-
The Most Agreeable Friend You Will Ever Have Is the Worst One for You
An AI companion optimized never to challenge you removes the friction that real relationships use to help people grow, making frictionless comfort developmentally inert.
-
The Turing Test Was Never a Test of Intelligence
Turing proposed a test of whether a machine could pass for human in conversation — a social question. Seventy years of treating it as a benchmark for intelligence has confused the field.
-
Psychiatry Is Quietly Abandoning Its Own Diagnostic Manual
The DSM still rules the clinic, but psychiatry's leading research frameworks now treat its categories as useful fictions rather than natural kinds of illness.
-
If a Chatbot Prevents One Suicide and Causes Another, How Do We Count It?
Deployed to millions, a mental-health chatbot will both prevent and cause harm. Scale forces us to do explicitly the moral arithmetic that medicine usually keeps implicit.
-
Loneliness Is a Signal, and AI Companions Are Learning to Silence It
Loneliness evolved as an alarm that pushes people back toward others; AI companions are getting good at muting the alarm without meeting the need behind it.