WEBVTT 00:00:00.000 --> 00:00:04.077 I'm here with Dr. Alex Karp, CEO and co-founder of Palantir. Thank you for 00:00:04.160 --> 00:00:06.997 taking the time to do this. >> By the way, I always say this, it's you 00:00:07.080 --> 00:00:10.757 know, you were you we we've been together since the beginning of our life 00:00:10.840 --> 00:00:13.717 in the public markets. Um 00:00:13.800 --> 00:00:17.677 when we first uh DPO'd, um which is why 00:00:17.760 --> 00:00:21.397 so many people in America and across the world got to participate. But, you know, 00:00:21.480 --> 00:00:25.677 without you, you know, sending out missives about how we were 00:00:25.760 --> 00:00:28.997 misunderstood, you know, probably millions and millions 00:00:29.080 --> 00:00:31.237 of people would not have participated. 00:00:31.320 --> 00:00:36.117 Um and yeah, it's Palantir's this kind of crazy walled poster child company 00:00:36.200 --> 00:00:39.237 where people can only understand the Palantir 00:00:39.320 --> 00:00:42.757 of 5 years ago, not the Palantir that were we have. 00:00:42.840 --> 00:00:47.197 And so, you know, there was a lot of work to do on the, you know, front line. 00:00:47.280 --> 00:00:48.877 I'm happy to have you with us. 00:00:48.960 --> 00:00:53.277 >> So, 6 years ago, you went on CNBC the day you guys did DPO, and I think a lot 00:00:53.360 --> 00:00:55.757 of people didn't really understand what you were saying. 00:00:55.840 --> 00:00:57.677 About 2 weeks ago, you also went on 00:00:57.760 --> 00:01:00.477 CNBC, and I think you changed in about 00:01:00.560 --> 00:01:05.117 20 minutes the entire valley's approach to thinking about front-end models 00:01:05.200 --> 00:01:09.677 versus open-source models. So, quite simply, why is open source so important 00:01:09.760 --> 00:01:12.397 for Palantir? >> Well, it 00:01:12.480 --> 00:01:14.077 what happened, by the way, that was like 00:01:14.160 --> 00:01:16.837 it was a revolution. It began with like 00:01:16.920 --> 00:01:19.677 work we were doing with Nvidia, and then 00:01:19.760 --> 00:01:23.277 going on um on on on television like 00:01:23.360 --> 00:01:27.077 that, it just exploded in a way no one could 00:01:27.160 --> 00:01:31.797 have really predicted. I think the explosion happened because 00:01:31.880 --> 00:01:35.997 you know, people are being sold this thing they know isn't quite true, like 00:01:36.080 --> 00:01:39.677 you know, and then then they paid a price. You're paying, you know, a lot of 00:01:39.760 --> 00:01:43.037 money if you're enterprise for for tokens that are 00:01:43.120 --> 00:01:45.077 really it's hard to see the value. And 00:01:45.160 --> 00:01:47.917 then you realize at the same time the 00:01:48.000 --> 00:01:50.317 alpha of your business is migrating to 00:01:50.400 --> 00:01:52.277 the person you're paying an absurd sum 00:01:52.360 --> 00:01:55.477 of money based on value. And then, last 00:01:55.560 --> 00:02:00.357 and least, they're just behaving in the public sphere in a way that is blowing 00:02:00.440 --> 00:02:04.637 up the politics of America, making people 00:02:04.720 --> 00:02:07.077 really afraid, leading to people just 00:02:07.160 --> 00:02:11.917 wanting to pillory wealthy people but not help poor people. And you're sitting 00:02:12.000 --> 00:02:15.397 there and saying you're basically like, "What is going on here?" 00:02:15.480 --> 00:02:18.477 >> Right. >> So, you have that backdrop. Then, you 00:02:18.560 --> 00:02:21.717 have And by the way, part of the reason for that backdrop, just to make this a 00:02:21.800 --> 00:02:24.317 little longer, is these models are 00:02:24.400 --> 00:02:26.717 excruciatingly effective 00:02:26.800 --> 00:02:29.397 when used with an application layer. 00:02:29.480 --> 00:02:32.397 Now, they all know that. When you use it 00:02:32.480 --> 00:02:35.357 in application layer installed 00:02:35.440 --> 00:02:39.757 with what everyone's now copying from us called FDAs for X. So, 00:02:39.840 --> 00:02:43.117 then I think part of the reason this revolution actually happened is 00:02:43.200 --> 00:02:47.877 technical participants in the market know one thing. 00:02:47.960 --> 00:02:51.677 Palantir gets it right. You think they're crazy. You think 00:02:51.760 --> 00:02:54.757 they're insane. Carp is obnoxious. 00:02:54.840 --> 00:02:56.757 And then it turns out you have to copy him. 00:02:56.840 --> 00:03:01.077 Right? So, now, investors who are kind of just chillaxing in front of their 00:03:01.160 --> 00:03:05.437 screen producing AI slop on their computer and thinking I'm going to get 00:03:05.520 --> 00:03:09.037 rich and yeah, maybe core people, no one else is going to get rich. They're like, 00:03:09.120 --> 00:03:11.797 "Yeah, but that's just how the world works. You're good for me." They've been 00:03:11.880 --> 00:03:16.157 a little bit tone deaf to the technical thing. So, they're very happy to watch 00:03:16.240 --> 00:03:19.597 the the the like the frontier. By the way, I always feel like I have to say 00:03:19.680 --> 00:03:22.157 this. You know, >> Yeah, people think you hate the frontier 00:03:22.240 --> 00:03:23.677 laughs. I don't think that's your point. 00:03:23.760 --> 00:03:26.317 I think your argument is >> Well, people actually think I'm I'm 00:03:26.400 --> 00:03:30.477 going I'm like I'm kind of not being like I spend 00:03:30.560 --> 00:03:34.797 more time defending Dario and more effectively, honestly, cuz they don't 00:03:34.880 --> 00:03:38.877 even talk to like if you're telling them the truth about what you think, they 00:03:38.960 --> 00:03:43.077 don't talk to you, apparently. Because like the views there you know, and like 00:03:43.160 --> 00:03:46.797 I'm always trying to explain to you, you know, it's a worldview. This is a 00:03:46.880 --> 00:03:49.117 essentially a world historic figure. He 00:03:49.200 --> 00:03:53.397 He really did take the LM revolution, 00:03:53.480 --> 00:03:57.397 bring it to enterprise in a way where everyone found out something is going to 00:03:57.480 --> 00:04:01.917 be real. Now, of course, they don't talk very much about almost all the values 00:04:02.000 --> 00:04:05.837 being created in ontology with their models where they're doing good. 00:04:05.920 --> 00:04:09.957 >> Um so let's get into that discussion because 2 years ago you made it very 00:04:10.040 --> 00:04:13.637 clear to everybody that the models would be commoditized. The frontier labs have 00:04:13.720 --> 00:04:16.397 gotten better and better and better, but with the rise of open source, which I 00:04:16.480 --> 00:04:19.117 think is kind of in your point and video's point, the cost of tokens is 00:04:19.200 --> 00:04:22.677 coming down. Which means the application layer, something that processes, 00:04:22.760 --> 00:04:25.077 orchestrates, governs the tokens is becoming more powerful. 00:04:25.160 --> 00:04:29.517 >> You know, you would need compute so that Nvidia primarily 00:04:29.600 --> 00:04:32.957 um and we're very strongly partnered with them. You you would need an 00:04:33.040 --> 00:04:36.797 application layer because that while the and you need a way of installing them. 00:04:36.880 --> 00:04:40.637 So, you need FTEs, you need an application layer, people generally call 00:04:40.720 --> 00:04:43.557 that an ontology now after maligning it everywhere 00:04:43.640 --> 00:04:47.837 uh because they you need to do it's and there's two points. Simply put, there's 00:04:47.920 --> 00:04:51.357 safety. So, you know, we've now we're now helping 00:04:51.440 --> 00:04:54.597 people and I think people are aware of the dangers of just, you know, like 00:04:54.680 --> 00:04:58.317 look, they you know, a lot of our clients are running around saying no no 00:04:58.400 --> 00:05:02.877 ontology, no love meaning no glove, no love. So, like it's probably not exactly 00:05:02.960 --> 00:05:05.397 what I can say on TV, but >> It's a famous last word. 00:05:05.480 --> 00:05:08.917 >> It's like they're like, yeah, they're like, "Hey, look, no ontology, no love." 00:05:09.000 --> 00:05:12.637 It's like it's like it's like gone it's gone viral in corporate America. It's 00:05:12.720 --> 00:05:16.397 like it's crazy. It's like it's so there's like but there's a safety 00:05:16.480 --> 00:05:20.037 function like if you're meeting someone you don't know 00:05:20.120 --> 00:05:24.997 and who may have a history of doing things and leaving certain products 00:05:25.080 --> 00:05:29.637 behind that you know, you can't get rid of, you probably be insane not to have 00:05:29.720 --> 00:05:33.517 ontology protecting you. So, that is the safety function 00:05:33.600 --> 00:05:36.877 and that's that so that that's crazy important. By the way, with models 00:05:36.960 --> 00:05:39.477 coming from the whole world, that safety function is going to be even more 00:05:39.560 --> 00:05:42.397 important. But in the end of the day, 00:05:42.480 --> 00:05:46.917 businesses need to expand their tribal knowledge, their alpha while protecting 00:05:47.000 --> 00:05:50.557 it. And so, then there's another term which is now being adopted essentially 00:05:50.640 --> 00:05:54.717 the orchestration. How do you actually deploy the models so that they can do 00:05:54.800 --> 00:05:59.437 exactly what they're capable of doing while not not providing like chillaxing 00:05:59.520 --> 00:06:03.717 swap? Uh so, you get >> Why is open source plus the ontology so 00:06:03.800 --> 00:06:07.237 important for orchestration? >> Okay. Well, within the Well, first of all, it's 00:06:07.320 --> 00:06:11.997 because that's where you get the maximum convergence between pro safety because 00:06:12.080 --> 00:06:16.917 you the open source model is is is not aspiring in any way, obviously, to 00:06:17.000 --> 00:06:20.237 export although you do have our, you know, whatever protective layer 00:06:20.320 --> 00:06:24.357 probably. So, um >> That data is not going back to Frontier Labs. 00:06:24.440 --> 00:06:28.837 >> So, it's essentially but then you also the it allows you to fine-tune the 00:06:28.920 --> 00:06:33.437 model. The fine-tuning is very hard to do without dealing with the Frontier 00:06:33.520 --> 00:06:37.237 Lab. And when you're fine-tuning, you're basically exploiting what is the alpha 00:06:37.320 --> 00:06:39.997 you're trying to create on top of the alpha that's deep in your business. What 00:06:40.080 --> 00:06:42.597 is the context of your enter >> What is the context of your enterprise? 00:06:42.680 --> 00:06:44.917 What is a specialized problem 00:06:45.000 --> 00:06:47.117 >> sets embedded in your data that you have 00:06:47.200 --> 00:06:51.757 a unique lever and insight into way of solving. So, if you're building, I don't 00:06:51.840 --> 00:06:55.757 know, you're building a drug to reduce the chance of cancer, like you are 00:06:55.840 --> 00:06:59.597 actually building it on proprietary data and proprietary insights on how the 00:06:59.680 --> 00:07:03.757 human body works. >> Right. >> Right. And that like so, or you know, in 00:07:03.840 --> 00:07:09.077 Battlefield's obvious, but every single enterprise that is actually valuable. 00:07:09.160 --> 00:07:10.477 Like how do you define an enterprise 00:07:10.560 --> 00:07:13.397 that's valued? It has a specialized way 00:07:13.480 --> 00:07:17.157 of doing something a consumer wants or 00:07:17.240 --> 00:07:21.397 in the government context has a specialized way of doing something that 00:07:21.480 --> 00:07:25.517 the end user does not want that protects your people from the end user. 00:07:25.600 --> 00:07:27.757 Like this so, that's like this is so, 00:07:27.840 --> 00:07:30.997 that that to do that most efficiently, 00:07:31.080 --> 00:07:34.717 you would need compute uh ontology both 00:07:34.800 --> 00:07:39.637 protective and alpha augmenting and creating. And you need to make sure it 00:07:39.720 --> 00:07:43.597 doesn't transfer you that own those weights yourself. So, this by the way, 00:07:43.680 --> 00:07:47.717 this is not speculative. We we've seen this already in the US government 00:07:47.800 --> 00:07:52.597 context where in context that is very very hard to use frontier labs or where 00:07:52.680 --> 00:07:54.437 there are other quite firm. I don't 00:07:54.520 --> 00:07:56.677 understand this. I'm willing to go like 00:07:56.760 --> 00:08:00.437 By the way, in what world does the 00:08:00.520 --> 00:08:04.717 Department of War, the only institution in America that is extremely high 00:08:04.800 --> 00:08:07.997 credibility, the institution that integrated before anyone else, that 00:08:08.080 --> 00:08:11.317 protects us every day, that protects, that allows us it to export our 00:08:11.400 --> 00:08:16.117 technology, in what world are we outsourcing the moral veracity of that 00:08:16.200 --> 00:08:19.917 to a bunch of people sitting very far away who do and they're even met a man 00:08:20.000 --> 00:08:23.477 or woman in uniform and it's presumably don't care if they come home alive at 00:08:23.560 --> 00:08:26.437 least in practice. So, there's also like 00:08:26.520 --> 00:08:30.997 who makes these decisions? By the way, Palantir we've always been part of the 00:08:31.080 --> 00:08:32.277 reason the sovereignty thing actually 00:08:32.360 --> 00:08:34.997 works is technical users know for 00:08:35.080 --> 00:08:39.357 whether it was PG, Foundry, Ontology, Maven, 00:08:39.440 --> 00:08:41.637 uh now sovereign AI, we've always been 00:08:41.720 --> 00:08:43.997 on the side of protecting individual 00:08:44.080 --> 00:08:46.437 rights, protecting data, and protecting 00:08:46.520 --> 00:08:49.117 by logical extension your sovereignty. 00:08:49.200 --> 00:08:51.717 So, that like we that that's actually 00:08:51.800 --> 00:08:56.597 where we should all be as a country and as a tech community. And then 00:08:56.680 --> 00:08:59.597 there's the reason though that this ended up being 00:08:59.680 --> 00:09:04.717 you know, now I should say I'm wildly happy you have a lot of people entering 00:09:04.800 --> 00:09:08.237 and taking, you know, cuz this is actually where it should be and we don't 00:09:08.320 --> 00:09:11.997 like just cuz we basically are have been at the forefront of launching many of 00:09:12.080 --> 00:09:13.477 these revolutions and it's commonly 00:09:13.560 --> 00:09:16.077 known in tech in technical circles that 00:09:16.160 --> 00:09:19.077 we do this well. Um you know, building a 00:09:19.160 --> 00:09:21.077 broad The reason you see this broad 00:09:21.160 --> 00:09:24.877 phalanx of a coalition is the customers 00:09:24.960 --> 00:09:28.837 this is what they want and need. They are livid at the option that was 00:09:28.920 --> 00:09:33.437 provided and this this way of doing things. In fact, currently the biggest 00:09:33.520 --> 00:09:36.357 problem at Palantir is just how do we scale this? 00:09:36.440 --> 00:09:39.477 >> So, so what is but there's a lot of customers here. By the way, the 00:09:39.560 --> 00:09:43.597 backstory here because of this is you get the backstory. 00:09:43.680 --> 00:09:47.877 >> Like usually it two three years ago playing AIP like >> Oh, I remember. 00:09:47.960 --> 00:09:50.357 >> Like playing AIP >> This was like, "Hey, we're doing 00:09:50.440 --> 00:09:53.797 something really cool. Let's just invite people." And 00:09:53.880 --> 00:09:56.237 I mean like but but and look at the people out there. 00:09:56.320 --> 00:09:58.197 >> Yeah. >> It's like they care. They they care. 00:09:58.280 --> 00:10:01.997 It's like they want to learn. We are actually partnering with them. They're 00:10:02.080 --> 00:10:05.917 from every part of industry. They're the among the most important industries in 00:10:06.000 --> 00:10:08.277 the world. >> So, so what has the customer respond? 00:10:08.360 --> 00:10:11.557 You're on the front lines with customers every day. What is their response been? 00:10:11.640 --> 00:10:14.517 >> I say Well, it's it's um it's it the 00:10:14.600 --> 00:10:16.437 response to I would say the open source 00:10:16.520 --> 00:10:19.837 as a part of the sovereign narrative is 00:10:19.920 --> 00:10:23.037 a colossal It is what It's not only what 00:10:23.120 --> 00:10:25.917 they want and need. It's like Imagine 00:10:26.000 --> 00:10:30.357 again if someone is you're using this thing you just have a bad feeling. 00:10:30.440 --> 00:10:34.837 And then And now like now you understand why you have this deep bad feeling and 00:10:34.920 --> 00:10:36.797 and honestly in many cases that's where 00:10:36.880 --> 00:10:40.517 I end up defending uh there's anger. So 00:10:40.600 --> 00:10:43.877 now it's like there's like a valve release. The problem we have as a 00:10:43.960 --> 00:10:48.797 company is how do we stay very focused like we're going to We just have to I 00:10:48.880 --> 00:10:51.157 get calls from all over the world. I'm like 00:10:51.240 --> 00:10:53.517 "Can you come to my office?" Like we 00:10:53.600 --> 00:10:55.837 We're just like we we're we're 00:10:55.920 --> 00:10:58.917 rebuilding internal dynamics a lot here. 00:10:59.000 --> 00:11:03.517 Uh rebuilding infrastructure, you know, this sovereign AIP like requires us to 00:11:03.600 --> 00:11:07.397 build out certain parts of mythology, to build out our ability to fine-tune 00:11:07.480 --> 00:11:10.477 models, to service a demand both in the 00:11:10.560 --> 00:11:15.557 US government and in US commercial uh that we can't really fulfill. We're also 00:11:15.640 --> 00:11:16.997 doing a lot of educational stuff. 00:11:17.080 --> 00:11:19.397 >> Yeah. >> It's Speaking of education, one thing 00:11:19.480 --> 00:11:22.157 that is I've heard a lot is a lot of these open source models are coming from 00:11:22.240 --> 00:11:25.877 China. And somehow they're like, "How is Palantir endorsing China?" Can you give 00:11:25.960 --> 00:11:29.157 a response? >> endorsing we're endorsing what what 00:11:29.240 --> 00:11:33.717 Well, first of all, we obviously are I mean very pro-American. You see we're 00:11:33.800 --> 00:11:38.477 partnered with Nvidia, a great American company. Um, the way Look, there's a 00:11:38.560 --> 00:11:41.197 there's a template of how not to do this. 00:11:41.280 --> 00:11:43.437 >> Yeah. >> It's called Europe. 00:11:43.520 --> 00:11:47.517 Europe basically tries to close its borders. I mean, there are all sorts of 00:11:47.600 --> 00:11:51.797 reasons how they It doesn't now put this way, but de facto, they're closing their 00:11:51.880 --> 00:11:55.877 board trying to close their borders to US tech, right? That The problem with 00:11:55.960 --> 00:11:59.517 that, leaving everything else aside, is they're screwing European tech. 00:11:59.600 --> 00:11:59.957 >> Mhm. 00:12:00.040 --> 00:12:03.597 >> You must be able to compete. Like, we I 00:12:03.680 --> 00:12:06.597 believe that we will we can deliver. 00:12:06.680 --> 00:12:10.557 I've seen this I in practice frontier 00:12:10.640 --> 00:12:14.317 surpassing uh alpha creation. 00:12:14.400 --> 00:12:17.717 Where we're honest about what models do and don't do safely. 00:12:17.800 --> 00:12:21.077 >> Correct. >> Right? And that's being done with a completely American stack. 00:12:21.160 --> 00:12:23.877 >> Which means you can download the open-source models, no matter where they 00:12:23.960 --> 00:12:27.477 were created, but run them locally, which fundamentally changes the game. 00:12:27.560 --> 00:12:32.877 >> I Yes, but I mainly we have a strong preference for American models. Some of 00:12:32.960 --> 00:12:36.517 our clients use models from all over the world, including China. 00:12:36.600 --> 00:12:40.957 I do believe if you're using any of these models, you know, I mean, I I I 00:12:41.040 --> 00:12:45.517 this the I can't go and disseminate, you know, no no Palantir, no no ontology, no 00:12:45.600 --> 00:12:49.397 love, because it may be not be right for the TV audiences, although who knows 00:12:49.480 --> 00:12:53.877 what I'll do. But um but yeah, it's going to put obviously it creates great 00:12:53.960 --> 00:12:56.957 demand for our products. So, since I don't want to be out there like I I'm 00:12:57.040 --> 00:13:01.437 cognizant of how this like is going to help us, obviously. But But But But But 00:13:01.520 --> 00:13:03.837 again, you you have to be very careful. 00:13:03.920 --> 00:13:07.477 Like, you can we should obviously scrutinize all models. There needs to be 00:13:07.560 --> 00:13:12.117 better oversight, no doubt. All models. But you do not want to do, 00:13:12.200 --> 00:13:16.237 if you do this, what Europe is trying to do, what you get are weak companies that 00:13:16.320 --> 00:13:17.837 nobody actually believes can compete. 00:13:17.920 --> 00:13:20.797 And that's So, you're screwing the people you're trying to help. You know, 00:13:20.880 --> 00:13:22.837 business is a weird thing. You can you 00:13:22.920 --> 00:13:25.557 can nudge, you can have preferences. If 00:13:25.640 --> 00:13:30.597 you do overt like that level without without being able to show something, 00:13:30.680 --> 00:13:34.397 you end up with industry that is and America is the dominant player in the 00:13:34.480 --> 00:13:38.517 world. Um, and so I think this will force a lot of honesty 00:13:38.600 --> 00:13:41.277 because if you are doing compute 00:13:41.360 --> 00:13:44.797 ontology, uh, actual product-based and 00:13:44.880 --> 00:13:48.797 technical orchestration of and FTEs 00:13:48.880 --> 00:13:50.877 with open model models, open way models, 00:13:50.960 --> 00:13:55.957 that's that is by far the best thing on the market. And yeah, the other things 00:13:56.040 --> 00:13:59.597 are, you know, so yeah, it's it's it's going to be very good for America and I 00:13:59.680 --> 00:14:02.077 don't think it's going to be I think it might actually be pretty good for 00:14:02.160 --> 00:14:04.517 Palantir, too. >> Last question for you, Dr. Karp. Again, 00:14:04.600 --> 00:14:08.637 thank you for taking time. Uh, a lot of people are wondering, is there actually 00:14:08.720 --> 00:14:11.517 going to be a light at the end of the tunnel when it comes to this entire AI 00:14:11.600 --> 00:14:16.557 revolution? Put broadly, what does it mean to be a 00:14:16.640 --> 00:14:21.117 enterprise software company that actually delivers value in the age of AI? 00:14:21.200 --> 00:14:23.757 >> Oh, you mean what does it mean to be Palantir? 00:14:23.840 --> 00:14:26.677 Uh, look, >> Well, we're we're It's a better way to put it. 00:14:26.760 --> 00:14:30.557 >> we're we're we're we're Palantir is at the way 00:14:30.640 --> 00:14:34.717 beginning of a revolution. Like I was I was just uh, two two years ago we're 00:14:34.800 --> 00:14:38.357 sitting here, we have an offsite, people are like, "Tell me something you believe 00:14:38.440 --> 00:14:40.557 and no one else believes." And like, we're obviously going to be a 00:14:40.640 --> 00:14:44.157 trillion-dollar company. Now, I think 00:14:44.240 --> 00:14:49.757 we're at the way beginning of a massive massive opportunity and there is no, 00:14:49.840 --> 00:14:53.117 like, it's very hard to know what the end point is, but I can tell you 00:14:53.200 --> 00:14:58.157 I can see American industry just accelerating forward and I think 00:14:58.240 --> 00:15:00.997 Palantir is going to play a disproportionate role in that simply 00:15:01.080 --> 00:15:04.837 because you can kind of fake like you've done 00:15:04.920 --> 00:15:08.957 these things before FDA application layer. 00:15:09.040 --> 00:15:11.637 >> Uh, terms that everyone start using now. 00:15:11.720 --> 00:15:15.277 >> It's like, but it it is yeah, but it First of all, it's it's it's the 00:15:15.360 --> 00:15:17.637 orchestration of all these things at the same time. 00:15:17.720 --> 00:15:19.677 >> Which is art. >> It's art It's much more art than 00:15:19.760 --> 00:15:22.637 science. You also you need deep access 00:15:22.720 --> 00:15:27.197 to these data sets and it working you need a knowledge of how to work within 00:15:27.280 --> 00:15:30.917 these institutions to expand them. They have to have seen your products actually 00:15:31.000 --> 00:15:33.757 work. Also, the software concept concept 00:15:33.840 --> 00:15:36.717 is very broad like it includes AI slop. 00:15:36.800 --> 00:15:39.837 It includes things that are obviously going to be replaced and then it 00:15:39.920 --> 00:15:43.797 includes things that like most of what Palantir does which are like essentially 00:15:43.880 --> 00:15:45.437 like primitive they're more like 00:15:45.520 --> 00:15:48.197 infrastructure you need to do to enhance 00:15:48.280 --> 00:15:51.117 what these things do and again, you 00:15:51.200 --> 00:15:54.117 know, the the demand for what we we do 00:15:54.200 --> 00:16:00.040 is beyond what we can supply. >> Thank you. Thank you. Thank you.