New book argues we can’t expect new tech to rekindle rapid economic growth

But then it slowed down after 2005 and now Bob Gordon is back with his theories about "productivity is over" and growth is finished etc. The new twist is that he claims today's technologies are not as important to productivity as the early 20th century were. Well, the 1988 paper has comments by William Nordhus who shows that in the period from 1900 to 1917 productivity growth was zero and it was very low between 1924 to 1937. For the whole period the productivity growth came in a burst from 1917-1924 and then 1937 to about 1948. The reason is productivity is based on combining technologies in novel ways and it doesn't proceed in a regular 3% manner like population growth or investment. It is likely "bursty" and non-uniform. I fully expect to see another explosion of productivity (and thus economic growth) as mobile/cloud/AI/ML start infiltrating the general economic system.

[1] http://www.brookings.edu/~/media/Projects/BPEA/1988-2/1988b_...

Until then, there isn't reason to think a tectonic shift will easily happen.

Steamships and canals were a powerful revolution, opening up huge parts of the country to international trade, captured well by how the Erie Canal opened the midwest up to the Atlantic.

Rail was another powerful revolution, allowing people to be in New York City and Chicago on the same day. Upon being in both cities on the same day, people observed that the cities didn't have their clocks set the same way, as that was never possible before.

Instead of a new human transportation revolution, the information age gave us a way for ideas to travel rapidly. At the speed of light in some cases. This is certainly profound, but consider how HUGE an idea this is. If we look at the start of the information age as 1971 when Intel opened its doors, then we're already 45 years into this one. A lot has been done. There's still more left, but it's probably about upgrading the rest of the planet instead of finding whatever US industries are left to be upgraded.

If I were to speculate, I might say the next revolution that could change society in a comparably profound way would be self-driving cars and other forms of automation. Perhaps around 2025-2030 society will be cool with cars driving themselves and many, many businesses will be started to explore the automation space. To compare, there were over 1800 car companies in the US between 1896 and 1930, but today there are 2 US car companies. Another comparison, there were once 500 US steel companies, but today there's basically 2. So perhaps 2030 starts with self-driving cars moving the same speed as human-driven cars and maybe by 2080 they go 300mph instead, which would upgrade humanity's transportation capabilities in a similar manner to when rail upgraded steamships or when cars upgraded rail. This would open up huge amounts of land that is otherwise inconvenient and could give rise to new housing technologies, as new houses would surely be built on this newly accessible land.

We are 45 years into the information age. It's completely reasonable that it's not blowing our mind asunder the way it did 10, 20, 30 years ago. I am VERY excited for whatever comes next.

I think in the short term we won't get a lot of economic growth. But I think these things happen in phases.

We are waiting for the next big breakthroughs. I think Artificial Intelligence, 3D printers, robotics and cheap sustainable power will be the key things to create a new potential for high growth.

E.g. if we perfect the technology to 3D print large structures that should bring down costs for house construction a lot. Companies are starting to do this now. But we are in very early stages.

AI and more advance robotics will allow us to automate so many tasks we are stuck with using humans now which really limits productivity growth. It is very hard to increase output from a doctor, teacher, lawyer, cab driver, cooks etc at the same rate as say a car factory.

This sort of development doesn't go in a gradual phase. As we develop AI we are not gradually making medical diagnosis cheaper or driving cheaper. The huge cost reductions will come when these technologies are completed and can start replacing people at massive scale. Once cars, trucks etc can be driven by and AI, transportation costs will get huge cost reductions.

Likewise with medical science. If an AI can diagnose you, one can save enormous amounts of money.

The Kaldor-Kalecki nonlinear dynamic business cycle model, the Kaleckian profit equation and Robinson's growth model are probably the most interesting results out of that era that still retain an influence on heterodox thinkers.

That said, one should be very wary of all such simplifying models since they all carry contestable assumptions. Nonetheless, they do pose intellectual curiosities.

I don't know if Robert J. Gordon explores this. The Post's review mostly rehashes the more naive dystopian arguments, so I'll have to check.

Let's not start saying everyone who maintains a position you don't like over a long term is "on a Jihad".

> I found a 1988(!!) paper where he is complaining that productivity has been slowing down for 20 years (in 1988)

Which is consistent with the claim that 1870-1970 was a special century for technology-driven productivity gains.

-Most tools we still have for enabling productivity are such crap - trying to emulate paper document based workflow, instead of offering a task oriented one.

- AI:s can probably observe employee patterns and raise red flags if someone seems to be goofing off. Less need to spend time in reporting, more time for value creation.

That's a bold misreading of the macro environment. Competition for funds expresses itself in higher rates, not lower. If the author's "competition for funds" intensifies further I think we may be in danger of being crushed under piles of money.