Negative effects of Generative AI on researchers: Publishing addiction, Dunning-Kruger effect and skill erosion

In this opinion piece, I strive to examine the negative effects of generative AI on researchers, highlighting three main issues: publishing addiction, the Dunning-Kruger effect, and skill erosion. First, generative AI may lead to publishing addiction. In neoliberal universities, merit is often based on the quantity of publications. Generative AI speeds up the writing and publishing process, causing researchers to focus on producing more work quickly rather than on quality. This shift may harm their well-being and relationships. Second, generative AI may worsen the Dunning-Kruger effect among researchers. Researchers might believe they possess expertise by merely engaging with AI-generated content. This overconfidence can mask their knowledge gaps, leading to a failure to recognize their own incompetence. Consequently, it may hinder learning and growth, as individuals might not seek further education or feedback. Lastly, reliance on generative AI may lead to skill erosion. As generative AI handles brainstorming, outlining, editing, and other scholarly activities, researchers might weaken their ability to develop rigorous research skills. I stress the importance of responsible AI use and ethical standards. Much like craftsmanship, true research requires careful effort and originality—qualities that AI cannot fully replicate. I also argue that efficiency in research writing is not the same as effectiveness. Just as King Midas learned to value life’s true treasures after his seemingly blessed golden touch was washed away, researchers should embrace intellectual humility and strive for excellence in their work.