This site is written with a large language model’s help. It organizes the entries, drafts them, and checks their sources. Every position is mine.
I say so openly because the stigma is real. Across 13 experiments, people who disclosed using AI were trusted less than people who did not. Students even trusted a teacher less for grading with AI than for grading with a teaching assistant. Same help, different helper.
That double standard is preposterous. At its best a language model is a careful editor and a tireless research assistant. At its worst it invents things, which is why the claims get checked. Nobody discounts a book for its editor. It enhances a writer’s work rather than replacing it.
Better still, it excels at exactly what writers try to master: words that inform and persuade, and entertain while they do it. So why the heck not use it? The journalist Jérôme Colombain put it plainly in September 2026: “Not using it would be the stupid thing.”
Checking claims is the remedy for all writing, with or without help. I suspect many writers use the same help and never say so. If this site’s help lowers the value of what it says, show me where.