One of the big concerns about AI is that it could cost people their jobs. Morgan Stanley Chief Global Economist Seth Carpenter says it's important to remember that "productivity gains have been part of the human economy since there have been humans, and it's not that it only goes in one direction." "If you can produce more things with fewer people, you can also produce more things, with the same number of people and not necessarily have to have a big wave of joblessness," Carpenter adds. Carpenter goes on to say that if it gets to the point "where the economy can just produce more at a lower cost, ultimately, that could be disinflationary in the short run." When it comes to what type of jobs could be impacted by generative AI, Carpenter says "information processing professions," including jobs like his where he analyzes markets, could be replaced, though more likely they will be supplemented by the technology.
Video Transcript
- The AI craze has been driving the market action since the start of the year with investors jumping in on the hype. NVIDIA is among the biggest beneficiaries with shares surging nearly 240% year to date. But what does it mean for the economy?
Here with a closer look is Seth Carpenter chief global economist at Morgan Stanley. Seth, you've got a new note out sort of looking at the implications for the labor market and the broader economy. And you've sort of say, look, take a step back, it's not all negative, it is about increased output. Ultimately, though, it could be disinflationary.
SETH CARPENTER: I think that's right. There's-- it's very easy to get excited about the topic. It's very easy to get caught up in all of the hype about it. And you see some very spectacular numbers about how many jobs could be destroyed or replaced or how many people could end up without work. And I think it's always important to keep in mind productivity gains have been part of the human economy since there have been humans.
And it's not that it only goes in one direction, right? So if you can produce more things with fewer people, you could also produce more things with the same number of people and not necessarily have to have a big wave of joblessness. And think it all comes down to where we are in the cycle, when the technology really starts to bear fruit, how that gets deployed, and then how businesses and individual markets really start to use the technology. And if it gets to the point where the economy can just produce more at a lower cost, ultimately, that could be disinflationary in the short run.
- So then, Seth, what does that mean then for Fed policy for Fed policy rates? Essentially here, I guess disinflationary factors playing in initially. But that longer term perspective, what does that look like?
SETH CARPENTER: It means lots of things. So it's very difficult right now to gauge just how soon any of those extra productivity gains would show up in the economy. Is it going to be a year from now, two years from now, three years from now, you know? My guess, and I will say it's hard to be precise right now-- my guess is that it's after we get past the current inflationary environment that we're in now and things do start to settle back down to normal.
But beyond that, maybe two, three, five years from now, then I think the Fed will have to ask themselves, OK, are we seeing more of the disinflationary force, in which case, maybe there'll be a little bit easier on policy than they would be otherwise, allow the economy to grow into that bigger productive capacity. But if productivity growth is just higher for five years, 10 years, 15 years in a row because of all of these new technological developments, that should ignite some extra desire for more investment spending and pull things forward, which means at the end of the day, you have a higher neutral interest rate.
So it'll be really choppy. It'll depend on when the Fed recognizes the extra productivity gains, and then how long those productivity gains last. In the short run, though, it does seem it should mean lower inflation and probably an easier time for the Central bank.
- Seth, there seems to be a real divide around this conversation on AI between those who say there will be mass job losses that come from it and those others who are particularly in tech who say, for all the fear of new tech, history shows that jobs ultimately were gained as a result of emerging technology. When you look at the labor market specifically, what are the jobs that are likely to be most affected? Is it what has traditionally been considered low skilled, or are we sort of not thinking about this in that way?
SETH CARPENTER: Well, I think this is part of what we still need to see. And I agree, I think there is a little too much pressure on how many jobs will be lost and not enough on how many new jobs might be created. If we go back to the 50s and 60s when computers were starting to be developed, I don't think anyone would have believed the sheer number of computer programmers that we have in this economy right now as a result of technological advances, computers getting faster, smaller, cheaper, more available.
And so one of the keys right now is what's called generative AI, where you're able to do much more with these large language models. And I think there we really could see part of the information processing professions. Maybe the legal profession. Maybe my profession and economic and financial market analysis either being replaced, or my guess is supplemented, being making it easier for people in the informational services to be able to do more with the same number of people.