A Tale of Two AI Futures - Citrini vs Citadel

Live from Schenectady, New York, it's the Metrics Brothers, Growth and CAC.

And I'm Growth, better known as Ray Reich, founder and CEO of Benchmarkit.

And I'm CAC, better known as Dave Kellogg, EIR at Balderton Capital, independent consultant, and the author of Kellogg.

And together we are the Metrics Brothers.

And we go together like SaaS and AI.

Dave, I could take that in 2 very different directions, but I am an eternal optimist. So hopefully this is not another episode about the SaaSpocalypse.

Another? I mean, Ray, we mentioned it. Okay, we did cover it fairly extensively in our episode on Avenir's Future of SaaS report, but we have not dedicated, dedicated an episode to the SaaSpocalypse.

I know, but it's so negative. And we did spend at least half the Avenir episode on the SaaSpocalypse.

Well, it is the biggest thing going on in SaaS right now, Ray. In fact, who was it, Crunchbase the other day that had an article out called Where Are All the SaaS IPOs? And it was like, bye-bye-bye, you guys aren't keeping up with your reading. And it may not all be good, Ray, but you're giving me kind of very strong other than that, Mrs. Lincoln, how did you enjoy the play vibes here?

Well, based upon that, I suppose you love the Citrini report, all dark and scary. It's even got ghosts with their ghost GDP, even though I actually preferred the Citadel response because to me it was a little more positive and uplifting and calmed some of the fears.

Well, alas, my metrics brother, let's look at this tale of 2 possible AI futures in today's episode.

It sounds like an epic discussion and one I'm sure we're always going to agree or not.

Okay, Ray, before we go red pill, blue pill here, let's hear a word from our sponsor.

The Metrics Brothers is proudly sponsored by Benchmarkit, the B2B AI and software industry's largest, broadest, and most contextualized benchmarking index. Dave Kellogg and Ray Groth Wright have put the Metrics Brothers podcast on hiatus for the remainder of 2026. All episodes we have ever recorded are still available on your favorite podcasting app. You can reach Dave at Kellogg and Ray on LinkedIn at Ray Reich. Thank you so much for listening to the Metrics Brothers podcast. Okay, Dave, so let me just give a quick high-level overview because we're going to take almost point-counterpoint using the Citrini Research Article that was published on February 22nd, entitled "The 2028 Global Intelligence Crisis," and then two days later Citadel came out with their 2026 Global Intelligence Crisis. And bottom line, it said Petrini perspective was a speculative future history, almost like a simulation where rapid AI agent adoption triggers this global intelligence crisis. They say things like. It ends up in June 2028 with 10.2% unemployment. The S&P is down 38% since the October highs. And by the way, I did a little semantic analysis of the article. It had a 3.4 to 1 ratio of negative crisis words to positive ones. And that was also part of Exponential View article on this. So Citadel comes back with this data-driven defense where Hey, unemployment's only at 4.28% today. AI CapEx is 2% of GDP, and that AI-generated commodities pricing is up 65%. So they argue that AI is a traditional productivity-enhancing supply shock that will generate real additional incomes and will generate new industries rather than rendering human labor obsolete. So That's the kind of macro summary of the two.

Okay. So, so before we start talking about what we think on this, I'm not sure it's all going to be point-counterpoint. We need to do our financial disclaimers. So one, we're talking about the stock market again, because Citrini is a stock market research firm and Citadel is a large asset manager. So it is the hot topic of the day. So that's why we're talking about it. But we are not financial advisors. We don't give financial advice. And if we did, you shouldn't take it. So we are software operators, but it's just hard not to look out the window and see the SaaS-pocalypse. So listen to our trailer for further and hopefully some humorous disclaimers. Now, we have nearly 20 sources and probably nearly 20 hours between Maria and I researching this report. And it's— look, part of it is the Citrini piece is, I believe you could fairly call it bear porn, meaning stock market bear, people who think the stock market's going down, and it's kind of porn to get them excited. Um, and there are people who have said on Twitter— I have no idea if it's true— that they're actually holding short positions. And I think one of the co-authors, Alop, actually, uh, admitted that. So, so you potentially have a piece of very well-written bear porn with a 3.4 to 1 ratio of scary words to nice words written by Citrini that, in my mind, hit a stock market that was kind of priced for perfection and kind of wanting to be scared. And Citrini produced this really— I mean, for what it is, it's a really— it's a little bit like the MIT NANDA paper. Like, I didn't think of much of it as a research project, but it was great content marketing. And this was great bear porn. If you wanted to scare a bunch of people and wipe out like $2 trillion in market cap, then this thing worked. Now, on the flip side, as Ray talked about, Citadel is a— by the way, Citrini is affiliated with the money manager, Ceterna Task Capital. Citrini Tax Capital, but not a particularly large one. Whereas Citadel is a top alternative investment manager, $65 billion in assets under management, founded in 1990 by Ken Griffin. Right. By contrast, Citrini was founded in 2023 by a famous healthcare entrepreneur called James Van Geelen, by the way. And they were famous, Citrini was, for calling weight loss drugs an AI. So they're kind of trying to pick sectors, if you will, looks like to me. Um, and Ken Griffin is obviously a famous hedge fund guy for 30 years and is sitting apparently a $74 billion gain, um, at Citadel. So the— I, I think, Ray, that's all I wanted to say before we actually talk about the report. Is there any other kind of metadata we need to talk about?

No, I think that at the end of the day, that's a great place to start. And I would like to dive into kind of the first point counterpoint between Citrine and Citadel.

Well, I gotta do my favorite tweet. I have a lot of favorite tweets on this, Ray, but my other one, I think it came from, oh, I can't remember his name, Chung Chan. Oh, what's his name? He's a finance shitposter. I'll get his precise name in a minute, but he posted the Franklin, a classic Franklin story, and it's got a picture of Franklin the turtle in front of 3 monitors that goes, Franklin learns how to put massive SaaS and payment short positions in place before writing AI doomer bear porn that tanks the market. And that's pretty funny, and that's certainly, One strong point of view on what Citrini was. But, but back to you, Ray, with a more serious commentary.

Okay. So kind of under the, does AI create this fundamental economic shift? Citrini says, and by the way, this is more of a simulation of what could happen. They specifically said it's not what they predict to happen, but they write it like they, they believe it. But they basically say AI is different than any other technology innovation because it displaces cognitive workers, white-collar workers, and that as labor income falls— because white-collar workers do represent almost 75% of total income in the US— as that labor income falls, consumption will weaken, and that will have severe economic impact. And then they made one more point, which is a little near adjacent because they basically infer that SaaS and the SaaSpocalypse begins this contagion across the economy, that credit markets that are tied to SaaS crack. And specifically, there's about $3 trillion of private credit markets in the US, and almost 25% of those are now tied to SaaS. So, and at the same time, I believe it's Bloomberg says almost $46 billion in tech loans are already at distressed levels. So Citrini basically says we got a lot of issues going on because of AI and it's going to create major economic collapse.

So 2 things, Ray. One, the shitposter in question is Trung Phan, P-H-A-N, Twitter @TrungTPhan, 750,000 followers. So this guy's a pretty well-followed shitposter, but he had some great commentary on this. Back to you. Citrini to me is what I'd call a crack in the window piece. It's a little bit like when you drive your car and a pebble hits the windshield and it cracks. And sometimes it just stays there and they can fix it for $50, right? And sometimes that crack starts propagating and eventually the whole window is shattered. And Citrini is very much like there's a little crack in the window. SaaS, which is less than 1% of GDP, like itself, SaaS becomes the chip in the windshield and that hits SaaS employment. Then it hits general employment. Then it hits every possible intermediary. And in one of the worst picked case studies ever, they decided to say DoorDash is going to be the problem. And there was a whole family of shitposting tweets about like, imagine looking at DoorDash and thinking the hard part was vibe coding the app, right? Like literally, it's like, yeah, imagine looking at DoorDash and thinking the app was the hard part. It's not the hard part. Right? The hard part is the network and getting the drivers and signing up the restaurants and building the marketplace and getting the subscribers and payments. And like, it's everything but a vibe-coded act to order food. So, but in any case, then they go through DoorDash and then they do the main economy because they say, oh, oh, then all white-collar workers, they're not going to have jobs, therefore mortgages, therefore housing. Right. And then the other one is therefore different direction, private credit, because SaaS goes out, it's 23% of private credit. So it takes out private credit. Ah, but the other side of private credit is apparently, I did not know this, don't know if it's true, insurance. So now the insurance business gets taken out. And did I miss anybody else, Ray? I mean, they literally take this chip on the windshield and look, I'm sure it's possible. And, but, but it's going to require very specific conditions for each of those propagations to happen. And they're kind of creating, in my opinion, like a 0.1% outlier story where all this stuff happens exactly the way they say it. But did I miss anybody else caught up in the carnage, uh, according to Citrini?

No, I think you hit them all. But they basically say that it's not just a SaaSpocalypse, it's an economic apocalypse.

Yeah.

And that's why I like Citadel, because Citadel came back with basically saying, hey, there's no evidence of a nonlinear shock, which means that there's no scenario in which productivity surges But aggregate demand collapses, and they basically said, "Hey, look at software engineering jobs. They're up from since January of 2024." They say that new technology adoption historically follows S-curves, and what that really means is it's going to go slower than anyone predicts. Kind of reminds me of that famous quote: We overestimate the impact of new technology in a short term and underestimate. the impact over the long term. So they're saying Citrini is overestimating the impact on labor and the economy in the short term. And they're saying there's no data right now that shows collapse dynamics.

So look, I like the Citadel piece too. I think it's more credible than Citrini piece. Um, I got 2 things I want to come back to here, Ray. Um, and maybe we'll do the second one second. Remind me later to talk about economists' reaction to this. But before we do that, I'll just tell you my take on Citrini, which is I think it is a labor shock. I think it's going to replace a lot of white-collar jobs, and that will have negative effects. And I think the reason that this bear porn is selling is because it's happening to us. I hate to be mean, but I think if we were debating the auto industry and those poor workers in Ohio and Michigan are going to lose their jobs, we would all sit very serenely in our offices writing software, trading stocks, doing real estate, doing other things and say, wow, those poor people who build cars, they're gonna get displaced, but I'm confident they'll find better jobs. Right. And if cars should be built, you know, if they can build cars better in Japan and cheaper in Japan, then whip out my David Ricardo International Theory of Trade 101 from economics Which says if they can build cars more efficiently over there than we can, then they should build cars over there and we should stop building them. So gosh, too bad for them, sucks for them, but good for America, good for the economy. That would be the reaction most people would have, in my opinion. And the reason that's not happening now is when we say sucks for them, them is us. And that's why this is so scary to everybody. Because the same very people who would sit down and very calmly talk about, well, autoworkers could be retrained, they can find new jobs, they can deploy to new things, right? Yeah, the autoworkers didn't feel that way about it, right? We know that. So I think it's how close it hits to home that is really making this story go like wildfire.

I agree. In fact, I got to double-click on that because one of my responses to this was, Number one, I think Citrini, their dark picture is way too aggressive, but it was focused on white collar. My initial thought went to, well, with AI plus robotics, I think there might be even, well, at least equal impact on blue collar jobs, like the 6 to 7 million truck drivers we have, the 1 to 2 million Uber drivers we have. So I think a lot of those jobs are going to be at risk over time. So I don't think it's going to be just a white-collar short-term issue. I think there's going to be some blue-collar impact. But at the end of the day, Dave, my, my real perspective is I don't think it's going to have as much impact in the short term as Sretrini said. And I think it's a matter of timing as to when we actually create new jobs and we can replace some of that lost work. From AI.

Yeah. And look, Ray, no one's arguing it's not going to affect Uber drivers and self-driving cars, yada, yada. But my main point is the reason I think the reason this is taking off is that once again, those examples people can talk about calmly because it's not them. But what makes this sell so much in the world that we live in, in the kind of knowledge worker world, is that holy cow, it's us. And that's why I think it's really taking off. I agree with you on the blue-collar jobs as well. And look, these transitions are rough. In the town I grew up, I grew up in Irvington, New York, there was a GM plant just up the river, up the Hudson River. And when you graduated from high school, you could either go work at the GM plant or go to college. And the kids who worked at the GM plant made more money than us for at least 7 years. By the time we were 25, the GM workers were still making more money. than those of us who went to college and got more white-collar jobs. And eventually we caught them and passed them. And then eventually, unfortunately, the factory shut down, see other discussion. But these displacements happen and they greatly affect the local communities. You don't need to talk to you about the Midwest or even on the Hudson River. We lost a car plant and that changed optionality for a bunch of people in terms of what jobs they could get. And then the alternative jobs, by the way, didn't drive the same productivity and wages that the car factory did. So, but it's all easy for us to say that. And it's all easy for us to go, you know what? In the end, it's good for the economy. And it was. Let's be clear. It is good for the economy. The only 2 questions are, one, who's it happening to? And the closer it is to us, the scarier it is. Because now it's us that has to find new jobs. It's us who need to learn new skills. It's us who had our jobs eaten, right? It's us who are unemployed with a mortgage, right? That's all scary. When it's somebody else, it's less scary.

In fact, yeah, well, that's it. So Trini, basically, to the point we've been talking about, you know, for the last 4 to 5 minutes, you know, it's different this time because it's cognitive work, because it substitutes white-collar tasks and high-income workers get displaced. And Citadel, their counterpoint to that is, says, hey, labor has always been displaced by technology, whether it was industrial machinery, automation, or software itself. And They actually brought this out, Dave. This is Citadel. They said, hey, when Microsoft created Office and Excel, it didn't replace office workers. It complemented office workers and increased their productivity, which I thought was a little bit of a stretch, but it was an interesting juxtaposition.

Yeah. I mean, we've talked about this before, Ray. I mean, look, technology always replaces labor. I'm not sure there's anything else you can replace. So we're used to the cycle of technology replacing labor. There's lots of jobs that we don't miss anymore. We talked about this in a prior episode, from slobber doffers to elevator operators to telephone operators, right? We don't miss those jobs. Again, if you were a telephone operator, it was painful. But if you were not, you just go, yeah, why does the world need telephone operators? Why does the world need elevator operators? We'll redeploy those people to higher jobs, higher productivity jobs. And in general, we do. So, I have a great deal of faith that over time this will happen. The issue to me is just how painful is it going to be and how quick is it going to happen? And obviously, the more quickly it happens, the more painful it's going to be. And I do think there are some risks there to talk about. And then the other one I would just say is it's about how quick is it going to go down? How disruptive is it going to be as a result? Oh, and what will the government do about it? Because, you know, we're in a democracy, at least for now. We have elected politicians. And there will be a political response to this. And I believe Citadel touches on that.

Yeah. And in fact, they do. And they talked about recursive capability doesn't equal recursive adoption. And what it was really talking about was, hey, when AI starts helping the AI get better, right? It's going to be even better at raising productivity. But 2 things you just said, number one, I believe that Government policy, specifically regulation of AI, because once they start seeing some real impact on unemployment and jobs, Dave, I think you're going to start seeing some regulatory issues around how do companies have to make sure they don't have too many, too much unemployment, that they fire too many people. I think they'll do that through some regulation. And then I think they're going to start putting a lot of regulation on AI itself, and that's going to slow adoption. across corporate America. And then the last thing is, hey, we know with COVID and every other emergency, at the end of the day, they'll print money to try to keep the economy humming.

So, okay, good. So there will be some political response. It is not lost on Washington and other governments that there's potential for large-scale unemployment if this transition goes down too quickly. But to me, long story short, Dave Kellogg's opinion is, We are— history is full of these examples of disruptive technology wiping out jobs, and the system empirically recovers and becomes more productive over time. And yes, it's really hard for those people impacted, but that is a cycle. And the reason this is so scary is it's hitting the white-collar people, and that's us. And that's why there's this whole extra, I'd say, 10x amplifier on fear. Now, the thing I wanted to say moving forward is the Citrini piece is largely macroeconomics. And Ray and I agreed that we were not going to try and get too deep into macroeconomics because neither one of us are economists. Uh, and if you want to debate economics, you should probably, you know, what's the old joke? If you want to have 3 opinions, ask 2 economists, right? Like, so, so, you know, we're not going to fall into that. But I want to talk about how real economists, quote unquote, responded to this piece because I think much like the MIT Nanda report, It was, which was impressive if you weren't actually a researcher. I think this, this is like impressive if you're not actually an economist, right? So it's kind of written for the pop market to scare the pop people. But, you know, Ray knows this. My older son is a PhD in economics and I mailed him this whole thing and said, what do you think? And he mailed me back a single link, which was to Noah Smith's opinion piece on this, right? His blog called Noah Opinion, which I think is pretty funny, right? Noah Opinion. So his piece back. was, the Citrini article is just a scary bedtime story. And basically dismissive. Like, in other words, Dad, it's not worth your time wasting your time on this. It's just a scary bedtime story. The other one in that vein was Jeremy Schwartz, who I think Ray has watched on TV. He wrote a very short piece that was just, there is no AI apocalypse. And he explains using macroeconomics, which I won't do because I'm not a macroeconomist, On just why, like, the piece is just kind of flawed. Ray, thoughts on this or any other economist reactions that you saw that you think made, you know, changed your view of how you saw all this?

You know, there also was a gentleman from Wharton. His name is Jeremy Siegel. So there's Jeremy Schwartz and Jeremy Siegel. He's another guy that had a pretty good, I would say, take on this, which is his basic Thesis of macroeconomic is that if we increase productivity, that there will be more opportunity for demand and more opportunity for greater income.

Yeah, Ray, and I crossed my wires there. Thank you for clarifying that. Jeremy Schwartz is the Chief Investment Officer at WisdomTree. He posted the Jeremy Siegel story. It almost looks like the Jeremy Siegel story was guest posted on Twitter under his account. So thank you for clarifying that. It's also on Siegel's blog. So, so excellent point. So look, we've got 2— I mean, basically my definition of real economist is someone with a PhD in economics, right? And if you don't, maybe you shouldn't be having these deep macro conversations. And we've got a couple of people who are well known in the field who aren't even— the issue to me, Ray, it's not that they're debating this fiercely. They're not even taking it seriously. They're just like, it's a scary bedtime story. Go away. And that's one of my messages for our audience here today. Ray, is, you know, I talk to a lot of operators all day, right? That's what I do. And at some point, like, I was having a talk this weekend with a founder about business books and whether he should read business books. And I'm like, for the most part, no. And he's like, what about education, executive education? I'm like, for the most part, no. Maybe once a year you treat yourself and you can go on a 1-week class or a 2-week class, but you're a founder. You got a full-time job to run your company. Let me worry about this. Hire me because I like to read business books and I'll read them all. And if there's one that I think is super apropos for you, I'll give it to you. But no, don't spend all your time reading business books. Don't spend all your time in executive education classes. Once again, if we have a specific problem, I'll help you find one. I've taken a couple at Stanford, between 2 and 4 weeks long, and they were great, right? But I like doing that. And I don't have a real job, quote unquote. And I feel the same way about all this AI doomer porn. Like, if you've got a real job, don't— Ray and I put at least 20 hours into this episode, just digging through all the articles and the sources and the point-counterpoints. Like, let us do it for you and we'll tell you what to worry about. And look, if you want to read it, you can, you can go find this Trini piece anywhere, but it's really not worth reading, even more so understanding or trying to understand it. It's a big investment of time unless you happen to know a lot about economics.

Yeah.

Yeah, totally agree.

And I'm just kind of a summary from me, Dave, and I think that's really good. Let us kind of spend all the time— same thing with a lot of these reports we read, right? There's like 5 or 6 things that are interesting, but you don't need to spend, you know, 5 hours reading Mary Meeker's 272-page report, right?

Ray actually enjoys doing that. Let him do it for you.

So, you know, I think both Citrine and Citadel had some interesting points. They have some points I agree with, points I disagree. And I think timing of how AI adoption is going to impact our economy, I think that's the biggest unknown is timing. Long-term, I think history favors Citadel. The market humans will respond to technology innovation that increases productivity. You know, I actually do feel that the impact of AI feels different, but as we've said here, a lot of it's because it impacts us The white-collar guys. And the last thing I would say is I do think the SaaS credit risk that was kind of overdramatized in the Citrini article— hey, if we've got $600 to $900 billion in private credit exposure that was based upon, you know, 15x revenue multiples, and you did a great job in your Top Line article talking about how that changes the entire financial equation for PE firms.

Yeah.

I do think we're going to see some issues with private credit and SaaS. I just don't believe it's going to create a big contagion across all of economy. And government intervention is a certainty. And I'm not sure how that's going to play out.

And just remember, Ray, it's going to have to be pretty bad on the private credit side because they're the first guys to get paid. Let's just remember, like, to me, I'd be worried about private equity because they get paid second in line. First in line is the debt holders. So, you know, if you buy a company with— and you take debt of 2x ARR, and okay, if you can't sell it 5 years later for 2x 5-year-ago ARR, right? Okay, I guess that can happen, but you're probably going to get 1.5 to 1.8x. The debt doesn't get destroyed, right? It gets back 1.8 when they wanted, you know, 2, right? So they lost some money, but they're not destroyed. The equity gets crushed because after the debt is paid, there's nothing left. So I'd worry more about the equity holders than I would about the debt holders. But it all depends how much leverage they took. And I don't understand the insurance argument, Ray, because they say this somehow blows up insurance. I didn't really get that. But I know a lot about software companies and their debt levels. And look, 1x ARR is more typical. And what if you can't sell a software company for 1x ARR? There's a massive, massive problem. And that's typically the debt burden. So I think it only even gets interesting when you've got more than 1x ARR on a company.

Yeah, not to pick on a company, but that means Domo would be a terrible credit risk right now because I think they're selling what, 0.8, 80 cents on the dollar?

Yeah, I don't know how much debt they have. But in any case, Ray, um, Ray's favorite punching bag, often mine, Domo. We do not give financial advice here. And that's it, ladies and gentlemen. Thank you for joining us for another episode of The Metrics Brothers.

Hey, that was fun, Dave. Thank you.

Thanks. The Metrics Brothers: AI, SaaS, and More is a production of the Metrics Brothers Growth and CAC and a member of the Benchmarket Podcast Network. By accessing this podcast, you acknowledge that the Metrics Brothers make no warranty, guarantee, or representation as to the accuracy or sufficiency of the information provided or the humor content of the jokes told. Ray, information, opinions, and recommendations presented are, according to our spouses, probably wrong and provided for information purposes only. This podcast should not be considered professional or for that matter, unprofessional advice. We disclaim any and all liability for any direct, indirect, indirect, misdirect, incidental, special, ordinary, consequential, inconsequential, or other damages arising out of any use of, or God help you, reliance upon the information presented here. Ray Grothreich is based in New York City and available on X @RayReich. Dave Kellogg is based in Silicon Valley and available on X @DaveKellogg. Kellogg. Schedectity, which is French for unspellable, is not our actual production location. Thanks for listening.