Σημειώσεις επεισοδίου
I talk with Justin Ranton, who designs clinical AI at SmarterDX, about why the decisions that determine whether an AI product works happen upstream of the interface — not on the screen. We get into why a clearer, more convincing explanation can make a clinical AI more dangerous rather than safer; why a responsible output should keep clinicians checking instead of accepting; and where design ends and informatics and engineering begin. Along the way: automation bias, the risk hiding in tools that feel effortless, and the genomics dashboard he built on his own DNA.
"You could redesign that…
Ιστότοπος: kellysmith.me

