Although computers and the internet have transformed society in significant ways, they have done far less than most people expect in improving economic productivity. The consistent 2% growth in post-World War II total factor productivity fell off the rails in the mid-1970s, and, despite a burst of improvement over 1995 to 2005 (as China opened to the world), growth in actual productivity remains dismal.
“What if AI ends up like the Internet” asks Eli Dourado, “transformative to our daily lives while somehow not actually delivering major productivity gains?”
Computers and the internet allowed improved monitoring of energy systems and demand prediction, but they did not counter anti-nuclear power sentiment or the drive toward sub-par “renewable” energies. Will machine learning end red tape or transform international lending policy?
How about housing? Would AI overcome the pernicious attraction to actively bad zoning and building requirements? Will AI overcome the U.S.’s seemingly inherent inability to build successful high-speed rail?
