Can A.I. produce writing that we want to read?

# Summary The article explores whether AI can produce writing indistinguishable from human work, particularly as it threatens traditional academic integrity. While current AI writing contains detectable flaws and stylistic tells, the author argues that as technology improves, it will become functionally impossible to distinguish AI-generated text from human writing in most contexts—potentially eliminating educators' ability to enforce academic honesty. The recent controversy over a Granta literary prize winner suspected of being AI-written highlights this emerging problem and raises questions about the future of writing and higher education.