Major innovation booms have historically coincided with bubbles and subsequent collapses - the 1920s had the most patents and innovation yet was followed by 1929 and the Great Depression - so one cannot assume AI innovations will create a productivity miracle quickly enough and in time to outweigh the other forces.