I think they meant a more vague definition software. Amazon isn't really a software company, but it used it to change distribution and squeeze inefficiencies out of the system. Same with Google and advertising. No one really knows what the ROI for a TV spot is, but every Adwords bidder knows their keywords' ROI, and if Adwords claimed TV dollars, that squeezed out inefficiency.
Yes, and we see this in disproportionate demand for software developers, and also in envy/resentment towards tech workers, who have benefitted the most in the recovery.
Sure, but this didn’t prevent recessions in 2001 and 2007; in fact, efficiency improvements have been occurring since the dawn of civilization. It’s not a lack of efficiency that causes recessions, it is over-expansion mediated by greed.
I don't know the intent, but I read the original comment as software's enabling function as well. So the direct sales of software may only be 2%, but I'd be curious if there's any data on the ancillary effects of software on GDP (like increased productivity through automation, increased business efficiency etc.)
Before software there was Jan an, operations management, industrial engineering, interchangeable parts, division of labor, agriculture... all the way back to math, etc. Software is just how modern day improvements are expressed.
I understand your point, but I think they can/do build upon one another. Agriculture is more productive than gathering, oxen are more productive than human farm labor, tractors moreso than oxen, and GPS controlled combines more than a tractor. But im not sure if we've quantified how much productivity is directly attributable to software