at RWTH Aachen University and Alexander von Humboldt Professor of Artificial Intelligence. “Development occurs in spurts that are very difficult to predict. You can see this in the example of ChatGPT. Suddenly something has become very good very quickly, something that even experts wouldn’t have thought could reach this level of quality so quickly.”

    However, he criticizes the polarization when it comes to AI. “Both the enthusiasm to say that soon we will no longer need all these job profiles, but also the great fears are often exaggerated. You have to see everything a little less extreme.” Actually, says Hoos, Nobel Prize winner Pissarides is also an AI optimist. Either way, Holger Hoos does not believe that the work of computer scientists will be fully automated in the foreseeable future.

    AI will provide more support

    Of course, he expects there will soon be more support from AI systems. “This is a development that is already underway and which I personally welcome because we can do more with it.” But it is also about understanding the systems in depth in order to be able to better assess, for example, when fail and when they cause problems.

    Ines Helm also sees it that way. As a professor at the Ludwig Maximilian University of Munich (LMU), she researches structural change in the labor market. “Humans still outperform AI in creative and social intelligence, reasoning skills, and ability to deal with uncertainty,” she says. So in most cases decisions will still be made by people. AI is more of a tool or decision-making aid.

    In practice, this could have positive effects. Helm gives an example: The Github Copilot, a programming aid based on generative AI. An analysis by American scientists last year showed that the Copilot can enormously increase the productivity of programmers: An experimental group that had access to the Github Copilot completed a programming task around 56 percent faster than the control group without access to it.

    Holger Hoos says that analytical and critical thinking will be all the more important in the future, regardless of which technology supports you. AI systems can only be used sensibly and responsibly by people who are good programmers themselves. “It is absolutely important that these subjects continue to be studied by people who have real talent and interest in them,” he says.

    One of them is Katarzyna Polewska. The 25-year-old student is currently in the third master's semester of her computer science studies at the BTU Cottbus. In her master's degree she focused on artificial intelligence. Whether she is disturbed by statements like those made by Nobel Prize winner Pissarides? No, just the opposite. “I talked to a lot of people at my university about it. Every person who works in IT here says this is nonsense. I don't want to be ignorant and say that AI won't take away jobs in the future. “It’s possible,” she says. “But it will still require computer scientists.”