Over-Reliance on AI May Silently Degrade Professional Skills

An academic paper published in the journal Human Affairs by researcher Kaisa Kärki explores how over-reliance on AI leads to unnoticeable professional skill degradation across fields like medicine, research, and software development. The study highlights that professionals often fail to notice this decline due to cognitive biases when interacting with seemingly persuasive large language models. This issue challenges traditional philosophical notions of moral responsibility, as practitioners may unknowingly make errors caused by undetected AI hallucinations or mistakes. Furthermore, it raises critical concerns about whether future generations will retain core human competencies and the ability to work independently without technological assistance. The paper suggests structural solutions, such as setting mandatory non-AI training phases in professional education, regularly evaluating senior staff on core competencies, and establishing industry regulations. Kärki emphasizes that this is a systemic societal issue rather than just an individual failing, requiring proactive risk management before widespread group skill loss occurs.