💡 A Quick Q&A with … economist Michael Strain on the economic future of AI

'I think the technology just seems too powerful to bet against it having a productivity impact.'

Quote of the Issue

"Things are only impossible until they are not." - Jean-Luc Picard, Star Trek: The Next Generation.

---

I have a book out: The Conservative Futurist: How To Create the Sci-Fi World We Were Promised is currently available pretty much everywhere. I’m very excited about it! Let’s gooooo! ⏩🆙↗⤴📈

• Amazon

• Barnes & Noble

• Books-A-Million

• Target

• Walmart

• Bookshop

---

Q&A

💡 A Quick Q&A with … economist Michael Strain on the economic future of AI

Michael Strain is the director of Economic Policy Studies and the Arthur F. Burns Scholar in Political Economy at the American Enterprise Institute, where he oversees the institute’s work in economic policy. Strain is concurrently Professor of Practice in the McCourt School of Public Policy at Georgetown University. He is also serves on an ad hoc committee of the National Academies of Sciences, Engineering, and Medicine that is studying the "current and future impact of artificial intelligence on the workforce of the United States across sectors." Strain is the author of “The American Dream Is Not Dead: (But Populism Could Kill It).”

1/ ChatGPT was released in November of last year. What is your current opinion as to whether gen AI, large language models, are an important general purpose technology? What have you seen over the last year that makes you think they are, or aren't, or too soon to tell?

I think it's pretty clear that they will be an important general use technology. They're clearly capable of providing services that will be of value to businesses, that will make workers more productive, and those services will be attractive to businesses and workers in a broad range of industries and across a broad swath of occupations, similar to the internet, similar to email, similar to the personal computer, similar to electricity. This technology seems very likely to touch millions of workers in many different kinds of jobs on a daily basis.

2/ Is it an important enough technology to change current expectations about long-term economic and productivity growth?

I think that the best forecast is that this will have a measurable impact on productivity. Whether or not there will be a discreet jump in the productivity time series I think is a very different question. I could easily imagine us looking back 20 years from now and seeing that this technology hits some industries much sooner than other industries, hit some types of workers much sooner than other types of workers, and that the extent to which it diffused throughout the economy was much more gradual than what would be required for kind of a discrete jump in the productivity figures.

I think the technology just seems too powerful to bet against it having a productivity impact. I'm of the view that the impact it does have on productivity will take a little longer than I think a lot of analysts expect. There are reasons to think that this technology will spread throughout the economy really rapidly, and much more rapidly than what we've seen with previous leaps forward in technology.

For example, unlike electrification, the infrastructure required to use generative AI already exists among everybody who uses the Microsoft Office Suite, everyone who uses the Google Office Suite. That's a reasonable argument, for sure, but I think there are going to be a lot of obstacles to the widespread adoption of generative AI by companies, and I think those obstacles are going to require solutions—in some cases, enterprise solutions that deal with privacy concerns—and those solutions will raise the cost and lower the benefit of the technology, which will slow it's adoption. I also think that this technology is going to be regulated, and that's going to have an effect on the speed with which it's adopted.