Surveying the landscape of interpretive AI in radiology, two researchers note a yawning gap between great expectations set in the recent past and actual clinical implementations as of spring 2023.