WEBVTT 00:00:52.280 --> 00:00:55.117 Obviously, we are loving these results 00:00:55.200 --> 00:00:59.717 and um uh loving what they mean for our 00:00:59.800 --> 00:01:02.077 customers and and broadly speaking the 00:01:02.160 --> 00:01:05.317 West. Um so 00:01:05.400 --> 00:01:08.397 reflections on how we got to 93% 00:01:08.480 --> 00:01:09.917 aggregate growth, 00:01:10.000 --> 00:01:13.917 just under 150% growth in US commercial 00:01:14.000 --> 00:01:17.277 and an aggregate growth of 115% in 00:01:17.360 --> 00:01:20.597 America, which is astonishing, uh even 00:01:20.680 --> 00:01:23.197 surpassing the already anomalous results 00:01:23.280 --> 00:01:25.597 we've posted in the past and at a very 00:01:25.680 --> 00:01:28.757 significant scale. Um And and this the 00:01:28.840 --> 00:01:31.317 the the story really does begin 00:01:31.400 --> 00:01:33.557 uh at the beginning when we dedicated 00:01:33.640 --> 00:01:39.037 ourselves to our most important partners in the US government and we we built 00:01:39.120 --> 00:01:42.597 products to deliver value for them. 00:01:42.680 --> 00:01:45.997 We delivered those uh value for those for them 00:01:46.080 --> 00:01:48.637 by looking at the world in its naked 00:01:48.720 --> 00:01:52.037 state. We did not have uh soft We did 00:01:52.120 --> 00:01:55.837 not have AI available. We had to work with NLP. So, we had to develop forward 00:01:55.920 --> 00:02:00.637 to play uh engineer model to extend the technology and deliver value. The The 00:02:00.720 --> 00:02:03.317 nascent version of ontology was developed. 00:02:03.400 --> 00:02:07.037 Sham, Aki, others strapped 00:02:07.120 --> 00:02:08.997 Blackberries around their head and made 00:02:09.080 --> 00:02:12.237 the code work in sensitive environments. 00:02:12.320 --> 00:02:14.277 And what we learned and what was built 00:02:14.360 --> 00:02:16.717 into this company is that there are 00:02:16.800 --> 00:02:20.397 there are things, values, structures 00:02:20.480 --> 00:02:25.237 that are more important than purely uh 00:02:25.320 --> 00:02:30.117 extracting value from a client. And we rejected the the way in which we were 00:02:30.200 --> 00:02:33.757 being taught in Silicon Valley to build a 00:02:33.840 --> 00:02:37.397 software company at the time. Palantir, of course, is now both infrastructure 00:02:37.480 --> 00:02:41.357 software FTEs, orchestration, and business know-how. So, it's a completely 00:02:41.440 --> 00:02:45.877 different hybrid. But at the time, we were being told our job was to trick the 00:02:45.960 --> 00:02:49.557 clients into giving us money for something that made them attached to us, 00:02:49.640 --> 00:02:52.517 but really uh really added no value. So, 00:02:52.600 --> 00:02:55.877 essentially a parasitic model. Um and in 00:02:55.960 --> 00:02:58.677 the rejection of that, we fully aligned 00:02:58.760 --> 00:03:01.557 with our partners. Now, we To do that, 00:03:01.640 --> 00:03:03.397 as we went through the years, we built 00:03:03.480 --> 00:03:07.477 we built PG, Foundry, Gaia, Maven, 00:03:07.560 --> 00:03:11.437 Ontology, AIP. Now, we're taking the AIP 00:03:11.520 --> 00:03:15.637 stack and extending it for sovereign AI, which requires us to be able to 00:03:15.720 --> 00:03:17.677 orchestrate and fine-tune models to 00:03:17.760 --> 00:03:20.317 provide a completely sovereign stack to 00:03:20.400 --> 00:03:22.677 our partners. But what What is the 00:03:22.760 --> 00:03:25.197 philosophical importance of that? We are 00:03:25.280 --> 00:03:28.237 offering a present that 00:03:28.320 --> 00:03:30.717 that augurs to a future that we want to 00:03:30.800 --> 00:03:32.517 live in. What does that future look 00:03:32.600 --> 00:03:35.357 like? We have more rights in the PG 00:03:35.440 --> 00:03:37.957 frame, there's it's safer. In the 00:03:38.040 --> 00:03:42.957 sovereign AI frame, which is arguably by far the most important because, you 00:03:43.040 --> 00:03:46.277 know, all these other things are downstream from GDP growth and GDP 00:03:46.360 --> 00:03:50.437 health. And what if we transform America into the only democracy that actually 00:03:50.520 --> 00:03:53.957 grows, where the where production is more efficient and manufacturing 00:03:54.040 --> 00:03:56.357 actually happens. What is that frame? 00:03:56.440 --> 00:03:59.557 That frame is not uh you are going to 00:03:59.640 --> 00:04:02.517 buy into a future where you have no job, 00:04:02.600 --> 00:04:07.197 where adversaries win, and everybody who does win is a small tiny group of people 00:04:07.280 --> 00:04:12.037 living in a tiny place that somehow believes because they eat vegetables and 00:04:12.120 --> 00:04:15.437 they don't support war fighters that they deserve to have the total means of 00:04:15.520 --> 00:04:17.517 production of this country, uh and the 00:04:17.600 --> 00:04:22.517 rest of us should just sit by back and absorb the cost of that revolution, 00:04:22.600 --> 00:04:25.197 which we're paying for. How are we paying for it? 00:04:25.280 --> 00:04:27.717 In the in the enterprise context, people 00:04:27.800 --> 00:04:31.037 sign up for token uh self-pleasurings, 00:04:31.120 --> 00:04:33.757 and those that and that at a real cost 00:04:33.840 --> 00:04:36.237 like other forms of self-pleasuring. 00:04:36.320 --> 00:04:38.317 Where you are paying for the right for 00:04:38.400 --> 00:04:43.797 them to migrate your IP, your know-how, your expertise to their model so that 00:04:43.880 --> 00:04:47.277 they can build a competitive business that doesn't require your business, your 00:04:47.360 --> 00:04:50.717 people. And why are they doing it? It's actually being done for what they 00:04:50.800 --> 00:04:54.797 believe are moral reasons. They are superior to you. They deserve to 00:04:54.880 --> 00:04:59.717 colonize your enterprise. You deserve to be colonized. And in Palantir, in A, 00:04:59.800 --> 00:05:01.237 it's interesting, it doesn't work as 00:05:01.320 --> 00:05:04.277 well or as efficient purely on the alpha 00:05:04.360 --> 00:05:08.757 side as having an application layer owning your compute witness we built a 00:05:08.840 --> 00:05:13.237 partnership with Nvidia. We're expanding our application layer. We are going to 00:05:13.320 --> 00:05:14.957 enter the market and already entering in 00:05:15.040 --> 00:05:17.277 the classified space as Sham alluded to 00:05:17.360 --> 00:05:20.557 of of fine-tuning models. So, the models 00:05:20.640 --> 00:05:23.037 actually fine-tuned by us in our 00:05:23.120 --> 00:05:27.797 enterprise on Nvidia stack outperform frontier models. And you own the 00:05:27.880 --> 00:05:31.677 weights. You own the alpha. You own everything. Every single enterprise in 00:05:31.760 --> 00:05:35.397 this country is going to either look at doing this, find ways of doing it, or at 00:05:35.480 --> 00:05:37.397 least avoid the alternative of 00:05:37.480 --> 00:05:42.637 unprotected interaction with frontier models. This is very dangerous. You're 00:05:42.720 --> 00:05:45.717 warned about this in high school. And now you're learning about it in your 00:05:45.800 --> 00:05:49.597 enterprise. And in Palantir, we are at the front line we are in the front of 00:05:49.680 --> 00:05:54.037 driving this revolution. I am driving the business to grow at a 00:05:54.120 --> 00:05:57.117 rate equal or above to what we have in US 00:05:57.200 --> 00:05:59.317 commercial for the next 18 months, which 00:05:59.400 --> 00:06:04.477 is a very high goal, but it is one we can actually get to because we are fully 00:06:04.560 --> 00:06:07.837 aligned with what's right and what's good and what actually works well in an 00:06:07.920 --> 00:06:09.837 enterprise. And for the first time, 00:06:09.920 --> 00:06:11.637 people believe us. And if you didn't 00:06:11.720 --> 00:06:14.877 believe us, you can believe 149% growth 00:06:14.960 --> 00:06:17.677 in the US, a rule of of of of 40 that's 00:06:17.760 --> 00:06:22.997 155, 93% aggregate growth, and 90% growth 00:06:23.080 --> 00:06:27.597 uh in US com with 62-63% 00:06:27.680 --> 00:06:31.677 uh free cash flow margins. People thought we wouldn't be profitable. 00:06:31.760 --> 00:06:35.917 So, this is one of the more exciting times to be at Palantir. It's one of the 00:06:36.000 --> 00:06:39.077 more exciting times to participate in Palantir. And for everyone on the 00:06:39.160 --> 00:06:42.477 sidelines, you got to get off the sidelines. This is a revolution that 00:06:42.560 --> 00:06:46.637 will affect the sovereign revolution, where you stand in it, will affect your 00:06:46.720 --> 00:06:48.717 livelihood, the livelihood of the people 00:06:48.800 --> 00:06:50.637 you care, and whether America and the 00:06:50.720 --> 00:06:54.197 West win. We cannot regress to a thin 00:06:54.280 --> 00:06:58.597 philosophical model where only a small group of people who think very 00:06:58.680 --> 00:07:03.357 differently than most of us actually absorb all of the revenue and value in 00:07:03.440 --> 00:07:06.677 our business and transfer all of the dangers to us. 00:07:06.760 --> 00:07:10.437 And that's what this revolution is about, and it's extre- extremely 00:07:10.520 --> 00:07:12.757 motivating for those of us at Palantir. 00:07:12.840 --> 00:07:16.117 Thank you. >> Thanks, Alex. Our first question is from 00:07:16.200 --> 00:07:19.277 Dan with Yorkville Ives. Dan, please turn on your camera, and then you'll 00:07:19.360 --> 00:07:22.840 receive a prompt to unmute your line. 00:07:30.040 --> 00:07:33.437 >> Where's the prompt? >> [clears throat] 00:07:33.520 --> 00:07:37.197 >> Well, while we're waiting, you can be entertained. 00:07:37.280 --> 00:07:40.637 >> Maybe we can circle back to Dan. Um our next question is from Mariana with Bank 00:07:40.720 --> 00:07:44.037 of America. Mariana, please turn on your camera and then you'll receive a prompt 00:07:44.120 --> 00:07:46.960 to unmute your line. 00:07:48.600 --> 00:07:51.357 >> I think I'm good. Um, thank you. Can Can 00:07:51.440 --> 00:07:56.117 Can you walk through this at the the sovereignty boot camp? 00:07:56.200 --> 00:08:00.917 What did you learn with talking to customers that maybe 00:08:01.000 --> 00:08:04.637 you weren't expecting? I mean, obviously just an overwhelming sort of group of 00:08:04.720 --> 00:08:07.517 executives. Can you just talk about that, Alex? 00:08:07.600 --> 00:08:11.637 >> Well, first of all, yeah, for those of you not in the know, we we we did a 00:08:11.720 --> 00:08:16.157 sovereign boot camp after this kind of revolution exploded. Um, and just to 00:08:16.240 --> 00:08:20.557 give the backdrop here, you know, 2 years ago, we were said we we spent four 00:08:20.640 --> 00:08:25.477 or five months, Sasha who runs this, organizing AIP Con. This was much more 00:08:25.560 --> 00:08:29.557 like, "Hey, let's invite our buddies over to lunch." And then all of a sudden 00:08:29.640 --> 00:08:33.317 we start getting bombarded by people, people we invited, people we didn't 00:08:33.400 --> 00:08:37.277 invite, all levels of the business. So, like when you're working in enterprise, 00:08:37.360 --> 00:08:40.517 it's really important that the operational people are interested. So, 00:08:40.600 --> 00:08:43.877 it was like CEOs, operational leads. Um, 00:08:43.960 --> 00:08:47.677 and there's a a huge educational component and Cham talks about this a 00:08:47.760 --> 00:08:52.237 lot. Like we people understand that they need a way of controlling their alpha. 00:08:52.320 --> 00:08:55.877 They understand broadly that token maxing is at their own cost and they 00:08:55.960 --> 00:08:59.717 certainly understand that token maxing is leading to them transferring their 00:08:59.800 --> 00:09:03.077 data, their prompts, the way they run their business, their expertise to a 00:09:03.160 --> 00:09:07.717 third party. Um, but they don't they they need education on what they can do 00:09:07.800 --> 00:09:11.357 about that with us, preferably, but also without us. How do you do contracts? How 00:09:11.440 --> 00:09:14.397 do you How do you work with How do you work with open weight models? How do you 00:09:14.480 --> 00:09:18.597 work with closed weight models? How does this work in ontology? Is Is ontology 00:09:18.680 --> 00:09:22.277 the protective layer that they that they they've been told? Does it create value 00:09:22.360 --> 00:09:25.557 the way they How would this look in their own business? And how would they 00:09:25.640 --> 00:09:28.997 would they How would they work with the compute stack? And so, there's a just 00:09:29.080 --> 00:09:31.277 this massive demand and we're in the 00:09:31.360 --> 00:09:33.597 business of educating people, both our 00:09:33.680 --> 00:09:38.437 customers and others. But it was a super heterodox group of people both in terms 00:09:38.520 --> 00:09:42.757 of the kind of people showed up, the demand for it, and also people we've not 00:09:42.840 --> 00:09:46.997 worked for. By the way, you know, one of the reasons the NDR number is so strong, 00:09:47.080 --> 00:09:50.157 it's ridiculously strong. I mean, that people always write these things about 00:09:50.240 --> 00:09:54.637 customer adoption and maybe, you know, but the NDR number is anomalously 00:09:54.720 --> 00:09:58.477 strong. It it That is also going to shift in a even as 00:09:58.560 --> 00:10:02.237 as hard as it is to believe, become even more positive because some of our older 00:10:02.320 --> 00:10:05.757 partners we haven't really interacted with, they also showed up. 00:10:05.840 --> 00:10:09.637 They're like, "Okay, now we now we get why we would need you." And not just 00:10:09.720 --> 00:10:12.997 Foundry, they're migrating across our stack. So, customers that were only 00:10:13.080 --> 00:10:16.197 using Foundry now want Ontology, now want to be part of the Sovereign AI 00:10:16.280 --> 00:10:20.357 stack. I would say last not least, you're asking the internal version, 00:10:20.440 --> 00:10:26.117 external version, internally, recruiting, retention, excitement at Palantir. 00:10:26.200 --> 00:10:30.237 It's it's I mean, I'm very excited. I think the people 00:10:30.320 --> 00:10:32.397 around this table are very excited. Uh 00:10:32.480 --> 00:10:35.237 the the legal department is excited. 00:10:35.320 --> 00:10:39.397 This I mean, that hasn't happened since like that doesn't happen. Like, they're 00:10:39.480 --> 00:10:42.597 out like "Rock on." >> [laughter] >> Uh 00:10:42.680 --> 00:10:47.517 so, it's just it's more fun. You know? Uh so, yeah, there's just it 00:10:47.600 --> 00:10:51.757 that that's the way this went down. And I think this is going to this is the way 00:10:51.840 --> 00:10:55.597 beginning. So, you the way the way at least we think about this internally is 00:10:55.680 --> 00:10:59.357 what portion of the market is available to people who want to create value, mean 00:10:59.440 --> 00:11:02.957 create value and keep it. That portion of the market has gone from a small 00:11:03.040 --> 00:11:05.117 portion of the market where we are we 00:11:05.200 --> 00:11:09.997 we're doing where till like a large large part of the US GDP. And that's why 00:11:10.080 --> 00:11:12.757 we need partners. Behind the scenes now we're trying to 00:11:12.840 --> 00:11:17.357 find partners. Now, partners we need technically exceedingly competent 00:11:17.440 --> 00:11:20.197 partners. Partners doesn't mean have We don't have to agree on every 00:11:20.280 --> 00:11:23.677 issue. We don't even have to agree on every client. They can occasionally 00:11:23.760 --> 00:11:28.317 compete against us. We've seen this in the defense tech stuff. Stack and Sham, 00:11:28.400 --> 00:11:32.397 you know, like the approach to defense tech where we partner with people. 00:11:32.480 --> 00:11:35.957 Partner doesn't mean we agree, sometimes we compete, but it means we're marching 00:11:36.040 --> 00:11:39.197 in a similar direction that allows us to scale. So, those things are going to be 00:11:39.280 --> 00:11:41.877 a very big important part of like 00:11:41.960 --> 00:11:46.717 essentially the guide. Why am I pushing the company to to grow not to the just 00:11:46.800 --> 00:11:48.637 to the end of the year, but next year? 00:11:48.720 --> 00:11:51.317 It's because that also forces us to to 00:11:51.400 --> 00:11:56.240 find ways to scale to meet the demand that's out there. 00:11:57.120 --> 00:12:00.877 >> Thanks, Alex. Our next question is from Mariana with Bank of America. Mariana, 00:12:00.960 --> 00:12:03.317 please turn on your camera and then you'll receive a prompt to unmute your 00:12:03.400 --> 00:12:08.037 line. >> Good afternoon, everyone. 00:12:08.120 --> 00:12:09.277 >> Hello. 00:12:09.360 --> 00:12:13.277 >> So, a follow-up to to Dan's question. 00:12:13.360 --> 00:12:16.997 When all this AI revolution started, right? It was really clear for 00:12:17.080 --> 00:12:21.357 enterprises that data and proprietary data and how you train your models was 00:12:21.440 --> 00:12:23.837 going to be the key. But then, we're 00:12:23.920 --> 00:12:28.637 like 3 years into that revolution and now everyone started to realize, oh, 00:12:28.720 --> 00:12:32.877 owning my data and what I really learn from my data is important. What happened 00:12:32.960 --> 00:12:37.197 there? Why you think that you positioned yourselves 00:12:37.280 --> 00:12:40.597 back then in a different approach to AI that 00:12:40.680 --> 00:12:43.197 enterprises weren't able to see? And why 00:12:43.280 --> 00:12:48.117 these numbers that reflect that you're the winner of AI today, 00:12:48.200 --> 00:12:53.160 it wasn't that clear for any other software application back then. 00:12:53.400 --> 00:12:56.917 >> Well, you can kind of divide it up into two parts. There's the first part is 00:12:57.000 --> 00:13:00.517 efficacy, you know, and then the second part is efficiency. That's not the quite 00:13:00.600 --> 00:13:04.117 the right word, but it's like how do you scale it? In the zero to one phase, it's 00:13:04.200 --> 00:13:06.957 much more important to focus on the application layer. Like, okay, this new 00:13:07.040 --> 00:13:12.277 thing has happened. How do we turn that into economic value? And then as people 00:13:12.360 --> 00:13:15.277 started to experience the economic value and as time started happening, you 00:13:15.360 --> 00:13:18.637 started to see that some of the people who were partners out there were 00:13:18.720 --> 00:13:21.797 building things that were competitive to you. And I think that took some time to 00:13:21.880 --> 00:13:26.677 kind of seep into the psyche and mindset of, wait a second, this is maybe not 00:13:26.760 --> 00:13:29.957 you know, now I know this thing is valuable in the right hands on the right 00:13:30.040 --> 00:13:34.357 platform, but I also am going to need to control the weights. That the alpha that 00:13:34.440 --> 00:13:36.877 is being generated isn't simply the data 00:13:36.960 --> 00:13:41.437 that's resident in my enterprise. It's also the metadata, the reasoning traces, 00:13:41.520 --> 00:13:46.117 the exhaust, the usage of this, which I don't yet have mechanisms to control. 00:13:46.200 --> 00:13:48.877 And now I understand that's actually probably more valuable than just the 00:13:48.960 --> 00:13:52.997 data in my enterprise. And that's been a clarion call, I think, for for 00:13:53.080 --> 00:13:56.557 the market over the last quarter, two quarters. 00:13:56.640 --> 00:13:59.677 >> There's the there's also the clash 00:13:59.760 --> 00:14:01.757 question implicit in your question is, 00:14:01.840 --> 00:14:04.477 you know, why did we get this right? 00:14:04.560 --> 00:14:06.877 And, you know, again, I think it's we 00:14:06.960 --> 00:14:11.877 are actually fully aligned with our partners. Sometimes we make decisions 00:14:11.960 --> 00:14:15.157 like, you know, that are against our economic interests. Like, we're 00:14:15.240 --> 00:14:19.477 supporting lots of institutions in Europe. The growth sucks. 00:14:19.560 --> 00:14:23.557 I mean, clandestine institutions. But, without our products, they would have 00:14:23.640 --> 00:14:27.237 rampant terrorism and their problems, their migration problems, and the 00:14:27.320 --> 00:14:30.957 results would be 10 times worse. It's not actually in our economic interest 00:14:31.040 --> 00:14:32.997 anymore to do this, but we still do it. 00:14:33.080 --> 00:14:36.877 And this is because we actually are believers, for better or worse. 00:14:36.960 --> 00:14:39.557 I would say also, um, this is a company 00:14:39.640 --> 00:14:42.157 that from beginning, from inception, has 00:14:42.240 --> 00:14:45.397 valued artistic insights. Meaning, you 00:14:45.480 --> 00:14:49.517 can't model something purely on science. You have to have an 00:14:49.600 --> 00:14:53.997 aesthetic or artistic appreciation for it. And we've made huge bets, everyone 00:14:54.080 --> 00:14:58.437 at this sitting at this table, and many, many hundreds of people at Palantir have 00:14:58.520 --> 00:15:02.477 essentially artistic insights. We've always viewed ourselves, we've always 00:15:02.560 --> 00:15:05.797 said we're a colony of artists, and people assume that means we're just 00:15:05.880 --> 00:15:08.157 difficult. And that's also true. But, 00:15:08.240 --> 00:15:10.517 you know, it's it's also we value 00:15:10.600 --> 00:15:13.037 insights that are way before anyone else 00:15:13.120 --> 00:15:17.997 would see them. And we build major parts of our business. That's very hard for 00:15:18.080 --> 00:15:21.517 normal business to do because, I mean, one of the jokes running around Palantir 00:15:21.600 --> 00:15:25.437 is we can we can definitely meet our guide next year if we get paid for all 00:15:25.520 --> 00:15:29.717 the people copying us. You know, so even we got a small portion of people copy 00:15:29.800 --> 00:15:33.757 you know of the of the of the of the funds for the copyers, we do very well. 00:15:33.840 --> 00:15:37.717 Normal businesses, again, no critique here, are built around there was a 00:15:37.800 --> 00:15:40.317 playbook, we execute on that playbook, 00:15:40.400 --> 00:15:42.837 and we're in a non-playbook 00:15:42.920 --> 00:15:46.117 world. So, executing on a playbook that 00:15:46.200 --> 00:15:51.397 worked 5 10 years ago essentially build parasitic software and monetize it, that 00:15:51.480 --> 00:15:54.397 doesn't work now. And there's hundreds of variants of that, but that is a 00:15:54.480 --> 00:15:56.477 central advantage we have. We are a 00:15:56.560 --> 00:15:59.517 colony of believers and artists that are 00:15:59.600 --> 00:16:04.437 very motivated to drive value. And that sets us aside much more than I would 00:16:04.520 --> 00:16:08.277 have imagined 10 years ago if you'd asked the same question. And it and then 00:16:08.360 --> 00:16:12.517 luckily for us, this is capitalism. I mean, DSA wants to get rid of it, but 00:16:12.600 --> 00:16:16.677 until they do, you got to look at the results. Because if our results and and 00:16:16.760 --> 00:16:20.557 the other thing I'll tell you that's very special for us, we are outsiders. 00:16:20.640 --> 00:16:24.597 Outsiders, like you're an outsider. You come to this country, you better have 00:16:24.680 --> 00:16:28.117 really good results. You know, and like the same thing with 00:16:28.200 --> 00:16:31.517 Palantir, we know we need the best results cuz people aren't buying our 00:16:31.600 --> 00:16:34.957 product because we're swinging the golf club correctly or paying for the steak 00:16:35.040 --> 00:16:38.077 dinner. They don't even invite us to steak dinners. 00:16:38.160 --> 00:16:41.077 So, that's and that outsider status has 00:16:41.160 --> 00:16:43.277 caused huge problems 00:16:43.360 --> 00:16:46.597 but in the first 18 years, but a lot of 00:16:46.680 --> 00:16:49.997 benefits in the next 18 years. 00:16:50.080 --> 00:16:52.877 And other people don't like being outsiders. I'm I'm You know, in fact, 00:16:52.960 --> 00:16:56.880 I'm struggling with our current popularity. 00:16:58.200 --> 00:17:01.877 >> Thank you. Our next question is from Gil with DA Davidson. Gil, please turn on 00:17:01.960 --> 00:17:06.240 your camera and then you'll receive a prompt to unmute your line. 00:17:06.640 --> 00:17:10.957 >> Thank you, Ryan. The topic of sovereignty, you focused a lot on how 00:17:11.040 --> 00:17:15.197 dangerous it is to give the keys to the labs because they could choose to 00:17:15.280 --> 00:17:20.037 compete with you. Is there another aspect of this as well though that if 00:17:20.120 --> 00:17:24.277 you choose a lab and you buy the orchestration and the consulting and the 00:17:24.360 --> 00:17:28.317 harnesses from them, you're beholden to their models and if 00:17:28.400 --> 00:17:30.677 something happens, if it's not the best 00:17:30.760 --> 00:17:33.237 model anymore, if the model gets pulled, 00:17:33.320 --> 00:17:36.237 then you as a customer are stuck and you 00:17:36.320 --> 00:17:40.757 may have a mission critical system fail and this isn't hypothetical. This 00:17:40.840 --> 00:17:44.517 happened a couple of times this year where one of the frontier models got 00:17:44.600 --> 00:17:48.037 pulled by either the government or or the company. Well, if you work with 00:17:48.120 --> 00:17:52.397 Palantir, it I would assume that when that happens, you can go to your 00:17:52.480 --> 00:17:56.277 customer and say, "Hey, if that model doesn't work, I can plug in another 00:17:56.360 --> 00:17:59.277 model for you." >> Now, I mean, I'll let Ryan and Shyam 00:17:59.360 --> 00:18:03.277 comment here, but we're already doing that across the USG, US government. We 00:18:03.360 --> 00:18:07.517 have a product that allows you to about switch out models. I've Look, at the end 00:18:07.600 --> 00:18:11.277 of the day, if if you are locked into a product, you're going to I mean, that's 00:18:11.360 --> 00:18:15.197 the nature of monopoly capitalism. That that's why people want the lock-in cuz 00:18:15.280 --> 00:18:18.397 then they can raise the prices and reduce the quality. We're against that 00:18:18.480 --> 00:18:21.277 because we're on the side of the American worker, the American people and 00:18:21.360 --> 00:18:24.957 its great institutions and other institutions across the West. Um but 00:18:25.040 --> 00:18:28.997 yes, there you know, and enterpr- people are running big enterprises in this 00:18:29.080 --> 00:18:30.717 country are very sophisticated. They're 00:18:30.800 --> 00:18:32.997 aware of these risks and they don't like 00:18:33.080 --> 00:18:37.037 people who are kind of flaunting like 00:18:37.120 --> 00:18:40.197 setting this up so that they feel they're being made fun of or they're 00:18:40.280 --> 00:18:44.277 being they're sucker, basically. So, there's a lot Honestly, there's a lot of 00:18:44.360 --> 00:18:48.757 anger here. I spend interestingly a fair bit of my time trying to explain to 00:18:48.840 --> 00:18:51.557 people that, you know, some of the people involved in these things are not 00:18:51.640 --> 00:18:54.317 the caricatures they think they are, but 00:18:54.400 --> 00:18:58.997 because the business setup looks like heads I win, tails I win and American 00:18:59.080 --> 00:19:00.997 business people don't like that. 00:19:01.080 --> 00:19:03.637 But then, I don't know, there's a lot of things on the contractual front that 00:19:03.720 --> 00:19:05.357 would that would support that. 00:19:05.440 --> 00:19:07.437 >> Yeah, I would say like that's, you know, 00:19:07.520 --> 00:19:11.637 our whole focus is converting tokens to value. The example Shyam gave, 00:19:11.720 --> 00:19:14.157 uh that's happening across the board in the conversations with customers. That's 00:19:14.240 --> 00:19:18.197 why they're looking to expand to convert how they position themselves in the 00:19:18.280 --> 00:19:22.797 industry with us in their industry. And so, we're seeing like extreme alignment. 00:19:22.880 --> 00:19:25.597 It's not about being beholden to one model, it's about being bringing the 00:19:25.680 --> 00:19:27.917 right models to bear for the right purposes. And our contracts, our 00:19:28.000 --> 00:19:31.797 structures are set up to do that to support and compound their alpha in the 00:19:31.880 --> 00:19:34.077 organization. >> I think you know, I mentioned it 00:19:34.160 --> 00:19:37.517 earlier, but you're going to see this like we've been beholden to a small 00:19:37.600 --> 00:19:41.437 number of benchmarks that people have been designing models to and then 00:19:41.520 --> 00:19:44.517 releasing models saying like look how well it does on this benchmark. But the 00:19:44.600 --> 00:19:48.477 benchmark has actually almost nothing to do with your business. Well, so how do 00:19:48.560 --> 00:19:51.317 you figure out how to make the benchmark that represents your reality, what 00:19:51.400 --> 00:19:54.197 you're trying to succeed at, what you're trying to get better at, and then see 00:19:54.280 --> 00:19:58.077 what model makes sense. And the natural consequence of doing that, even even 00:19:58.160 --> 00:20:00.797 leaving aside all of your other arguments for sovereignty, it's like oh, 00:20:00.880 --> 00:20:04.917 how do I how do I climb that hill? How do I figure out what it is that I do as 00:20:05.000 --> 00:20:08.557 a business that I feed back into weights that I can control, which presupposes an 00:20:08.640 --> 00:20:11.957 open model and sovereignty. And then you're not just going to do that where 00:20:12.040 --> 00:20:15.437 you're waiting, hey, maybe I'll switch it out when this model gets pulled, when 00:20:15.520 --> 00:20:18.117 the rug gets pulled out from underneath me. You're actually going to be 00:20:18.200 --> 00:20:21.557 leveraging the automation to do that constantly to figure out when you have a 00:20:21.640 --> 00:20:24.957 next checkpoint that you can pivot maybe there's a new model. I I mentioned the 00:20:25.040 --> 00:20:30.277 example I I literally almost felt gaslit when within 24 hours of getting Nemotron 00:20:30.360 --> 00:20:34.557 up with no post training. This is vanilla Nemotron Ultra. It did better 00:20:34.640 --> 00:20:37.557 than Frontier. You know, if you just looked at the numbers, you would say, 00:20:37.640 --> 00:20:38.677 well, it's nowhere near the Frontier. 00:20:38.760 --> 00:20:41.957 That shouldn't even be possible. But of course, the benchmarks are wrong. I 00:20:42.040 --> 00:20:44.557 mean, the benchmarks are right for what the benchmark's measuring, but that's 00:20:44.640 --> 00:20:47.597 not my business. Those are not the tasks my customers had that they were trying 00:20:47.680 --> 00:20:51.637 to solve. And so then moving this to an empirical basis is I think is is how 00:20:51.720 --> 00:20:57.480 we're going to accelerate the realization of tokens to real economic value. 00:20:57.800 --> 00:21:01.237 >> Thank you. Alex, as always, we have a lot of individual investors on the line. 00:21:01.320 --> 00:21:03.557 Is there anything you'd like to say before we end the call? 00:21:03.640 --> 00:21:06.677 >> Well, you know, uh Uh, your support has 00:21:06.760 --> 00:21:10.437 was crucial to getting us this far and 00:21:10.520 --> 00:21:14.157 uh, you know, crucial to getting us to where we're going to go, which is a much 00:21:14.240 --> 00:21:16.757 much much much larger company. 00:21:16.840 --> 00:21:19.317 Um, this is one of the most exciting times 00:21:19.400 --> 00:21:21.477 to be involved in the Palantir mission. 00:21:21.560 --> 00:21:25.797 Um, we are going to help transform especially this country but allied 00:21:25.880 --> 00:21:29.197 countries both in commercial and government and 00:21:29.280 --> 00:21:34.877 uh, this sovereign frame that is we're using as our organizational principle is 00:21:34.960 --> 00:21:36.197 the one that is 00:21:36.280 --> 00:21:38.877 inclusive of everybody who wants to have 00:21:38.960 --> 00:21:44.117 a better world today and tomorrow and we invite everyone to engage with it in 00:21:44.200 --> 00:21:48.240 some form and thank you.