Looking for Trouble: AI, Writing, and the Labor of Thought
摘要
AI writing tools extend a problem Plato identified in the Phaedrus. Where Socrates warned that writing would give readers words without the author’s knowledge, AI writing tools now risk giving authors words without knowledge. With AI, students can produce competent academic prose about concepts they have not genuinely grasped, mistaking the generation of text as intellectual comprehension. Recent neuroscience research on “cognitive debt” shows that writing with AI can significantly reduce brain connectivity associated with deep learning, with effects persisting even after the tool is removed. Drawing on Plato, Dewey’s philosophy of educative experience, and Haraway’s call to “stay with the trouble,” this chapter argues that the cognitive struggle AI promises to let us bypass is actually a critical component of learning. To help students and faculty notice how AI reshapes the writing experience, the paper demonstrates a posthuman framework for making visible the relations that shape writing encounters, organized around six relational dimensions: Mind, Body, Time, Space, Sociocultural, and Technological. A structured relational audit translates this framework into a practical exercise that guides writers through phenomenological investigation of their process. A comparative audit of writing with AI and writing on a manual typewriter illustrates the framework in practice. The chapter addresses ethical reasons for refusing AI tools, including hidden labor and environmental costs. The framework equips writers to see what is at stake when they use AI, restoring the opportunity of deliberate choice. It does not prescribe whether or not to use the tools.