AI使用引发认知能力担忧

A growing body of research is beginning to examine whether heavy reliance on generative AI can weaken the mental processes people are supposed to practice when they write, study, and solve problems. The most widely discussed example is a study led by MIT research scientist Nataliya Kosmyna, which used electroencephalograms to monitor brain activity while participants wrote short essays under different conditions. The study divided 54 participants into three groups. One group used ChatGPT, another used Google Search without AI-generated summaries, and a third wrote using only their own knowledge. Across multiple sessions, researchers measured neural activity, analyzed essay patterns, evaluated writing quality, and asked participants about their sense of ownership over the work. …