All right. Hello everybody. Welcome back to another episode of Basis Points. We are here live at Zeta Live with the CEO and co-founder of Zeta, Mr. David Steinberg. Thank you for taking the time.
Oh, it's great to see you guys. You know, I love being with you.
So, real quick, the event's amazing. 3,000 people.
Your team is amazing. So let's just give them their flowers because your team has done such an incredible job that this event is great. So, congratulations.
Thank you, Steve. I'll tell you, I'm so proud of the team. A lot of stuff comes together a little later than you'd like, but it's been, you know, just so far today a home run.
So, I'm going to start it off with, when I see the people that are here, it's not people like admitted. You're talking about CMOs are here, talking about decision makers that have direct lines to their board of directors and CEOs about why they should make a decision.
Oh, we have a lot of CEOs here, too.
A lot of CEOs. You got a lot of CFOs here. You got C-suite here. If you had one message, what would you want them to leave with, really capturing your day?
Let me start by saying that I know there's been talk online, oh, why would I want to see somebody carrying a ball or so on and so forth. You bring the athletes, you bring the movie stars, you bring the rockstar entrepreneurs. That brings in the customers. That brings in the potential customers. In the room today is $200 billion a year in decision makers for marketing and technology today.
And usually those celebrities and athletes are involved in marketing and decision making.
Of course. And when you think about it, it's funny. Last year, for every dollar we spent on this platform, we generated booked business of 10 times that amount.
This year, I think we'll do better than that.
Love that. So the reality is, the message is, why Athena operating system, why AIM, why ultimately other things. How do we continue to drive a higher return on investment for our clients than anybody else, bar none?
6 to 700% on marketing spend goes up dramatically in an AIM environment. It goes up even further when you get to the chip level. 297% confirmed return on marketing spend today—I'm sorry, I apologize—on technology spend. 6 to 700% on marketing, 297% return on technology spend today. That's what I want CEOs to hear, CMOs to hear, and CFOs to hear.
Athena OS, that was one of the major four announcements today. What is the new trajectory that you think this can lead Zeta onwards towards? Because it's not just about marketing intelligence anymore. It feels like it's much broader than just marketing for a company.
It is. Well, we've moved out of marketing, you know, a number of months ago and really added the business use case. That's what drove this, right? So when you added the ZBI, what happened was, perfect example, we had a client, we're doing a tremendous amount of marketing for them around skirts in the Northeast for women because that's what the demand was. Problem was, all the inventory was in the Midwest. So now all of a sudden you put together a system where inventory can know where the demand is before it's there. Let's take it a step further. You can then use that for manufacturing, right? What is the demand, what do we need to make? Then you can use it for the best logistics. You can adjust weather. You can adjust everything that's necessary to make the entire system work. Each decision makes the next decision smarter. So when you look at return on investment, our goal is to be the highest return on investment across technology and marketing for our clients, bar none.
And I think that was part of the relationship you guys expanded on today with Palantir, bringing their ontology together with the Zeta ecosystem. Karp was on stage. He doesn't go to many conferences. So—
He does not. And let's just stop and say how amazing—
Dr. Karp is an incredible individual, and just incredible. And to call him a friend—I mean, I don't know if you heard—I thought my epitaph would be his line from the press release, that, you know, it's an honor for Palantir to work with Zeta.
Yeah. Right. My new epitaph is, 'David Steinberg's really funny in public, but he's much funnier in private.' I was like, that's a compliment. He's very funny.
And I tend to be very funny. So, we have a good time together.
So, I got to ask you, we learned a lot of things. We tried getting it out of you last night. You did not break. It's like you were trained by the CIA. You did not break. We tried. But I got to ask, with what I saw, Palantir Foundry rearchitect—and one of the things I told you originally why I became an investor is because of your speech where you redeployed 400 engineers, and you said to the board of directors, which is not an easy thing to do even if you founded the company, we made $50 million this year in EBITDA and we're going to make zero for the next three years because we have to re-engineer everything for the future. I want to know specifically how the chip fits into this whole flywheel and give us really the insights behind this decision.
And let's also back up and say you guys announced that you're working with a company to develop a microchip.
Yeah, let me make this very clear, just to be clear. We're working on a microchip with a third-party manufacturer who's going to put up all of the money to make the chip. This is not going to affect our free cash flow. I know that's created some confusion in the marketplace. When you look at our business, last quarter we grew top line by 44%, free EBITDA by 56%, free cash flow by 73%. Yeah, I think we're going to continue to grow our free cash flow at very, very high percentages. This chip will not affect free cash flow. It will not affect our share buyback, and it will not affect our ability to create cash to buy new cool companies. We have a manufacturing partner who's going to make it. We'll have another third-party company to buy the chips from them and host them, and then we will use those chips on demand exclusively. So, it's a very differentiated type of relationship. I probably should have been clearer about it online because I think it's created some major confusion in the marketplace.
No, because the market sees capex, chips, all this stuff. So, it's a fair point.
People saying, oh, they're going to put all their free cash flow into making a chip. That is absolutely not the case. Why make the chip? Today we drive 6 to 700% return on marketing spend, 297% return on technology spend. The chip will speed everything up. That speed will help us. What's our goal? I've been very clear. I ultimately want to get to 10 times return on marketing spend. How do I get to a thousand percent return on marketing spend? And how do we get to 5 to 700% return on technology spend? We do that by getting deeper and deeper into the infrastructure level in a way that doesn't cost us any money.
Is this going to be part of the sovereign AI ecosystem as well?
100%. This all sits on top of Foundry. It's going to be in partnership with Palantir. Their FTEs are helping us all along.
Well, Elias told us that you guys have developed a very strong relationship over the past couple of months and the outcomes-based methodology that you both have when approaching enterprises. That's what makes this special.
It's almost crazy when you look at a guy like Elias, you look at a guy like Dr. Karp, Alex, and me, and we're just totally aligned. The organizations are totally aligned. Our FTEs are now going into deployments together.
Yeah, that's what he told us, right?
And it's been just unbelievable.
Can you talk a little bit about the Neotron Nvidia partnership?
Yeah, listen, we're going to run these inference models on Neotron in partnership with Foundry—I'm sorry, it's on top of Foundry. It's been a very long day—in partnership with Fireworks AI. And, you know, listen, if in the first nine months of the year, we'll have now announced partnerships with OpenAI, Snowflake, Palantir, Fireworks, and Nvidia.
We need to do some more. It's been a slow year.
Really slacking off now. We need a fifth announcement today.
Mr. Steinberg, I have someone else who actually introduced me to your company.
Mr. Steinberg—Jesus, now I really feel old and formal.
This person introduced me to your company and he has a question. So I would love if he could ask a question right here.
Okay. So, well, thank you for letting me be here and have the opportunity to even ask a question. My question, I want to probe into the Palantir partnership a little bit. We know that Palantir opens up many more doors beyond commercial, and that's kind of what Zeta focuses on, is the commercial sector. And I know you have the Zeta Economic Index, which tracks consumer data.
But do you see a pathway into the public sector where, yes, for example, Zeta might help the Fed?
Yeah, we're already working on stuff.
No. Did you see my introduction to Karp today?
What did I say in addition to commercial?
Country. He may have said—
I said countries in my introduction. I said enterprises and countries.
And you may have hired somebody.
Well, you guys track who we hire.
You may have hired somebody in the past couple of days that is very influential overseas and may be from a very large company that may know a thing or two about countries.
Yeah, listen, I see a public sector business for us. We're already building it, and I think it's going to be something that's very interesting.
Love that. I love how you're expanding, expanding fast, doing it with the right people.
It's an amazing journey to be on, honestly.
I'm excited to be on it. We love having you guys with us. How was Benjamin's last night?
I told you, best steakhouse in Manhattan.
No, no, no. See, that's—we're going to both there and Peter Luger's. We're going to end the debate.
It's not. I have to agree with Steve. Luger's, the day before, first time, unbiased opinion—better.
I'm just telling you, the '86 and Fort Charles are both pretty good too.
So I have a question about OpenAI.
As you know, our group tends to speculate a little bit, and we tend to dive into things much deeper than most people. And one of the things that we have been thinking about collectively is, for OpenAI to reach a lot of the projections people are expecting, they have this real estate. And if you look at Amazon, they built a hundred billion dollar advertising business out of nowhere over a two, two-and-a-half year period. I think, and other people think, that in order for OpenAI to hit the revenue projections as expected of them for the amount of capex that they are allocating, they're going to have to utilize their retail space. And when you think about where Anthropic and ChatGPT from OpenAI are—Anthropic is really enterprise level, ChatGPT is consumer level. So are you going to be a little—
We already are serving ads through there too. So, we're serving ads there today. I'll be careful here. Yes, we're serving ads. It's a very meaningful and growing partnership, and I think it could be very, very big. But I think it's unfair to just call OpenAI retail.
So, let me phrase that from a coding aspect. Anthropic has the majority of coding, and ChatGPT has just huge consumer.
I think, once again, they—I think they both have a lot of both. I've got to be a little careful. But to your point, OpenAI has made the decision to go into advertising, and Anthropic has not. So that's a major difference. And I've said publicly that we are partnering with them, and we are actively serving ads into OpenAI. It's growing very quickly, and I think it's going to be a very interesting partnership.
And by the way, the Seno AI deal is going to be very, very important there.
David, thank you for your time today. Last question for you. What do you think—and you've spoken about this before—but what do you think the marketplace in general misunderstands about Zeta? I mean, this event is amazing. You've got thousands of people. I think the event and the company's brand has progressed so much over the past year. What do you still think people don't really understand about the culture of this company?
I think first, people think we're an advertising company, which we've never really been. We were a marketing technology company. 60% of our revenue is subscription, 40% is utilization. You know what they call utilization this week? Tokens.
Right. I mean, we don't call it that.
With 120% net dollar retention, right? I mean, gold standard is 100% net retention. We were at 120 last year. You could look at all the different metrics and we perform ahead. And then we make a mention about chip, and then all of a sudden everybody's, 'Oh my god, they're going to spend all their cash flow.' It's just, some of the noise that comes out just makes no sense.
Now, it creates opportunities.
Yeah, because if people don't understand what's really going on, others do. The opportunity is there.
That is correct. But what I would tell you is that we are building the next generation of AI-driven software company. We are going to build a $10 billion a year business at a 30% plus operating margin with the vast majority of that falling to free cash flow. And anybody who's betting against us is in the end going to probably lose. But what I'll tell you is, different moves on a day-to-day basis, week-by-week basis, are not as relevant as how do we build a hundred billion dollar business, and that's what we're working for.
Well, I wouldn't bet against you, especially with what I'm seeing—
Oh, that I'm not even betting. I'm correct on that.
David Steinberg, thank you so much.
Thank you so much.
Hello everybody. Welcome back to Basis Points. We are here live from Zeta Live 2026 in New York City. We are here with someone very special, Mr. Elias Davis of Palantir. Elias, thanks for sitting down and doing this.
Of course. Thanks for having me. Good to see you, buddy.
And before we get started, Ahmed is making us wear the headphones, and it is weird. Is it not weird?
Yeah. Yeah. I feel very uncomfortable. But anything for my guy here.
Everything is amplified in our ears right now.
Basis Points in the future. We'll have headphones. The biggest smile I've ever seen because it makes us uncomfortable, so it makes it all better. He's like, 'Steve's uncomfortable. We are doing this when we build the studio.'
Elias, yes. I've known you for a while now.
You are a rockstar at Palantir. You get things done. Steve and I have been covering Zeta for the better part of about a year now. Around November, December 2025, we got into it a few months ago. One of the companies we've been covering for five years decides to partner with Zeta. Palantir doesn't really do that many partnerships, especially with companies in marketing intelligence. How did the partnership actually start?
Yeah, it's a great question. First of all, thanks for having me. Good to see you guys. How did it all start? I think I forget exactly who actually got connected or who David connected with originally. But in very typical Palantirian fashion—thank you for the compliments. Luckily it's not just me. Everyone at Palantir is a rockstar. Got a note from, I forget if it was KK or Karp or whoever that might be, and we were off to the races. I think David had originally gotten connected to, I forget exactly who, but had a couple cycles. Didn't exactly go where everyone intended it to go. And then, and I'm never going to forget it, I went to his office on Park Avenue. We sat in a room for 45 minutes and we walked out shaking hands, sort of kind of, with the details of the deal figured out. And then we were off to the races. I think he said on a podcast that we're tied at the hip. So, yeah, my wife is a little jealous. But it's been great. He's an incredible individual, as you guys already know. And in very typical Palantirian fashion, sure, we'll talk about what the partnership and its foundation is all about. But ultimately it comes down to the people. You've covered us, you know us super well. Obviously you know everyone at our company. Everything we do with a third-party company, partner, customer, whatever it might be, it really starts with the people. Do we like them? Do we align with the same values, same mission? And every single conversation we had with David, 100% aligned with that. And obviously we saw Dr. Karp and David on main stage open the show up this morning. They really, really like each other, and obviously that helps with making sure the foundation's strong and things are kicked off in the right way.
It's exciting. It's a seven-year partnership.
And David Steinberg is known for ripping out and just redoing everything. Prior to 2020, he deployed 400 engineers to rip and replace the system, and now they're on Foundry. So, I can only imagine what this is going to look like.
Yeah. Look, I think the beauty of the first meeting we ever had was that there was no expectation that was set. And so I think, very similar to what you just said, David is the type of human being who thinks from first principles, similar to the way that we do as an organization. And yeah, I mean, through that 45-minute meeting—and I can sort of kind of walk you through some of it—the whole vision of what this partnership would look like was sort of kind of orchestrated. And obviously now, for the last few months, it's been a lot of hard work from our team, our forward-deployed engineers, their version of forward-deployed engineers as well. And then obviously me and David running around doing what we do to make sure that this is as successful as obviously it has been. Obviously their stock has been ripping. Thankfully ours has as well. It's always nice when a company partners with Palantir and the market genuinely recognizes the value that could be there. You know, I mean, Palantir may or may not have gone past 200 this morning.
I saw that. I saw that. Let's talk deeper about the synergies of the partnership. So, Zeta is marketing intelligence. Palantir is much more broad-based business intelligence at a very deep layer. One could argue that in this AI revolution, the models are important, but the orchestration of the models is really what's important. Dr. Karp was talking about this back in 2023. That thesis has started to play out. How do you see marketing intelligence from Zeta and their relationship with many customers in the Fortune 500 working with the integrated stack on Palantir? Where does the synergy lead to business outcomes that are differentiated for customers on a marketing and a business side in the general operations that Palantir works with, with the ontology?
Yeah, absolutely. So I think the first question we had to ask ourselves in that meeting that we sat down for 45 minutes was, what is this really going to be all about, right? We can do a bunch of press, our stock will probably go up, but unless there's strong foundation to the actual use cases and the groundwork we have to lay that makes this partnership strong, that won't sustain itself long term. Right. And so I think the question we really had to start asking ourselves is, what primitives on Foundry are core to benefits that we can drive for Zeta customers? I think if I distill it down to two main areas where the partnership focuses, one is the acceleration of onboarding. Obviously AIT knows this super well, but the data connectors, the billions of dollars of R&D that we at Palantir have put into building Foundry and getting it to where it is today is obviously extremely difficult to replicate, impossible to replicate, some would say. And so obviously that is a key part of this, right? The onboarding process, the acceleration of that onboarding process for Zeta customers just accelerates the path to value creation. Now, touching on what expected value Zeta customers intend to see, ultimately the foundational value proposition for Zeta is leveraging their data and Zeta's proprietary data that they don't sell, that hundreds of millions of users opt in for, to drive marketing ROI. And so when you distill down the primitives and first principles of what Foundry is, ultimately it's the acceleration of go-to-market and/or onboarding, then go-to-market with data connectors ingesting all the data, creating that ontology so that the Zeta customers have a 360 point of view of where their data sits, how they can approach different campaigns, and then ultimately the integration of that with the Zeta cloud platform is ultimately what drives and accelerates the path to the expected 600, 700% ROI.
Exactly. So it's almost like if a customer wanted to sell a water bottle that was like a really cool water bottle, they need data from Zeta on who is this type of person, maybe that goes to the gym often, maybe that's into these niche water bottles that would actually buy this. But they also need the supply chain of how to build that water bottle, and a lot of that stuff comes from Palantir's ontology. So, do you think that's a clear example of how you need the marketing intelligence on who's going to buy it, you need the logistical operational intelligence of where we're going to source the water bottle parts from, and putting that together is where you get a much better return for customers and, quite frankly, a much better return for the company that's trying to optimize revenue and decrease costs?
100%. And actually, I unfortunately can't disclose the name of the customer, but I think this is like part two of what this partnership stands for. We're working together to go to mutual customers and Palantir commercial customers to actually provide this offering to them. I'll give you a perfect example, and I think you touched on this a little, but if we, Palantir, on the commercial side, started a lot of the use cases in global supply chain, right. And so if you think of a very, very large publicly traded consumer-facing product goods business, obviously you can leverage Zeta's proprietary data and the customer's data to drive an increase in ROI and user acquisition, yes. But ultimately you can't actually deliver that product to that end consumer after you acquire them unless you have visibility and unlock all the bottlenecks in that global supply chain, which obviously is powered by us on our end, which is why, you know, everyone's like, you know, go to market, do we really do that? Ultimately, everything happens at Palantir, because it makes sense for our customer. If we can help drive incremental value to the bottom line to our customer, in this case both on the marketing user acquisition side as well as the supply chain, it ultimately allows the acceleration of delivery of the product to the end consumer, which obviously drives incremental unit economics for our business as well as Zeta's.
So, what can a new customer expect? Palantir revolutionized the RFP process. It revolutionized actually doing an implementation from what could take six, twelve, eighteen months to several weeks. When you combine the Palantir supply chain and Zeta marketing and how implementation is now—Zeta services 51 out of the top Fortune 100s.
For those 49, what would the message be to them that aren't using it yet?
Absolutely. So I think I'm not 100% if Zeta sort of kind of took this from us or sort of kind of did it themselves. Either way is fine. But the idea here is like the actual go-to-market process is very similar to ours now, right? You do a small pilot, boot camp, agent camp, whatever the thing might be, within a quarter. How do you do two things in parallel? Convince a customer to try it and have the technology slash the actual integration piece align with the process of onboarding, go-to-market, value creation. How do you do everything within a quarter? Obviously we're public, obviously they're public, obviously their customers—many of their customers—are public. Everyone wants a value, you know that super well. We're just helping make that happen. And so if you're a prospective customer for Zeta, I think the way we have to think about it is: do I spend a lot of money on marketing? Am I optimizing that spend? Am I utilizing all the data that I have to accelerate the economics and catalyze the economics for my core business? Let's try it out. And it's not—similar to the way we say this—it's not like, let's try a pilot six to twelve months, throw hundreds of millions of dollars.
It's an old way of selling software. It's done, right?
It's done. It's done. And I think I didn't catch all of the news and announcements that Zeta reported at opening this morning, but I think there's a lot of forward-looking perspective of new markets that Zeta, in partnership with us, are looking to break into to re-energize the market of what this new vision for software looks like, which we're obviously very, very excited to support them on.
And it's, you know, Palantir fundamentally changed the go-to-market approach for software. Fundamentally changed what it means for a company to be able to grow—in particular, forget the top-line growth of 90%, but the operating margins—like, you just don't get software companies that are this efficient. And I think the exciting thing for me and Steve when they partnered with Zeta was like, okay, Zeta's hitting the inflection point of operating leverage. If they had a Palantir partnership, that could probably supercharge what it means—rule of 180—but like to get those types of numbers, you really need the type of characteristics that you're talking about. Dr. Karp doesn't really go to these events and speak live on stages. He's a very busy man. He has a lot of things going on. He's running a half a trillion dollar company. So, it's obviously hard to get his time, but he came to this one. What is so special between Dr. Karp's relationship with David Steinberg that actually compelled him to come to this event?
Look, I think it's not an artificial relationship. It's not transactional. I think ultimately—I hate to say this—but in corporate America people do business, and sometimes, most of the time, it's transactional. You know Dr. Karp better than most people. He does things he enjoys. He does things with people he likes to spend time with. The relationship is genuine. He was asked to do it, and he was excited to participate. And obviously, you know, we all saw it this morning.
Very cool. It's absolutely incredible. Last question for you guys. Thank you so much for taking the time to do this. You do a lot of stuff at Palantir. When a partnership like this comes into place, what is your role? Like, are you managing relationships? Are you trying to understand the financials of it? How are you, as a rockstar at this company, working with other rockstars to make a rockstar deal happen?
I mean, look, I think it comes down to the origin of every Palantirian, right? Like, all of our jobs are to create value for our own respective organization and the customers that we work with, our partners that we work with. And so whether it's, you know, being on a bunch of podcasts and fireside chats with David, working on the go-to-market piece to make sure that this partnership is successful and that obviously everyone is extracting value from the hard work that our teams are working on overnight or around the clock, working very, very closely with our forward-deployed engineers that are on site in Zeta's office, collaborating on a lot of the hard work to actually make all of this possible—a lot of what we discussed this morning. But it comes down to the same thing. David and I, I'll sort of kind of loop it back to where we started. He's an incredible individual. I look up to him a lot. I've spent a lot of time with him. I've added a lot of value to him. He's done the same to me. And that's what makes a lot of what this partnership is about very special for us as well as, I think, for them.
And that's how business should happen. You should like working with the people that you're working with.
Thank you for having me, guys.
Thank you. Thank you very much.
Oh wow. Everything sounds better now.
Oh—
It actually wasn't that bad.
Thank god that's still rolling for you saying that. Oh my god. It actually sounds a little better now. Yeah. Can I get a clip of just when Ahmed said that at the end?
Hello everybody. Welcome back to another episode of Basis Points. We are here live from Zeta Live with the one and only, the man, the myth—I think we got to say he's a legend on Wall Street. Ladies and gentlemen, Dan Ives.
So, Dan, we met back in 2023 because you were the first Wall Street analyst. Isn't that crazy? The first guy—
And look at everybody jump on the bandwagon now.
The first guy at an institution that is responsible for tens of trillions of dollars to say, 'Hey, I think Palantir at $15 looks kind of compelling.' Honestly, this year I think you were one of the first couple of people to say Zeta is kind of interesting, and you did an interview with Steinberg and said maybe this company is something that people should look at. What has it been like over the past seven months in terms of the progress you've seen from the company?
Yeah, I mean, you know, and obviously as an advisor for David and Zeta, look, my view is this: in this AI revolution, there's going to be so many companies, and I'd say obviously Palantir being the first of them, but it's really about like the second, third, fourth derivatives that are happening across AI. And I think like if you look at Zeta, I think obviously the Palantir announcement, which is so important.
And so many others where I think it's a good example of, in the SaaS apocalypse, whether it's Zeta or so many of these others—ServiceNow, Salesforce, Adobe—like, software companies definitely have structural issues, but you cannot just label. Like, if you look at what Zeta does from a marketing perspective and how integral they are from an AI strategy perspective, customers—they're not, everyone that's here, they're not going to Zeta because they think that they're going away to some extent. It just speaks to them spending more and more money, right, in terms of with the Zeta platform as they embrace AI.
So, let's talk about that, what you just said. You know better than anybody. My biggest thesis for Palantir was net dollar retention when nobody was talking about it, and we had that conversation on Basis Points, and I think the net dollar retention at that point was 146%. Maybe I'm wrong. Maybe it was 138. I don't remember exactly what quarter it was.
It was somewhere right in that area. And you look at what Zeta is doing, and you look at the 20 beating and raising quarters, you look at all the underlying metrics, and you look at that net dollar retention at 120% right now—it still gives me Palantir vibes from several years ago because it just looks like it's going to expand, especially with what they announced today. How are you looking at this?
I think, look, for Zeta it's about install base. You're going to have new companies, but it's like the opportunity for them is really—and look, David, they've laid out everything you know here at the conference, and obviously a lot of the announcements, it's just further monetization of the platform. Because look, the reality is, take all the customers that are here and so many other user conferences, they're trying to figure out—they have an edict from the board: you have to embrace AI. Okay, so like they spent, whether it's like getting in line for GPUs, data centers, whether it's on Palantir, whether it's the hyperscalers, when it comes to strategically with customers, marketing budgets, analytics—who do you go to? You can't go to a hyperscaler for that to some extent. I think it just speaks to why, like, if you could say, like, a lot of their partnerships with some of the biggest tech brands in the world, that's not coincidental. I think it just shows they recognize the opportunity, but it all speaks to customers needing to spend on AI use cases. The use cases are going to be the key to everything we see playing out.
When it comes to the Palantir partnership, Palantir is like broad-based intelligence via the ontology. Zeta is like very hyper-focused in marketing. Where do you see the synergies there over the coming years on that partnership?
I view it as like the engine is really Palantir, right? But Palantir—that's not who they're going to be. They need to partner, whether it's Zeta on the marketing side, whether it might be other companies in the financials, others on government, you know, depending on the different use cases and projects. Look, I think it's really Palantir recognizing, like, sovereign AI to me is the key, because customers have to have control of their own data. And if you look, let's say a Zeta customer—they're not just going to put that out to like a frontier model. That needs to be within their ecosystem. And I think the monetization to go at it with Palantir, look, I think that's like sort of the magical one-two punch that we've seen play out.
And when you have that one-two punch, we saw a Gap come over not too long ago. Do you think this is the start of major brands recognizing that there may be a better way to do things than what they were previously doing? And part two of that, Athena OS—do you think that that's the driver to get more brands like Gap, who have a huge customer install base, over to Zeta?
I think Athena is sort of the Trojan horse, like that's sort of the linchpin to the whole strategy when you have Gap or other retail, because I see it as someone that just follows AI, you know, like around the US and globally. So many of these customers have gone at it themselves. They felt like they could build something internal, right? Then all of a sudden you realize, like, I'm trying to build a coffee shop, then Ahmed's sitting there in Starbucks with his cappuccino. Okay, why would I build something here? And I think that's what you're recognizing in terms of what the opportunity is, what they've actually built in terms of the stack.
What are your current thoughts about the frontier labs right now and the models? I mean, they're super important to the AI trade because they are a lot of the demand for the capex, but at the same time you've got this sovereignty open source stuff that's kind of competing with them. Like, how do you resolve everything that's going on right now?
Look, listen, 5% of companies in the world have gone down the AI path. And I think as you start to see it all—
Oh, wait. So AI is not fake.
Look, I know. No, but it is so true. It's like, I think part of what's happened—even a lot of the bears that have called 10 of the last two downturns—they're trying to figure out in terms of like, oh, are these use cases real? What happens if companies actually don't build out data centers? The theme's accelerated. But to your point, the frontier models—my view is like, frontier models over time will commoditize. OpenAI and Anthropic are the first ones that understand, well, they got to go after the enterprise. So I'm just not of the view where it's like with open source or things that, you know, whether it's Jensen or others, even DeepSeek in terms of what we see in China—the cheaper the models get, the better it is for the AI theme. But you're going to have three to four trillion being spent over the coming years. Companies, financial institutions that have been so far down the AI path—they can maybe slow down some tokenization spending with the model players, but where are they going to spend it? They'll spend it with software companies, use cases, sovereign AI, Palantir, others. I think that's ultimately where it's going. I think we're still in the second, third, fourth derivative of the AI trade playing out.
So, I'm glad you brought that up because I have a bit of a disagreement. I don't even think we've left for dinner yet. So, you and I, you and I both were—we were talking about what we were in last. I'm like, I know I've redone my thesis. I don't even think we've left for dinner. I don't even think we're at happy hour.
But I like the thesis because just take a step back. My view is consumer AI hasn't started.
Apple, we're just starting. Meta, we're just starting. Physical AI, not even started. But that's why it's like, we're in year three of what's going to be a 10 to 12 year build out of the AI revolution.
And I think as that plays out, look, chip—we'll be in Korea next week—but it's like, chip equilibrium at this pace doesn't happen till early 2029.
What I've been saying to people is you have to look at the cloud buildout, because there are companies—and you made a great point—with less than 5% of companies have gone down the AI rabbit hole. When you look at the on-prem to off-prem off-ramp, it took 15, 20 years and companies are still going through a cloud migration. The idea that AI is just going to somehow stop in a year or two is crazy. And I think Dan's 100% correct that companies like Zeta, Zeta in particular with what they did today with Athena OS graduating from just the marketing to a full operating system that gives you everything predictable is going to be the next iteration. But also, this is not—you can't even walk because of the crowds here. People aren't just here for the cappuccino and the fruit. Although the food is really good—
They're here because in those companies, whatever the vertical is, they need to come up with their use case and how they're going to go out.
Yep. It just comes down to that.
Yep. It really reminds me of AIP.
It's so similar, if not this last AIP. Well, I'm going back—if you go back like two years.
Two years. Yeah, it was like a couple hundred people and like you could see those early use cases with a couple clients, and now it's become a lot bigger. Palantir's brand has grown. I mean, if there's a world in which Zeta keeps executing at the level that they're executing at and they start hitting those inflection numbers on free cash flow like you talk about so much with Palantir, it can get exciting.
Oh, think about how many people here are decision makers at their companies or direct links to the board.
Well, the thing I think that is so important in terms of, like, who ultimately is—like, who's the crowd here? The crowd is not us. The crowd is CMOs. It's CMOs across the board, or other C-level.
Yeah. Because you think about all the breakout sessions they're having now. This is directed at CMOs and what we can do for you. But if you're a CMO, you can't just come out here like, 'Yeah, conference was good, we're going to continue.' You need to have answers how you're going to build out your AI strategy for your—because guess what? If you're in advertising and you slow down, and I'm your competitor, I excel.
Yep.
No, it's like—and I just think we're getting to that point to where it is.
Yeah. Companies cannot stand still anymore because their competitors are not standing still, and every board is asking those questions about how do we get a bigger piece of the pie.
But this is a barometer for the broader—as we all three of us talk about—the broader AI trade: where demand's going, where spending. Second, third—remember, for every dollar spent on a video chip there's an 8-to-10 multiplier across, you know, across the rest of tech. And that's why, like, the bears, when they're in hibernation mode, they can see AI in the spreadsheets.
Oh, they just—they get to the point where they don't understand that capex being spent can take two, two and a half years to turn into revenue from the RPO that turns into cash from operations that becomes free cash flow. They're worried about negative free cash flow for a year or two, but investing is a long-term game. As a shareholder of the large companies, I want them deploying capex because it fortifies their moat, and they are dictating where tech goes.
Vegas strip build out, 1955. That's where we are. There's one strip.
And guess what? If you don't build on the strip, somebody else will, and you'll be in Reno.
And who the hell goes to Reno? And Reno actually—I will say, Reno actually has like really good food under the radar, and that's obviously where Tesla and a lot of other tech companies are. But the point is, is that there's one Vegas strip, one sphere.
You just initiated your top five picks for 2027: Apple, Palantir, Nvidia, Tesla, CrowdStrike. What do those names have that are standing out to you right now?
I mean, to me it's like when I think about the AI revolution, and I think it all starts with the one chip in the world fueling the AI revolution, led by the godfather Jensen, you know, in terms of what they're—and I think investors continue to underestimate earnings power for Nvidia. Microsoft, it's their backyard on the enterprise. New York City cab drivers bearish in the stock six months ago, but if you see it continue to play out, that stock could have a six or a seven in front of it, I believe, over the next year. Then you look at CrowdStrike, Palantir obviously we're another big, huge fan of. But surface area is expanding, cybersecurity—I think budgets could ultimately double over the next few years. And of course, look, Palantir—I just think Palantir is now going into the next phase of its growth frontier when it comes to everything that's happened on the enterprise.
Companies where they are—I think incremental spending—if a company was spending 5% of their budget related to Palantir, and now it could be 10 to 15%. And Apple, they're a toll collector on the AI highway, right?
Tesla—you think robotics is on the way?
I think it's just this is a physical AI play that's right now in a transition from an EV company to what ultimately is going to be Optimus, CyberCabs, autonomous.
Are you upset RoboTaxi hasn't really played out as fast as people want it to, or do you think it's inevitable that it will happen?
I think, look, has it been disappointing, the ramp of robo? Of course. But is a lot of that because of regulatory? It's also they know, like if they hear a squirrel on the road, they'll have some regulatory assessment. So I think part of it is that. To me, it's a matter of when, not if, as they build it out. And I just continue to believe by the end of next year there's over 80% chance that SpaceX and Tesla merge.
Why? Why do you say that?
