1 Comment
User's avatar
The Interior Frontier's avatar

Signal 01 has a long history outside AI. Psychiatry has had exactly this argument about its own categories. In 2013 the director of the US National Institute of Mental Health said the DSM’s weakness was validity, not reliability: clinicians could agree on a label without anyone showing the label matched something real. A calibrated decision inside a wrong schema is the same problem with better arithmetic. And Gohel’s point that low-confidence cases show where the rulebook is incomplete is how good clinicians already treat the cases that don’t fit.