Accelerated technology? No.

“Technological change is accelerating,” I hear at a symposium. They’re talking about Artificial Intelligence (AI), but it’s a phrase I’ve heard, in one context or another, since I was a kid.

However, the phrase doesn't cease to be false just because it's repeated more than once. In reality, technological change is slowing down.

The idea that AI will bring us a time of abundance is based on flawed premises.

Consider the case of my grandfather Puig, who was born in 1888 into a family well-off enough that his father never had to work, despite having 12 children. In his world, there was no running water or electricity—and therefore no household appliances—cooking was done with wood or coal, and travel was done on foot or by cart. By the time he died in 1974, he was familiar with electric lights, the washing machine, the refrigerator, the gas stove, the telephone, antibiotics, the automobile, and the airplane.

Let's now consider my case. Over the course of my life, the utilities my grandfather already had will have been paid by direct debit, incandescent light bulbs will have been replaced by LEDs, gas will have been replaced by induction, landlines will have been replaced by cell phones, and, of course, the internet will have been introduced.

It's obvious that technological change was far more significant in our grandparents' lives than it was in ours. Robert J. Gordon reached this same conclusion in his monumental examination of technological innovations in the United States ("The Rise and Fall of American Growth"), where he concluded that the radical change in living standards took place between 1870 and 1970, and that subsequent developments have centered on entertainment and information.

The idea that AI can bring us a period of abundance comparable to that provided by the Industrial Revolution is based on two flawed premises. First, that AI is intelligent. AI mimics intelligence, but does not emulate it, so algorithms are unable to reach novel conclusions that a young child would reach if given much less information, and the hope of experts is that intelligence will appear in some unknown way as an emergent property from structures that mimic it. In fact, the very name was a tactic by the researchers involved to attract financial resources.

The second erroneous premise is that the material progress we enjoy is the exclusive consequence of human intelligence (this is the basis of Sala and Martin's two volumes on the subject). In reality, it is a consequence of the application of intelligence to the cheap energy sources—that is, fossil fuels—we have found on Earth. To the extent that we have decided to do without them, we must forgo the growth of material abundance.

I think it's reasonable to envision a future with more therapies, more robots, and more virtual experiences, but not more material abundance. It's certainly not a small amount, and therefore there's no need to exaggerate it.