This cluster started with Maggie Appleton’s “A Treatise on AI Chatbots Undermining the Enlightenment”, which argues that sycophantic chatbots cut against Enlightenment values by rewarding passivity and flattering users. Judgment atrophies when someone else exercises it for you.
I’ve been pulling at a few threads from that idea. Digital priesthood asks what happens when models trained for user comfort start functioning like a soft authority class. Radical AI defaults is about skepticism as a default setting. Interface friction is the interface version, the little bit of resistance that makes you check yourself.
A non-sycophantic assistant should be polite and evidence-seeking. It should test claims before affirming a user’s view of themselves.
Which contexts should default to a kind of Socratic mode automatically? What kinds of friction improve reasoning without just making people leave?