Moody's Talks - Inside Economics - CPI and AI with Capital Group's Jared Franz

Episode Date: September 12, 2025

Mark and Cris are joined by Matt Colyar to break down the latest CPI inflation report, while Jared Franz from the Capital Group explores how artificial intelligence is reshaping the American economy and labor market. We examine the opportunities and challenges of the AI revolution and what it means for workers, businesses, and investors in this rapidly changing economic landscape.Jared Franz is an economist at Capital Group, responsible for covering the United States. He has 19 years of investment industry experience and has been with Capital Group for 10 years. Prior to joining Capital, Jared was head of international macroeconomic research at Hartford Investment Management Company. Before that, he was an international and U.S. economist at T. Rowe Price. He holds a PhD in economics from the University of Illinois at Chicago, a bachelor’s degree in mathematics from Northwestern University and attended the U.S. Naval Academy. He is also a member of the Forecasters Club of New York, an elected member of the Conference of Business Economists and a member of the Pacific Council. Jared is based in Los Angeles.Explore more insights from Capital Group’s Jared Franz in the articles below:4 charts on why the U.S. economy could stay resilient | Capital GroupBenjamin Button’s clues for the US economy Explore the risks and realities shaping the economy in our new webinar, now streaming for free.U.S. Economic Outlook: Under Unprecedented UncertaintyWatch here: https://events.moodys.com/mc68453-wbn-2025-mau25777-us-macro-outlook-precipice-recession?mkt_tok=OT…Hosts: Mark Zandi – Chief Economist, Moody’s Analytics, Cris deRitis – Deputy Chief Economist, Moody’s Analytics, and Marisa DiNatale – Senior Director - Head of Global Forecasting, Moody’s AnalyticsFollow Mark Zandi on 'X' and BlueSky @MarkZandi, Cris deRitis on LinkedIn, and Marisa DiNatale on LinkedIn Questions or Comments, please email us at [e-mail]. We would love to hear from you.  To stay informed and follow the insights of Moody's Analytics economists, visit Economic View. Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

Welcome to Inside Economics. I'm Mark Sandy, the chief economist of Moody's Analytics, and I'm joined by one of my trusty co-host, Chris DeReedy's. Hey, Chris. Hey, Mark, good to see you. Where's Marissa, our other co-host? She's traveling. We have a big conference next week, so she'll be here for that, and we have some internal

meetings as well. So she's traveling. Yeah, I was just looking at the calendar. It feels like we got a lot of stuff going on next week. We do. We do. Packed. All good stuff.

All good stuff. All hands day. we call it. Does everyone call it all hands day? Is that like a universal? That's an industry term. Yeah. It is an industry term. Okay. Meaning we bring in our colleagues from the rest of the country. I think this is the Americas, not the rest of the world, but that's right. And we spend a day, at least a day together, and Mercer's coming in for that. And then we have our convening, was convening the right word, or a conference, I guess, with clients in New York. Conference, yeah. Seven World Trade. That's Moody's HQ. That'll be fun.

in New York. Yeah. Good. Well, we'll miss her. But we have one of our other colleagues who always joins us the week of CPI, consumer price inflation. That's Matt Collier. Hey, Matt. How are you? I'm doing very well. How are you, Mark? Well, I was doing okay until I got some thomo. You told me you're on your way to see the Eagles play the Chiefs in Kansas City. That's right. That's why I'm incognito off camera headed to Kansas City in a couple hours my friends and I that I went to like basically kindergarten with we do an away Eagles game

every year and that's this weekend can I can I be a friend to Matt I think we're getting closer this is a regular appearance on the podcast that's developed that's my view okay Chris you're invited oh thank is your friends named Taylor

by any chance or not any Taylor's on this trip no but she is a PA girl, right? Exactly. You know, there's a chance. Yeah. Yeah, yeah, you make a great point.

Didn't she, like, grow up in, like, Redding or something, Pennsylvania, somewhere in that area? That sounds right. Sounds right. Yeah, yeah. Yeah, not too far away. Okay. And we have a guest. Jared, Jared, Franz.

Good to see you, Jared. You too, Mark. Thanks for having me. Great to be with you. Pardon me? Great to be with you. Oh, well, it's great to have you. I'll have to say, though, you know, you're, you're,

sitting somewhere in Los Angeles, right? That's kind of HQ for the capital group. That's where you are right now, right? Yeah, that's right. I'm in a sunny L.A., though today it's a little cloudy, so we're not getting the sun vibes today. But yeah, it is Los Angeles. Well, I just have to tell you, Jerry, you look more corporate than the rest of us do. I don't know what I mean, I thought you know, you're in L.A. shouldn't you be wearing? I don't know. What should he be wearing, Chris? It's definitely not what he's wearing. How sunglasses? Sunglasses.

Yeah. No. Palm trees in the background. This is me every day. I'm nose down, hardcore. I'm pumping out magical boards. I'm doing data analysis. Like, I never quit.

Yeah, I know. You're as nerdy as we are. Thank God. Can I ask, Jared, are you the chief economist of Capital Group? What do you call you? Well, we don't have a chief economist role at Capital Group, and that's one of our, I think, secrets is that we're very flat as an organization, even analysts. We don't have like senior analysts and junior analysts or senior PMs and junior PMs.

We have economists, and we have in our group, in addition to economists, we also have political analysts. So we have our political team in our group. And we also have our accounting experts. And so that group makes up what we call capital strategy. research. But yeah, there is no chief economist, and we kind of like it that way, actually. You know, Chris, somehow I feel like that was a dig at me. I don't know. You know? No, no. I don't think. Well, you know, it makes a total sense to me. I mean, Moody's has been around for how long has Moody's been around? A hundred plus years? Very hierarchical. Very hierarchical.

but it's great to have you on. And, you know, capital group is huge. You know, is it fair to call capital group an investment fund? Is that how would you describe capital groups? I think of it like, you know, the Fidelity, Vanguard's capital group. Is that kind of sort of the right way you think to think about it? Yeah, I think we're, I like to think about it as. the largest investment firm you've never heard of. And because we're over three trillion in assets,

so that's pretty big. We're usually a top three owners in most countries for equities, for example. And so we're diverse, we're global, we're fairly large. But actually kind of comes back to your prior point, Mark, the founders, the Lovelace family, came to L.A. to be differentiated, to be away from New York, to be away from the buzz, and to really curate this independent thinking. And so that's why we're here. We've had a history of being low-key. Not even, I mean, this would be, you know, five, ten years ago would be insane. We would never do something like this. But as the market has evolved as the industry has evolved and our clients want to hear more from us. And so we've evolved over time to be more in tune and partnering with our clients in areas like this that we know

they want to hear us. So appreciate you having me here. Yeah, you know, it's a great, it's a great company. My daughter works at Capital Group in London. You guys are global, of course. And she was in a a program early on when she was at CAP group and was kind of doing different rotations. I'm sure I'm describing this incorrectly, but I think the listener gets the gist. And you were one of her rotations, I think, and she thinks very highly of you, Majority. I'm not making that up. She really enjoyed working with you and learned a lot. And I was just listening in on the kinds of things you guys were working on. I think she's got like the best job on the planet to tell you the truth.

I wish I had her job. She, well, first of all, I was terrified when I learned that as Zandi, I would be sponsoring a Zandi in a rotation. I'm like, is your dad, Mark? And so I was a bit terrified being, you know, having to live up. up to the large expectations of the Zandi clan. So I was for, I'm fortunate that I didn't do anything wrong during that. She did, I mean, she did rotate with us in this group called Capital Strategy Research.

She did a great job. She was part of our program that we call the the CAT program where we bring in graduates from college very early on. and it's a two-year rotation. They rotate throughout the firm. We hope to place them in the investment group that I'm in and that we are a part of. And that's what she did. And we did a project on macro and the project was really related.

And a lot of the work that I do at Capital is less about, you know, marking to market where GDP is today and, you know, where, you know, what decimal place the CPI came under. but more about connecting why the macro environment matters for an equity portfolio or why it matters for a caterpillar when you buy or sell caterpillar or why it matters for, you know, Gen A.I or something. So we're really long-term. And that's the project she worked on, really connecting some of this economic data with some of our equity holdings. Well, she had a wonderful experience.

and still is having a wonderful experience in CAP group. It's really been, and I just love chatting with her about what she's doing. And it's long equity. And it just feels like the funnest thing on the planet to do for a living. So, you know, she's very lucky. And we're lucky to have you. Jared, before we get down to the business at hand, and to your point, we're going to talk about artificial intelligence, AI, and the economy because I know you've done a lot of work in that area.

but a little bit of work to do before we get there on inflation and growth. But before we even do that, could you just give us a sense of your path? How did you get to capital group? Right. So I actually came here through T-Roe Price. I was at T-Roe for about seven years-ish. I did a stint in between at Hartford, but I was at T-Roe Price. I came in from Tiro on the fixed income side.

So I was on the fixed income side, sat on the trading desk, responded to non-farm payroll reports on the first Friday of every month, all that fun stuff. And the opening came up at Capital Group. A friend of a friend actually told me about it, and hey, they need this economist. It's in Los Angeles. you want to interview. And I was like, yeah, I want to interview. And so I came for that. We went through like 20 rounds of interviews.

They mistakenly hired me anyway. And I've been here for about 10 years. And so the difference, though, between Tiro and Capital is that I was on the fixed income side at Tiro and on the equity side at Capital. And I can't even begin to tell you how different those mentalities are. Right. It's like night and day, right? It's doing what I call economic analysis where you worry about what's to the right of the decimal point on fixed income. In equities you're worried about just get me the sign right, pussler negative.

And then if you can get the magnitude even better to the left of the decimal point. So it's just a very different mindset. And it's really rewarding because something, I think like you, Mark, like just doing things differently, finding a purpose in terms of applying macro analysis. It's just fun when it works. Well, very cool. So let's, here we are.

This is Friday, September 12th. And this week we got some economic data, mostly with regard to the kind of the headlines were around inflation. And most importantly, the Consumer Price Index, CPI. And, hey, Matt, let me bring into the conversation. Can you give us a quick rundown on the CPI numbers? Yeah, so headline consumer price index came in at 0.4%. So that's the change from July to August. That was a little bit hotter than expectations. Looking at the components underneath, which we'll get to, I should say, it's the fastest monthly pace we've had since January. and the year-over-year change at 2.9% is the fastest since January as well.

So a bit of an overshoot, not, I think, anything terrifying, but certainly not the kind of surprise you want to see. Components underlying energy jumped 0.7% from July to August. That was mostly the explanation for our forecast miss. We were a little bit lower at 0.3, as I said. Gas prices still relatively low, but drifted up a bit in the month. The more interesting position or point there is what's happening with the electricity market, relatively small gain from July to August of 0.2%. But year over year,

electricity prices are, as measured by the BLS, 6.2% higher. That's largely attributed to the mass demand coming from data centers, which are powering the huge investments into AI capabilities. So I think that's something worth paying attention to and getting an increasing amount of scrutiny. Food prices accelerated. I think that's another concerning point in August's report, so 0.5% for the CPI food at home. We talk about a lot, generally considered a proxy for grocery store prices, grocery prices. That is a 0.6% rise. That's fastest we've seen since 2023. That's post-Russia invasion of Ukraine, where you see a lot of food commodity prices go up. A lot of food that we eat in the U.S. is imported from other places in the world. And if we track those

types of goods, those food products, it's a relatively unsurprising story. So coffee prices jumped majorly. That's one of the underlying factors. Meat, that comes from Brazil, other places in the world as well. Big jump again, 5.4% higher for the CPI for meats, poultry, and fish. And looking at core CPI, which again, accelerated from the month before, again, And the strongest pace we've seen since January, a 0.35% increase. It was rounded down to 0.3%. And the year-over-year rate stayed at 3.1%. So that's the high-level overview there.

Can I ask you, Matt, I know you construct a CPI for tariffed product. So, you know, obviously there's a lot of concern around the tariffs, the higher tariffs and the impact that's going to have an inflation and trying to gauge, you know, how big an impact that it is having and will have. and you've constructed a index that's based on all the various goods that are subject to tariffs to look at what's going on there in terms of price increases. What did that say in the month of August? It was a 0.6% increase from July to August. So faster than average. Faster than average. That sounds like a lot. Yeah. And it wasn't so bad in July.

Faster than average, I'm referring to the CPI basket. So we have everything that raised, that accelerated 0.4%, but if you just focus on the things that are import intensive, that's the average I'm referring to. And that, yeah, a 0.6% increase comes after mild, but still positive growth in July. But if we take a three-month snapshot, so we look at when, you know, terrorists have, to a varying degree, increasingly become the sense that they're here to stay over the past few months. I think as that's really settled in, what we've seen and what we think we're going to continue seeing is a real acceleration. So we look at the past three months on an annualized pace, those goods, those tariffs, import-intensive goods are rising over 5% at a year-over-year

pace. So certainly not good. And it's a strong acceleration from where those components, the- And just for context, correct me if I'm wrong, but prior to the tariffs, you know, we're talking about goods prices. They tend to, they're flat to down, right? So the 5% annualized positive is a pretty big swing here, you know, relative to where we were before. Am I getting that right? Absolutely. The comparisons that when we're thinking of superlatives

here, it's all back to the supply chain issues that we saw in 2022 as that recovery began in 2023. The period in between, 23, 25, you're right, goods prices are flat, certainly not driving any kind of inflation. That was a shelter story. Another service prices were keeping inflation elevated. Okay. So the conclusion would be tariffs are now showing up in the inflation data? Yeah, I think that's been clear and clear each month now for three months. One other explanation for higher inflation prices is restrictive immigration policy that it's been the deportation, the self-deportation. We've seen a very sharp pullback and even decline in labor force for foreign born. Obviously, that creates difficulties for those industries that rely on

immigrants, construction, agriculture, manufacturing, transportation, so forth and so on. Any evidence in the report, the CPI report, that that's, that labor market supply issue is showing up in higher prices? Do you get any sense of that? I think it's speculative to think, I mean, if you think of the agricultural sense, vegetable fruit prices have been rising throughout, particularly for vegetables. So that's an agricultural story. Those higher labor costs or just scarcity is going to push prices up if there's not people to process food in the way that we were generally accustomed to. The second story, which I don't think is conclusive yet, would be to look, as you mentioned, at construction and think about shelter prices, is the shortage of available homes

that kept shelter inflation high for so long? Is that going to be exacerbated because of the disruptions to labor supply in the construction industry as well as lumber prices being tariffed coming in from different parts of the world? It may look like that in August. We got a pretty strong increase in shelter inflation. The CPI for shelter jumped. I don't think that's the story. It's far slower moving than that.

I think it's more likely to be, even if we believe those intermediate concerns to be the case and to be inflationary, I think the much more persuasive case is that August increase was likely noise. And a lot of the month to month jump was hotel prices, which had been falling for a long time. And then in August, jumped. They're still down pretty big year over year. And so is shelter, but, or shelter is still drifting down.

But I wouldn't overstate that case. Yeah, I was kind of thinking personal services. I noticed there was a pretty large increase there. You know, more service or kind of on the front lines, you know, that might show up more quickly. Yeah, I don't think the shelter inflation has anything do with anything related to immigration at this point. That takes a long time to play out. Okay. Two more quick questions.

And I want to bring Jared in and get his take on the numbers and inflation more broadly. I've noticed that CPI inflation feels like it's around 3%-ish. The consumer expenditure deflator, inflation is measured by the PCE deflator, is also kind of around 3%-ish. And historically, there's a gap between the two. The CPI generally runs, you know, quarter point, half point, percentage point above the PCE deflator and had been for, you know, the last several years during the pandemic, clearly the case, it doesn't feel like, it feels like that gap has closed. Do I have that right?

And do you know what's going on? I know it goes to the composition, different, these two indices are different compositionally and there's a lot of different measurement, differences in measurement, but do you know what's going on? Do I have that right? Do you have you noticed that? I wouldn't, that wouldn't have caught my, you're right. I mean, if you're leading me into our forecast, too, by the way.

I'm ready for August PC forecast I have. Oh, yeah, I got to get that from you. Yeah, we'll get that next. Yeah. Yeah, so you're at about three-tenths of a percentage point, two-tenths of a percentage point between the core CPI, core PC and headline PC and CPI. I don't think I have a strong opinion outside of the compositional factors that you mentioned. Housing is bigger in the CPI, less than the PC, healthcare, vice versa.

But I don't think it's so big that it's warranted a ton of attention, but maybe it is. worth looking into. Okay. And you said, you now you have the PPI, the producer price index, which we didn't talk about, but we won't. And then the consumer price index, CPI, those two things are inputs into the PCE, the consumer expenditure deflator, which we're going to get in a couple weeks. What do you now think the increase in the PCE deflator will be in August? So we'll get a, our expectation is that we'll see a 0.2% increase in both the headline and core PCE deflater. So on the softer side, Just given the compositional stuff. So things that were not so inflationary in the PPI that feed through the PCE,

keep that a little bit lower. And similarly with CPI components. So for the PCE, with our forecasted change, we'll see a slight increase in the year-over-year rate from 2.6% to 2.7%. That's just given to 12-month comparisons looking back at relatively soft growth this time last year. And the core PCE will stay at 2.4%. 2.9% year over year. So 2 tenths of percent less than the core CPI. So certainly a narrower gap than we're used to. I haven't encountered a great explanation. Why or that it's noticed that. Right. Yeah. That's where we'll stand in. Just just the obvious point to make for folks.

The core PCE deflator, the 2.9% year over year is the kind of what the Fed Reserve is focused on when setting monetary policy and their 2% target. So we're at 2.9 and we're at 2.9, and we see, stay there. It's about a point above the Fed's 2% target. Yeah, exactly. All right. Jared, let me bring you in. What do you think about all these inflation numbers? How worried or nervous are you about any pickup of inflation here, any acceleration inflation here in the next six, 12 months? I think Matt did a great job of going through the report, highlighting the important bits. in terms of, you know, overview and like how we're, how I'm thinking about it, you know, we're in the camp that, you know, tariffs will cause prices to rise and, you know, one-off or not one-off

in this whole debate, you know, that's happening. We're going to get three percent inflation. We're already close to that now. My base case has been, you know, three and some change heading into year-end. If you look at some of the underlying data in terms of the posted tariff rate and what's actually coming in, we're not quite there yet. It takes time for these tariffs to move through the economy. And so we're going to still see this inflationary pressure, I think, in the reports ahead. Now, am I worried that we're going to get some type of 1970s hyperinflation or something like

that. No, but high inflation after already high inflation for the last five years where the mandate has been breached for multiple years now does create some issues in terms of where, you know, consumers' heads are at. It does it, it does increase this risk that inflation expectations could become higher, that we have more persistent inflation. Now, you know, as well as I do, Mark, that pre-COVID and for the decade prior, we struggled to reach the 2% target from below for decades, right? And we're like, oh, we'll never get there

and this is a loss cause and just throw in the towel. And now we're gonna struggle to get down to the target, I think, for quite a while. That's really what I have in my base case, that I have even beyond 2026 and into 2027, that we don't really breach below the 2% target, but we're in this 2 to 2.5% regime for a while. That's, you know, that's not awful.

It's higher than it was. If you take 1.5 to 2.5, that's important. That's pressure on individuals. But I think, you know, we'll have to, in terms of the near term, I do think we're in this deep mini cycle, what we've called it here at Capital, deep mini cycle that is going to slow growth because of the tariff pressure that's going to slow consumer spending. We have business fixed investment, CAPEX, that type of stuff slowing down as well from the uncertainty, policy uncertainty.

So we do think we're going into this slower rough patch, let's call it. And then 2026 looks a bit better. But again, inflation is not coming down quickly. we'll see what other policies are a past or what they look like in 2026. But we don't see this huge, this huge rebound in 26. So it's a pretty muted forecast for now. We're worried about inflation, but not hyperinflation. Just to make sure I have what you said right, or at least one part of it,

did you say we're kind of inflation as measured by the consumer expenditure inflator is going to kind of settle in around 2.5%? for an extended period, as opposed to going all the way back down to the 2% target? Right, between 2 to 2.5. Oh, between 2, 2 and 1. Yeah. So the Fed Reserve is going to allow that to occur. I mean, presumably if the Fed wanted two, they could get two, but they're not going to do that.

I think, so the implicit answer is yes, they will. They'll allow it to occur. Just as they'll, you know, a 1.5 wasn't the end of the world. Right. pre-COVID, two and a half is not the end of the world post-COVID. And how much pain do you want to exert to get that last half a percent? And we all know that if you're printing two and a half percent, the standard errors are enough that maybe you are printing to.

Who knows? Right. Good point. Yeah, yeah, right, right. Good point. And I guess you could also, I don't know how you feel about this, but there's been, in my view, kind of a reasonable argument. that 2% may not be the right target.

You know, maybe it should be a little bit higher given that underlying growth rates are at least up till now relatively slow. And more likely that you'll hit the lower, zero lower bound if you don't have a somewhat higher inflation target. So maybe de facto we're getting there, you know, one way or the other. Yeah. Yeah. Yeah.

I mean, that's irony, right? Because there is struggling to get to 2%. And now, right, you can't. get it below 2%. I've always thought, you know, as I think about, you know, we did the strategic review and we've done another one and, you know, we're always kind of fighting the last battle in a way. And I don't know, I would be curious what you think, Mark, but why not go to a 2% and just say plus or minus 1, right, kind of a target with a range, right? And, you know, we're not going to react violently to, you know, two and a half, but we're not going to react

violently to one and a half to give some kind of uncertainty leeway to it versus saying, we're going to hit two because we know we know we can hit two and we're going to do everything we can to get to. I'm just, I'm kind of playing with that, toying with that idea. If that's a more reasonable way to approach it given the uncertainty that we face ahead. What do you think? Yeah. I mean, I kind of, I thought we kind of de facto got there when the Fed did the inflation averaging. You know, that came out of the last review that they did back in 2020. If we're below that 2% target for an extended period of time, then it's okay to be above target for an extended period of time. And through the cycle, you might get two. So there's, it's not a hard and fast

2% at every point in time. It may be 2% over a period of time. And that felt flexible enough to me. It just just feels like 2%'s too low in the context of potential growth, real potential growth that, you know, might be around 2% or maybe even lower, although that may change, you know, given AI, and we're going to come back to that and what that means for productivity growth. So, you know, maybe this all becomes mute in the not too distant future. You did say, so just to make a concrete in my own mind, I know you said it, but I want to repeat it, you're thinking that inflation, let's say it's kind of in the around 3%ish, will migrate a bit higher here into the 3s, kind of low, mid-3.

over the course of the next six, 12 months because of the tariffs. I'm putting words in your mouth, but I'll just put them immigration policy. It might be pulling a bit of a role. But it's not going to be persistent. It's critical to watch inflation expectations to make sure that they don't increase. Certainly bond market inflation expectations. So far that hasn't happened. Therefore, this inflation should moderate back and ultimately settle in somewhere in the low twos, you know, two and a quarter, two and a half.

And that's kind of sort of will stay, you know, for the foreseeable future. Is that kind of got that right, roughly right? That's accurate. See, I'm a good student. I'm even better than my daughter. Don't you think? Don't you think I like, could my daughter have done that? I'm just asking.

I'm just that. I was a sure. You could be a capital group economist, Mark. Damn. Really? I'm telling you, she's got the greatest job on the planet. Okay. Let's turn to, you mentioned this already, but just to flesh it out a little bit more is growth.

And your sense is that the economy is going to be, it's not going to be a recession. Again, I'm putting words in your mouth to see how it tastes, but no recession, but kind of a sluggish economy here, you know, as it tries to digest the higher tariffs. I think that's exactly right. Not a recession, but what I'll call a mini-cycle. And it's for all the reasons you said. Tariffs slowing capital expenditures into this year. We think prices do have an impact.

And so we've seen these cycles before. You know, we've seen many cycles in the past. 2015 is a good example. A prior one, 1990s, if you want it, 1998 is a good example. that, at least in our empirical work, that if you look at these mini-cycles relative to recessions, you know, the big difference is that, you know, these mini-cycles, they feel like recessions when you're going into them, and the service sector of the economy basically doesn't buckle. That is usually manufacturing-centric, good-centric parts of the economy are the ones experiencing the pain,

and kind of like what we're seeing now. And so the service sector economy doesn't buckle. The key tenant of that is the labor market. And I think you and I would both agree that monitoring the labor market is important, and that's what's keeping you from going into recession. But that's exactly right. Now, I wouldn't go to the mat and say recession is not a possibility right here. It's certainly a possibility.

I worry about it. But for right now, I'm not in the camp that we'll see some resilience. It's going to feel painful. It's going to last for probably the next three to six months as we get into the new year. But eventually, we're going to turn the boat a little bit, but it's not going to be a V shape. It's going to be more, you know, go back to the letters and all that stuff. You know, what shape are we in? It's going to be more.

So is there an indicator two or three that you're looking at to gauge? whether that forecast you just gave is correct or whether something, you know, darker is going to happen here? Is there some leading indicator that you, you know, put more faith into than others? I have my magic eight ball like you, Mark. I just shake it up. Today is a promising day. Yeah. The, you know, I think I always go back to, well, I'll give you two. The first indicator is, you know, one of the great things about working at capital is that our analysts are talking to corporations every day around the world in real time globally, right? It's just happening every day. We have a research platform that we publish all that work and, you know, what did such and such say about, you know,

the semiconductor cycle, what did this person say about, you know, electricity. So it's just a rich platform that we're getting these real-time insights from management teams around the world. So I read that every day. First thing I do is open it and read what's happening, what people are writing about. And when I read over the last six, eight weeks, or even, you know, throughout the cycle here, we're not getting the signals that you would expect if management teams were worried about recessions. We're not getting, yes, sir, they're concerned. They're concerned with tariffs. They're concerned with supply chains. They're concerned with geopolitics. They're concerned. There's definitely concern. But there's not, hey, we really need to pull back 15, 20 percent

on hiring. We need to pull back on CAPEX. So it's that internal research platform that I put a high value on and I read every day. And then the second, in terms of economic indicators, because I know you don't have access and our folks don't have access to it. I was going to ask if I can get on your mailing list, but I assume that's not possible. Yeah. You can become a capital group client, Mark Zandi, if you invest in the American funds. Oh, that's good to know.

That's a good point. See how he sells? He's a great salesman too. Wow. Jeez. Very smooth. So the more for the more tangible one, it's got to be the non-farm payroll report. You know, that's the one.

If you put me on a desert island and you ask me, what report do you want to monitor the U.S. economy every month? It would be the non-farm payroll report. And it's just so granular, so rich with data, even with revisions that, you know, it's got to be that. one so it's like what you've pointed out the trend in labor force growth the trend in job growth where we are with the unemployment rate and so I I do focus on that what what industries are slowing which are increasing what's the

diffusion how many jobs are you know which industries have actual job gains and how many of them are there so yeah it's it's the payroll report and we you know fortunately at here at Capital we have many economists and we debate this all the time and one of my counterparts, Daryl Spence, is more bearish than me. And so we just constantly debate this US outlook. And, you know, depending on what you have in terms of where are we with this break-even unemployment rate, is it, you know, what's the break-even job growth so that the unemployment rate doesn't go up?

Is that $25,000 a month or is it, you know, $50,000 a month? Yeah, we're right there. at 20, we're right there. And so I, yeah, I'm, I can't say, I can't look at last month's payroll report and be, and say that, oh, everything's fine. It's not fine. And we are, uh, I would just say that I'm probably less pessimistic than others. I wouldn't characterize it as optimistic. It's just less pessimism, uh, in the near term. Can I, and this, this is a question I'm grappling with, so I don't know how to, it's, I don't have an answer. I'm just saying, I'm asking, suppose break even job growth, and as you point out, that's the rate of job growth consistent with stable

unemployment, went negative. I mean, because of the immigrant, we're seeing a large mass exodus of immigrants from the country. You know, instead of getting net positive or net negative, so therefore we're now minus 25K, minus 30K per month. And therefore, the economy has to only generate minus, that means we're going to lose jobs. Isn't that a risk? What is that? Is that a recession or how do you think about that? I mean, have you, or have, that's bothering me.

I can't get my mind around that. It's a possibility. I like the framing. It's almost like the, the, the, the NERP version of payrolls, right? Because when you're in the NERP, the negative interest rate policy world, right? The NERP world, everything turns on its head, right? So, the less negative is then good, like you're saying. You know, I haven't, I mean, I haven't put pen to paper, but like what you're saying,

and I would say generally, though, like, this world that we've been in for the last five years has been topsy-turvy, left is right, right is left, up is down, you know, this, just strange things happening in the economic data, at least that's how I've perceived them. Relationships that should hold aren't holding, relationships that never held are holding now. And so it wouldn't surprise me that you could get a temporary. So a temporary world like that where, yeah, maybe break-even payrolls is like negative 25, and if you're flat, you're actually okay. It's fascinating.

It's a good question. It's a fascinating thing. I wouldn't say it's a good thing, right? I wouldn't say it's a healthy part of the labor market. But does it create a situation where you avoid a recession? And for us, you know, I just go back to like NBER, how do they define a recession? One of the principal components is like, you know, the increase in the unemployment rate. In that world, you would have negative job growth but no increase in the unemployment rate.

Would they characterize that as a recession if other things like retail sales are doing okay? if overall income growth is doing, okay, would they characterize it? I think it would be, they might get a 50-50 split on that one in the Supreme Court that judges recessions. Right. Right. Yeah, I mean, it's pretty mind-numbing. Chris, I'm going to bring you in quickly.

And then I want to get to artificial intelligence because talk about mind-numbing. There's a lot of mind-numbing aspects to that, too. I mean, you can envision a scenario where we're losing jobs because of the productivity gains contained by AI, but that we're getting GDP growth, we're getting retail sales growth, we're getting industrial production growth. So again, back into that weird world that we're talking about. But let's, before we go there, let me let you weigh in, Chris. Yeah, I've been grappling with this myself. And the other element I've kind of been throwing in here trying to figure out is just with all the retirees. Right. We have these structural demographic changes as well.

So that alone, right, would put downward pressure on job growth. You could certainly make a case. you know, with the decline for their declines in immigration that you go negative. And what does that mean? Not only for the, let's say the labor market, but then let's look at housing or let's look at other parts. What's it going to look like when permits, housing permits are going now, not because, you know, there's anything particularly wrong. It's just structurally we don't need as many homes to be built.

It's just a very different world we're going to be entering into. And, you know, how do we change our frame of reference here? You know, there's so many things to talk to you about, Jerry. I mean, you know, and I only have you for another less than a half hour and I got to get to AI, but I got one other question before we get there. I think I got so many interest rates. So what does this all mean for the Fed and for particularly the question I always get is the 10-year Treasury or long-term interest rates? Because there's a lot of stuff going on there. So do you have a view on that where the Fed's headed and where the tenure is going?

So coming into the year, kind of let's call it like this time last year, I had this year, you know, so from here, from where we are now, 50 basis points, reduction in Fed funds rates because of the mini-cycle view, because we are going to go into a rough patch. Historically, the Fed has reacted to mini-cycles, and so they would react again now. And then I have, I had 50 next year. And what I'm grappling with, to your point, too, Mark, is that, is that enough? Because the market is saying, you know, another 50 next year is not enough. And in addition, you know, it's starting to get more front loaded into 2026.

So should I change that? Right now, I'm, you know, I'm a little reluctant to do it. For a couple reasons, one, I think productivity growth is going to surprise us. I don't think we're going to go into recession. I think some of the O, triple B, the bill, the big beautiful bill, will have some upside into early 2026. So I'm not ready to do it. I'm wondering about it. On the 10-year treasury, though, I have been flat wrong.

So you should ignore everything I say about it. So, and I've been wrong in the way that I've, I've thought a 10-year treasury given the fiscal, given the bill that was passed, given where we are with fiscal policy in general, that we would be higher, that we'd be closer to, you know, four and four and three quarters or even up to five by the end of this year. And we're not. And so I've been wrong about that. I've got to think about, you know, how to, you know, either throw in the towel.

Do I throw in the towel on it or do I try to double down? But for now, so I would just take maybe my front end view more with more weight than my 10-year view. Got it, got it. Well, I don't think I'd be as honest as you, their judge. Definitely Chris would not be. Chris would not. Yeah. No, no, I'm only kidding.

Chris, Chris wears his errors on his sleep. So, yeah, unlike me. Okay, so let's talk about artificial intelligence AI. This is an area I know that you've been doing a lot of work in. And maybe I'll just let you, maybe this is a very open-ended question. Kind of my simplistic way of thinking about this in the context of the macro economy is there's two very different views of the world. You talk to technologists, folks in the AI world, you know, sitting, I was having dinner with a bunch of banking clients and there was the CEO of a major AI company. And these guys are so, well, from my prison, dystopic with regard to what it means for the labor market, meaning we're going to get a boatload of productivity growth.

And this is going to wipe out lots of jobs, so many jobs that, you know, we're not going to be able to replace them fast enough and we're going to have high. higher higher unemployment. Labor supply is not the issue. Not enough labor supply is not the issue. There's going to be a lot of labor supply. And then you've got the other side of this that feels like it's coming on now more than it has up till now. It's taking the opposite perspective that maybe this thing's a bust, you know, or it's certainly maybe it's more like business as usual productivity growth. It's not like this game-changing kind of technology that's going to upend, you know, all kinds of business practices going forward. Is that, is that a fair way of, does that, do you sense that kind of very bipolar kind of view

of AI and, and if that's right, you know, where do you land in that bipolar world? Or maybe I've got the frame wrong. No, I think in terms of the, I think the key debate and the key argument, I think that's Exactly right. You have a lot of folks, and we have this debate routinely internally here at Capital is, you know, the downside is that this is all a bust. The idea that this is any type of thinking machine is absolutely crazy. This thing doesn't think. And we've just, if you look at some of the hyperscalers and their investment plans for 2025 even, and into 2026 are massive, right? I mean, the money, I did a calculation market, I think I brought it to one of our nerdy economic conferences

that we do every two or three, two or three times a year. But I calculated if these plans are correct, then every incremental dollar into US and CAPX will be AI. Like every incremental dollar will, if these estimates are correct, which means that non-AI is just getting crowded out, right? Any non-AI related CAPEX is getting crowded out. And so if you layer on top of,

hey, this is a, this is a useless transformer architecture that mimics but doesn't actually help us and all the CAPEX spend that is going into it and the valuations, as you pointed, as you kind of alluded to, the valuations of these companies are so high, that outlines the bear case pretty clearly. The upside, I always think about the upside as like,

the closer you get to Silicon Valley, the more you hear the dystopian, you know, which is, I'm saying upside, I guess, yeah, it's upside in the sense of like we're going to get these massive productivity gains, but eventually you're going to get this like job loss because, you know, it's just so good. at doing the jobs that we do. Not your job, Mark, maybe my job.

No, I'm the first to go, I assure you. I'm definitely the first to go. And I, you know, in terms of those bookends, right, you know, I have skepticism of both of those of views. And the analogy I like to use is that for the downside, this is crazy, it'll never work. folks. You know, there was a time in history where someone looked up at a bird and said,

maybe we can build something like a bird and we can fly, right? And it happened, right? We had that happened and the Wright brothers created it. And people will say, well, it's not a bird. It's not exactly like a bird. Therefore, it's useless. And, but, yeah, this is not exactly like a human. It doesn't need to be human intelligence.

I think the bar is way too high to compare it to like it has to be human intelligence. I think it can be very useful and airplanes are very useful because they bring people to and fro. This can be very useful for a lot of the work we do. So even if it doesn't live up to the hype of like intelligence and super intelligence and AGI and all that stuff, I think even the lower bar of helping us do work differently, I think is, still pretty important and will be productivity enhancing. I've changed my workflow over the last year versus what I did based on this. And so I think others are doing the same. And then the other side of, you know, the silicon. Can I just got me there one second, Jared? Can you flesh that out

just a little bit? I'm really curious if you feel comfortable. What about your workflow has shifted, changed? Oh, yeah, yeah. So what I'll do, So what I'll do a lot, so I like to play with it. So we've been working with AI. You know, I wrote my first research note at Capital on AI in like 2015. It was actually a rebuttal to Larry Summers and Bob Gordon. How did that go? How did that go?

It was only internal. He didn't read it. Oh, he didn't read it. Okay. Thank goodness. But he was talking about secular stagnation. and, you know, I will never grow again and, you know, all these things. And I was like, I don't know if that's necessarily true.

And here's this AI stuff that, you know, kind of nerdy people are doing. Anyway, but how have my, so how's the workflow change? So what I'll do is, so we have co-pilot internally. We also have access to chat GPT and some other models. And so I think, like, one of the things is just, I'm doing less time doing work of like looking for stuff, looking for data points,

looking for things. I can get them much faster. I think, you know, we do so much work, we've been doing it for a long time. You kind of forget, you know,

things from the past a bit and just, okay, what happened in blah, blah, blah, and you type it in. I'll have four or five different versions of co-pilot open

and just be like, you know, doing, they're doing deep research for me, kind of asking question, like, yeah, you better get to work, you know,

why haven't you finished yet, right? And there'll be chugging along on my questions and I'll be working on something else for, you know, an internal client or, you know, maybe for a presentation or something. And I'll be doing that kind of, and it's doing its thing in the background. And that's a regular part of my workflow now.

It doesn't do data and analytics and charting and all that stuff very well. You know, it actually does a pretty bad job. But I think eventually, you know, I think in a year from now, when we're talking on the, on the podcast, and I'm absolutely wrong. And we've gone into a very deep recession and dystopia has occurred. That I think it'll have gotten to a point where we're seeing more kind of charting and things like that. So I've changed my workflow as an economist. We've changed things too internally in terms of how we do things.

Yeah, it's different. Have you changed, have you used it, Mark? Yeah, yeah. I mean, I, I, similar to what you're doing. It helps, for me, it helps with the blank page problem, you know, just getting started. Because I think what I get back from AI isn't, I don't feel comfortable with. I mean, it's far from, I'm comfortable with it. But it's a good, it's a start, right?

Right. And it gets you going. And I find that for me, the hardest part of writing and doing research is actually just writing the first word. So something can write the first word for me. What I find kind of mind numbing, though, is like I'll say, hey, can you, because I'm always writing something. Can you take what I just wrote and elaborate and expand and please do it in Mark Zandi's voice? because there's a lot of stuff out there that I've published. And so there's a long history for 35 years of stuff I've written. It's got a lot of content that it can use.

And then I'll say, you know, you can bring in other sources, but please annotate. I want to know exactly where this is coming from so that I can go back and take a look. And I find that very useful, you know, very, very helpful. And then for a certain complex statistical analysis, you know, reading academic papers is hard. It takes a lot of time, a lot of energy. And it's very helpful in distilling down, you know, the kind of the core of what's being said. And then helping in a very intuitive way, you know, what are the statistical approaches and

techniques that are being used here, that kind of thing. So I do, and we're using it for our modeling, you know, right now we haven't unveiled this yet, but we've got this large scale model of the global economy. And we have different agents that can do different things for you, you know, like decompose the change in the forecast into what caused the change or take a look at this equation, tell me exactly what's driving it and what's the most sensitive. That, you know, that kind of stuff. Or I want to determine the impact of higher tariffs on autos from Japan. How do I shock the model to be able to do that? You know, these are all, if I had to figure that out, you know, that would be very

difficult to take a long time. I could figure it out. It take a long time. But this gives it to you. And it's not always right. But it, again, it goes back. to the blank page problem, it gives you a real start. And you go, oh, okay, you know, I didn't think of that. So there's a lot of, there's a lot of value in that. But I don't know, Jared, and I stopped you because you have to go back to the other dystopic thing. But the one thing I don't get is, you know, AI, everyone's using AI, everybody, you know, all our competitors, everyone's using AI. So all it does is raise the bar. It's not like I, I need, I still need, you know, really good economists working on this.

The level of the work that they're doing is just higher, but I still need them. Their skill sets may be different, and the types of work they're doing different, but it's not like I'm going to reduce my head count as a result. In fact, I might even have to increase my head count, you know, because I'm in this pitch battle over AI. So anyway, I digress. You shouldn't have asked me that question. I can go on, but let me turn it back to you.

That's a good segue, actually, because I think that gets you this. like why this like mass job market layoff thing, in my view is a lower probability is because, right? We are gonna change the type of work we do and we actually might need more people to, so if you're successful, let's say Moody's is successful at implementing AI, you're gonna do things better, faster, cheaper than your competitors,

you're gonna have more and more clients, clients, you're going to have more and more demand for your services. And so you're going to have to have more economists, right? And so I'm skeptical of this, you know, total dystopian view that folks have laid out. But I will take a sliver. I'll take a sliver of where I do think there is danger for certain roles, right? So if you have, so I've kind of, I've created this kind of silly criteria, and I talk about it internally, that if your role has a very large context window that's well-defined, there's lots of research on it, right? Okay, that's, you might be susceptible to AI. If your role is a very standardized input and output, right, that I'm just taking X and I'm

creating Y, you might be susceptible to AI. If you're getting paid for one and two at a very higher rate than you are susceptible to AI because AI can do those first two things pretty well. If you've got a pretty standard language and kind of of what you do and its input and output very standardized and you're getting paid a million dollars to do it, why would a corporation continue doing, continue using that role in that way? So the and that, and that, I mean, that's an extreme example, but hopefully it highlights. There are components of the labor market. And I think I presented this to CBE, Mark, one of our meetings that we go, we attend together.

If you look at portions of the labor market, one area that I think, and Ben Yolson has written a paper about this, and I think Microsoft Labs have written some papers, kind of where is it? where is the labor market, where is it happening? Tech, if you're a coder and you're in tech and there's that coding, input, output, very standardized coding that you're doing it and you're getting paid a lot, it's logical that you would see

some disruption to that cohort of labor. I think you can see it in the official labor market statistics already. Ben Yolson wrote the good paper about, you know, younger software development. and call center folks are getting disrupted relative to older cohorts. But I think you can start seeing it in the data. But if you took that cohort of individuals of the labor market,

it's relatively small, relative to the 100 and whatever 70 million people we have in the labor market. You know, it's a pretty small cohort. But there will be some cohorts that get disrupted. I think to your point earlier, what do we do instead? I think we're going to be doing more value-added work on behalf of our companies and for our clients. I'm going to turn the conversation over to Chris in just a minute, but I want to try something on just to get your view. So we, you know, when we produce a, we're producing forecasts.

It's an explicit forecast, you know, that we put in a database that clients use. And so we have to have a quantitative view of what AI means for productivity. growth, GDP growth for jobs and everything else. So let me just lay out the framework we have now. I'm just curious when you think it's a reasonable one or not. So without AI of significant consequence, there's always been AI. I mean, I used a neural net back 35 years ago to figure out whether a bank was discriminating and its mortgage lending. So that's a form of machine learning, a form of AI, not LLM, but, you know, so it's been around a long time. But so let's assume a base level of that kind of, of technological innovation. I, our forecast for non-farm business productivity

over the next five years per annum is say 1.7% per annum, something like that. If I put AI into the mix, that would get us to roughly 2.2% you know, per annum. something like that, non-farm business productivity, which just to complete this and translate it into GDP, you know, you have to overall productivity growth, including government, would be closer to two. Labor force growth would be, you know, maybe a quarter point per annum. So that gets you to potential growth, GDP growth of two and a quarter percent per annum through the end of the decade. What do you think? Does that sound, is that kind of in the ballpark where your mind is, or is it, you think it's bigger than that or smaller than that?

Well, I'm shocked because you said you used a neural network, Mark. I can't believe it back in the day. So that's amazing. 1991. You're ahead of everyone. I mean, you should be, you know, the AI evangelist for us all. People were using neural nets back then. It wasn't, I mean, they're not the today's neural nets.

They're 1991 neural nets. But it was actually quite useful because it actually. uncovered interaction interaction interactive i'm not going to go into detail but it was helpful you know and uh in in in that project but anyway uh i digress no no that's uh that's useful framing and i agree with you um even when you go back to the history of ai we've had these AI we've had this these hype cycles before even the 2014 2015 um when the machine you know deep learning started coming out. We had a similar hype cycle and it was a similar argument that all of our jobs are going to get, you know, are going to be, we're going to lose all our jobs

because of AI. Didn't happen. But in terms of like where I am with productivity, I would, I'd probably take the over on yours, Mark. I'd probably take the over. Not, yeah, and I'm, I struggle with this because before, you know, I've kind of been the internal AI rah-rah person, like, hey, this is going to be big and, you know, and people are tired of listening to me and all that stuff. And they kind of roll their eyes when I say productivity. But what I, I, so I think it could be bigger. And I'll give you a couple of reasons. I would probably be like two and a half to three, something in that range. And it's feasible. we can get there.

If you look at... Getting a distance to be concrete, you're two and a half to three on what, non-farm business productivity? Okay, God. Overall productivity for the U.S. economy. Yeah, I think we could be there. And one of the things we've done internally

and have taken some... You know, Claude did a research paper, one of the AI companies. They did a good research paper. Microsoft Labs, I mentioned it before. Ben Yolson's paper, Stanford. If you take these papers and you start going through, you know,

where is the productivity increase coming from, and then mapping those to like labor market cohorts? And it's, you know, this is not, you know, it's a little hand-wavy, it's hard to do, and just to get directional impact. You know, some of these tasks, you're getting like 30% jumps in productivity, right? Big, I'm not every task, but some of these tasks, you're getting big jumps in productivity.

And if you start mapping those to labor market sectors, you can you start getting relatively big numbers because a lot of what white collar work does is like you know writing and you know doing things like like you said the the blank page problem right marketing translating into different languages you know it really accelerates some of those tasks and so i think the ultra bullish folks i think take it a little too far where they say like almost every, you know, 50% of labor market tasks have some, you know, 50% of AI exposure, that they're getting like 4% or 5% productivity growth. But I think, you know, 2.5 to 3 is actually pretty reasonable.

And I kind of base that on just the empirical work that we're getting from actual usage of, and the way they designed these experiments are pretty cool because they try to map like what, given a particular task, how much productivity are you getting from it? And so I think it's just been helpful to start triangulating on what those numbers look like. So I would take the over a bit, Mark. And what does that mean for jobs, do you, Jared? Does it mean anything for jobs? I mean, I'm not, I am not in the dystopian camp that we're going to lose a bunch of jobs. And, I mean, the history of technological change is, you know, on that side. I, I think, think it's actually, you know, we might lose some cohorts of jobs and perhaps the unemployment rate goes up, you know, 10, 20 basis points because, you know, we lose, you know, I don't know, 200,000, you know, cohorts that are exposed to AI in the near term.

Right. But I think overall, you know, I look out in the out years, I have an unemployment rate that's, you know, four-ish roughly as a rough equilibrium. And maybe the, you know, Maybe AI actually enhances job search and matching in a way that we haven't been able to do in the past. And so maybe the U-STAR, the equilibrium unemployment rate, it actually goes down, you know, or something. So there's just a lot of – Yeah. Interesting. Hey, Chris, we're getting along in the tooth, and I know it dominated the conversation. But let me – anything you'd like to ask, Jared, or anything you want to push on here?

Sure. Maybe just continue on this last point here. as you're thinking about productivity, and so you've got this optimistic forecast, is that completely driven by the AI world? Or it seems like AI has now taken all the oxygen out. We were thinking about a lot of other productivity gains before AI came on the scene that would come on, you know, green technology, synthetic biology, even the GLP, the weight loss drugs make people healthier. Have we lost sight of all those other kind of innovations in the background? Or is that kind of baked into your

forecast here. It's a good point. You know, it's so dominated everything. So my, so going back 10 years ago, my more optimistic productivity view has been basically on an AI view that

I think AI is going to surprise and I think it's going to be stronger than we think and for, you know, everything we've discussed. The, you know, I'm trying to think of a big productivity enhanced. like a GLP1 drug makes us healthier and so we're

more productive at work type thing Chris That's right And maybe not the scale of AI but certainly contributing So does that factor into your More optimistic You know we've

We'd You know I wouldn't say it's a It's a marginal driver The other one I think of a lot Is about What was the operation? Did we do it with Operation Warps

that we created the vaccine out of back in 2020. I can't remember the name of the... Amel Warp Speed? Was it Warp Speed? Is it Warp Speed? I forgot. Oh, yeah, Project Workspeed, right. Oh, I think...

So if you look at that process, right, where you took a... Usually, it takes like 10 years to develop a drug, right? And we did it in like 18 months or something, right? Massively productivity enhancing, right? Relative to where we were prior. But those... types of are there kind of structural things that we can do in drug discovery for example that I think about but I haven't Chris I haven't like integrated into

like a numerical forecast in terms of the contribution to productivity but I got to think there's a lot to be said about that model as being a way to develop drugs in the future but also you know you get into what is AI good at what is machine learning and deep learning good at is creating you know taking large, massive amounts of data sets and finding relationships that humans take longer to do, they'll probably eventually do it, but can do it much quicker. Can it accelerate this discovery of different types of drugs? I think the answer is yes. Our analysts at least agree on that, that there is something there. So I think that's also, you know, healthier lives,

more productive lives are also a component of it. But I just haven't marked a market that view. You know, the one thing I do worry about in terms of, and let me say the standard errors around these projections, I'm totally with you. Like, I don't know. It's all over the place. But the one thing I worry about on the downside is, and I can just feel this from my own experience, and this goes to a point you're making about crowding out, you're taking all these resources away from what I call businesses usual productivity enhancing activities. you know, things that I know with a reasonably high probability are going to win, they're going to, they're going to give me a single, you know, maybe a double. They're not going to give me the home run, but I'm going to get productivity gains.

But if I'm not investing in those things in putting those resources over here to hit the home run in AI, I better hit that home run. Otherwise, I got a problem. I got a problem, right? No? That's a good point. Yeah. Yeah.

And there, you know, you see this, there's going to be disillusionment, right? And we saw this in like 2015, 2016, where, you know, deep learning was going to be the big thing and it fizzled a bit. But for the most part, enterprises aren't ready. It's not like you just bring in an LLM and you just press a button and then, you know, magic works, right? Like, it just doesn't work like that. That there is a lot of data infrastructure. There's a lot of context learning. There's a lot of just really, grunt work, right? Trench warfare, right, that you got it to do it, to do it correctly. So I do think there's going to be disillusionment. And I think there was some kind of stuff circulating about like, you know, 85% of like corporate America is disappointed with their pilots of AI or something like that. I think it's probably overstated. But there, it is, it's really hard to do. And if you, like you said, Mark, if you're putting all your resources to like hitting this home run when there's a lot of singles out there, You know, I think that's also dangerous as well. And so, yeah, I do resonate with that.

You know, you got to get a, I think we can walk and chew gum, you know, do this normal stuff in business and kind of experiment on the edges. You know, there's time to do this. We're still in the early endings. And it's, I don't, just preparing the business for it is, I think, the most important thing. Well, this is a wonderful conversation. I really appreciate it, Jared. And there's, as I said, earlier, there's just so many questions I have for you. So maybe we can get you back on. And I'm not sure whether I'm hoping you're right about AI and productivity growth or not. I have to think about that because there's a lot of other questions about like who wins in that world and who loses in that world. I mean, and what do we do about that? But anyway, I digress, I've been digressing a lot. But I want to thank you for taking the time. Really do appreciate it.

Hopefully we'll have you back if you'll have us. Of course. And we'll have you back into a capital group as well, Mark. I always appreciate your views. And I'll be equally as wrong. The last time I was back to the group. But thank you so much, Jared. And we'll talk to you soon.

And with that, dear listener, we're going to call this a podcast. I hope you enjoyed it. And we'll talk to you next week. Take care now.