| Summary: | The article examines the evolving role of large language models (LLMs) in programming and broader intellectual work, framing their impact through the metaphor of "magic" as a screen that obscures understanding. It highlights how LLMs have increasingly automated tasks once central to programmers—such as researching, writing, and puzzling through code—leading to a decline in traditional programming textual communities and skills. This shift parallels concerns about the erosion of literate scholarly communities, as reliance on AI-generated summaries and outputs may weaken engagement with disciplinary norms and original texts. The author situates these changes within a cultural context referencing Tolkien’s concept of magic and power, questioning what is lost as human creativity and communal knowledge-making are supplanted by AI approximations of intellectual labor.
|
|---|