I didn't claim that I could determine whether AI wrote the prose, or even parts of it. Nor am I trying to discourage folks from writing (actually the opposite, I even thanked Ben for the demo!).
What I observed are overused, and recognizable, stylistic patterns that have become quite common in LLM generated prose. In this article alone there are three "here's the X" rhetorical signposts, well over a dozen neatly constructed reversals/antitheses ("not X but Y," "solved X and created Y"), repeated setup->complication->punchline paragraph structures, rhetorical questions used as transitions, and a high density of anthropomorphized technical metaphors.
None of these alone prove AI authorship. In fact their existence (and the authors own words!) are exactly the point I am making: his validateHypotheticalHooks linter flags "formulaic AI-tell openers," and he describes catching himself writing one because he'd "absorbed the cadence from reading too much generated text."
Comments
I didn't claim that I could determine whether AI wrote the prose, or even parts of it. Nor am I trying to discourage folks from writing (actually the opposite, I even thanked Ben for the demo!).
What I observed are overused, and recognizable, stylistic patterns that have become quite common in LLM generated prose. In this article alone there are three "here's the X" rhetorical signposts, well over a dozen neatly constructed reversals/antitheses ("not X but Y," "solved X and created Y"), repeated setup->complication->punchline paragraph structures, rhetorical questions used as transitions, and a high density of anthropomorphized technical metaphors.
None of these alone prove AI authorship. In fact their existence (and the authors own words!) are exactly the point I am making: his validateHypotheticalHooks linter flags "formulaic AI-tell openers," and he describes catching himself writing one because he'd "absorbed the cadence from reading too much generated text."