待翻譯：The importance of teaching students what AI can't do

AI 服務暫時不可用，以下為來源正文，待恢復後補全翻譯。

It’s disconcerting to envision human thought becoming redundant or deemed inferior due to advancements in artificial intelligence. But for all its ills, AI cannot be ignored or wished away. It’s here to stay, and I want my students to understand the stakes of becoming too reliant on this technology. I can’t blame them for wondering why they should dedicate so much time and effort to honing their writing and critical thinking skills if large language models appear to do a passable job. Yet this situation has major implications for culture and societies. It’s easy to envision a future where AI dependency leads to intellectual conformity and the erosion of critical thinking. Researchers are already documenting these effects, which threaten to render the hard work of wrestling with ideas a lost art. This is why I think it’s critical for educators to confront AI head-on. The key, I believe, is teaching students to distinguish AI’s pattern recognition skills from its inability to daydream, inhabit human perspectives and authentically express itself. Tell me what you think In a course I recently taught on Southern Gothic literature, I gave my students an exercise. They were tasked with prompting AI to analyze the “emotional atmosphere” of a scene from Carson McCullers’ 1940 novel “The Heart is a Lonely Hunter.” Many of the characters readers meet in Southern Gothic works like McCullers’ are freakish and brutal. Rendered from a distorted, unnatural lens, they speak in odd dialects and conjure what Flannery O’Connor called the “grotesque.” They live in isolated places that are stuck in the past and, to use another term of O'Connor’s, are “Christ-haunted.” Many readers have difficulty synthesizing the strange worlds of these characters. AI has an even harder time generalizing the complex tensions between faith and unbelief, or good and evil, that animate these stories.