WEBVTT 00:00:00.160 --> 00:00:02.717 the World Economic Forum in Davos. It's 00:00:02.800 --> 00:00:06.397 my pleasure to introduce Alex Karp. Um, 00:00:06.480 --> 00:00:11.437 I want to just start off on something more intimate between he and I. I'm 00:00:11.520 --> 00:00:13.277 pretty proud of what I created at Black 00:00:13.360 --> 00:00:16.237 Rockck, but my total return since I've 00:00:16.320 --> 00:00:19.997 been a CEO has only compounded at 21%. 00:00:20.080 --> 00:00:23.197 Uh, since Alex and Palunteer went 00:00:23.280 --> 00:00:25.837 public, uh, his compounded return is 00:00:25.920 --> 00:00:29.837 73%. So, um, 00:00:29.920 --> 00:00:32.637 congratulations, Alec. 00:00:32.720 --> 00:00:35.917 Uh, but more importantly, um, 00:00:36.000 --> 00:00:39.917 we're in the middle of a profound technological shift. I think we all are 00:00:40.000 --> 00:00:43.837 hearing about it, reading about it, feeling it, being a part of it. [snorts] 00:00:43.920 --> 00:00:45.517 And everybody's asking the question, you 00:00:45.600 --> 00:00:48.877 know, what can AI do for me or how can I 00:00:48.960 --> 00:00:51.037 translate it? What can it do for growth? 00:00:51.120 --> 00:00:53.037 What can it do for workers? What can it 00:00:53.120 --> 00:00:57.197 do for countries and national security? 00:00:57.280 --> 00:01:00.797 We are talking about it that it has the potential 00:01:00.880 --> 00:01:03.997 to grow capacity to modernize industries 00:01:04.080 --> 00:01:06.157 expand opportunity. 00:01:06.240 --> 00:01:09.277 It also will transform how we worked and 00:01:09.360 --> 00:01:11.677 where we work and how we work. And the 00:01:11.760 --> 00:01:14.557 question is are governments prepared for 00:01:14.640 --> 00:01:18.637 that real transformation of a society? 00:01:18.720 --> 00:01:21.037 So we need to make sure that as it's 00:01:21.120 --> 00:01:25.677 deployed it deploys in a way that empowers people, empowers institutions 00:01:25.760 --> 00:01:27.597 and builds a more resilient global 00:01:27.680 --> 00:01:32.157 economy. Um few leaders 00:01:32.240 --> 00:01:34.237 are truly at this intersection though I 00:01:34.320 --> 00:01:36.797 I I am certainly not that leader but in 00:01:36.880 --> 00:01:39.437 that intersection of technology, 00:01:39.520 --> 00:01:42.317 national security and the real economy 00:01:42.400 --> 00:01:45.037 the way Alex Karp is. 00:01:45.120 --> 00:01:48.077 um as co-founder and CEO of Palunteer, 00:01:48.160 --> 00:01:50.717 uh Alex has closely worked with defense 00:01:50.800 --> 00:01:52.637 and government 00:01:52.720 --> 00:01:55.437 private private organizations to apply 00:01:55.520 --> 00:01:58.957 how AI can be used and and in more of 00:01:59.040 --> 00:02:02.317 the critical areas um and it's really 00:02:02.400 --> 00:02:06.477 important and I must say my conversations that I've had and with 00:02:06.560 --> 00:02:08.797 Alex over over the last year has really 00:02:08.880 --> 00:02:11.917 enlightened me. So I'm looking forward 00:02:12.000 --> 00:02:14.317 to this Alex. So let me just start off. 00:02:14.400 --> 00:02:16.557 Sovereign states have often been early 00:02:16.640 --> 00:02:19.197 adapters of advanced technology and I 00:02:19.280 --> 00:02:21.677 think we're seeing that really very very 00:02:21.760 --> 00:02:26.317 intimately in the United States. But from your perspective, how is AI 00:02:26.400 --> 00:02:30.317 supporting decision-making in defense and security? 00:02:30.400 --> 00:02:33.677 >> Well, first of all, delighted to be here and with that introduction, maybe I 00:02:33.760 --> 00:02:37.117 should just stop. It's good downhill from here. Would you like to talk about 00:02:37.200 --> 00:02:39.757 our return some more? And I'll just uh 00:02:39.840 --> 00:02:43.437 I'll just time. I've been as I've been 00:02:43.520 --> 00:02:45.197 public now 26 years. So 00:02:45.280 --> 00:02:48.557 >> So um uh yeah. No, I I think one of the 00:02:48.640 --> 00:02:50.477 things to remember [snorts] uh like the 00:02:50.560 --> 00:02:53.437 the background backdrop of your question 00:02:53.520 --> 00:02:57.357 I I think for is though you have I in 00:02:57.440 --> 00:03:00.557 America and I would say also in Europe 00:03:00.640 --> 00:03:02.957 uh historically 00:03:03.040 --> 00:03:06.157 um uh industrial development and 00:03:06.240 --> 00:03:10.557 military technology were obviously co- 00:03:10.640 --> 00:03:12.877 in is a generalization but more true 00:03:12.960 --> 00:03:16.157 than not um you developed a product for 00:03:16.240 --> 00:03:17.597 the military 00:03:17.680 --> 00:03:22.157 that then was dual functioned and raised 00:03:22.240 --> 00:03:25.997 the standard of living of the country. 00:03:26.080 --> 00:03:29.597 Um and then um and then for lots of 00:03:29.680 --> 00:03:34.077 reasons I'll just leave aside for now that that's not the way at least 00:03:34.160 --> 00:03:36.397 technology was built until now there's 00:03:36.480 --> 00:03:37.757 all these defense tech startups and 00:03:37.840 --> 00:03:44.237 Palunteer. But um the uh um uh 00:03:44.320 --> 00:03:47.597 it it was that you had this thing where you were going to create something that 00:03:47.680 --> 00:03:48.877 had to work under the harshest 00:03:48.960 --> 00:03:52.717 conditions that was presumably 00:03:52.800 --> 00:03:55.917 significantly better than anyone else's 00:03:56.000 --> 00:03:58.397 to the point where it would give you a 00:03:58.480 --> 00:04:00.717 more than a slight advantage on a 00:04:00.800 --> 00:04:03.597 battlefield, especially if combined with 00:04:03.680 --> 00:04:06.397 um your way of fighting. And you saw 00:04:06.480 --> 00:04:09.917 this um there was a very famous uh 00:04:10.000 --> 00:04:14.717 social socialist German historian uh who 00:04:14.800 --> 00:04:18.717 was um said well one of the problems Germany had was the war fighting machine 00:04:18.800 --> 00:04:22.397 was so good that they just said well we'll decide on the battlefield who's 00:04:22.480 --> 00:04:25.437 right and that that that led to 00:04:25.520 --> 00:04:27.517 obviously lots of disjunctions and real 00:04:27.600 --> 00:04:30.557 problems in the in the American context. 00:04:30.640 --> 00:04:33.357 Um uh and I would say there you have a 00:04:33.440 --> 00:04:36.237 dislocation between what has happened 00:04:36.320 --> 00:04:40.797 recently in America and what is happening in Europe and what I think so 00:04:40.880 --> 00:04:45.517 I think America and China are are kind of very successful and I spent most of 00:04:45.600 --> 00:04:49.357 my adult life in Europe and I'm very pro- European but I think any honest 00:04:49.440 --> 00:04:51.037 assessment is it's not that's not gone 00:04:51.120 --> 00:04:56.157 very well. Um you built things that were 00:04:56.240 --> 00:04:59.517 could be used uh in an adverse 00:04:59.600 --> 00:05:03.597 condition. So rough, dirty, morally gray 00:05:03.680 --> 00:05:07.197 conditions. So how do you change the morality to fit with how we fight in the 00:05:07.280 --> 00:05:12.317 west is also a big vector. So the morality is difficult. Um the conditions 00:05:12.400 --> 00:05:16.877 under which you use the technology are difficult in in in software context. You 00:05:16.960 --> 00:05:19.677 you're [snorts] not you don't have you're not directly connected to the 00:05:19.760 --> 00:05:24.237 network in many cases. You have constraints under which you have to 00:05:24.320 --> 00:05:27.437 fight even though that's not the optimal way to fight. and you have very 00:05:27.520 --> 00:05:29.277 specialized ways of fighting in each 00:05:29.360 --> 00:05:32.717 country. Um, but the the positive side 00:05:32.800 --> 00:05:35.597 of that was that you also were building 00:05:35.680 --> 00:05:40.077 something that could be deployed that had an obvious value to the average 00:05:40.160 --> 00:05:43.757 citizen, >> right? Um and um so now you have this 00:05:43.840 --> 00:05:47.037 the AI thing is there's [clears throat] many things that are really interesting 00:05:47.120 --> 00:05:48.717 about it but if you start with the 00:05:48.800 --> 00:05:53.757 benality that I think um until very very 00:05:53.840 --> 00:05:58.797 recently all kind of adversaries of 00:05:58.880 --> 00:06:01.117 broadly defined the west assumed that 00:06:01.200 --> 00:06:04.237 investments in softwarebased 00:06:04.320 --> 00:06:08.317 uh defense were some kind of crazy thing 00:06:08.400 --> 00:06:13.267 Americans do for marketing. uh get rich, company blows up, you're on 00:06:13.350 --> 00:06:13.917 [clears throat] your beach in the 00:06:14.000 --> 00:06:16.237 Bahamas, but shareholders are happy and 00:06:16.320 --> 00:06:21.117 it's gone kind of thing. And um uh and 00:06:21.200 --> 00:06:25.117 um what you've seen uh so first of all 00:06:25.200 --> 00:06:27.437 you could say well that's changed but 00:06:27.520 --> 00:06:29.837 Germaine to your question the learning 00:06:29.920 --> 00:06:34.797 process in building how do you build this for sovereign governments is also a 00:06:34.880 --> 00:06:38.717 learning process for how do they adopt these technologies. So it's it's not 00:06:38.800 --> 00:06:42.317 just the technology because if you're building a tank like you know the 00:06:42.400 --> 00:06:45.917 British and then the French and then the Germans kind of optimized tank 00:06:46.000 --> 00:06:50.397 technology it's easy to see how you would deploy it but how do you deploy a 00:06:50.480 --> 00:06:54.157 system that's primary value is organizing parts on the battlefield 00:06:54.240 --> 00:06:58.317 without seeing the parts on the battlefield and seeing does it work how 00:06:58.400 --> 00:07:01.997 much does it work is it much better than what we had can we do things we could we 00:07:02.080 --> 00:07:03.837 couldn't do in the past and then there's 00:07:03.920 --> 00:07:07.757 a hidden thing about uh software AI 00:07:07.840 --> 00:07:10.237 that a lot of the value interestingly 00:07:10.320 --> 00:07:12.717 people always assume the value is from 00:07:12.800 --> 00:07:15.437 um where you are to where you should be 00:07:15.520 --> 00:07:19.997 and that's obvious but in most sovereign nations in the world I mean we deal with 00:07:20.080 --> 00:07:22.077 almost all in one form or another that 00:07:22.160 --> 00:07:25.437 are broadly defined as in a that would 00:07:25.520 --> 00:07:28.077 be you know it it could deal with a 00:07:28.160 --> 00:07:30.717 company that's you know American 00:07:30.800 --> 00:07:33.597 the actual technological 00:07:33.680 --> 00:07:36.797 uh uh rigor of the of the enterprises 00:07:36.880 --> 00:07:40.237 has significant holes in it. So it like 00:07:40.320 --> 00:07:43.917 as you know it's like it's very I'm dyslexic. It's very dyslexic. There are 00:07:44.000 --> 00:07:46.637 like whole pieces of the enterprise that 00:07:46.720 --> 00:07:51.117 exist on a PowerPoint that when you go to battle you will find out do not 00:07:51.200 --> 00:07:53.597 exist. And this whatever country you're 00:07:53.680 --> 00:07:56.237 in in this room if you are in the west 00:07:56.320 --> 00:08:00.957 this is a problem you have. Uh a day of battlefield you will find out this is 00:08:01.040 --> 00:08:04.157 one of the advantages the Ukrainians had. They essentially started from 00:08:04.240 --> 00:08:08.317 nothing. And so there wasn't you didn't have to rediscover that your enterprise 00:08:08.400 --> 00:08:10.957 didn't work after you're in fighting. 00:08:11.040 --> 00:08:15.517 And one of the huge advantages America has is just for you know better or worse 00:08:15.600 --> 00:08:17.437 I believe it or not think mostly for 00:08:17.520 --> 00:08:19.277 worse I was always against 00:08:19.360 --> 00:08:22.557 interventions. I'm not a neocon. Um it's 00:08:22.640 --> 00:08:26.717 uh um uh we had just all this experience on 00:08:26.800 --> 00:08:28.157 the battlefield so you could see what 00:08:28.240 --> 00:08:32.237 worked and what didn't work. Um, but 00:08:32.320 --> 00:08:35.997 pulling it back to the so the first thing that sovereign nations really 00:08:36.080 --> 00:08:38.237 struggle with is can I identify which 00:08:38.320 --> 00:08:43.277 tech is better objectively? Can I even rate it? Can I rate I'm looking at I 00:08:43.360 --> 00:08:45.837 guess say y >> but don't you need to know where you 00:08:45.920 --> 00:08:48.237 want to go to ask the right question. 00:08:48.320 --> 00:08:51.757 >> Um, you know that's you that's that's 00:08:51.840 --> 00:08:55.677 you actually let me reframe it. Y >> you have to know where you are to know 00:08:55.760 --> 00:08:58.637 where you want to go. >> Okay. So like the the point I'm making 00:08:58.720 --> 00:09:01.517 is with this exogesis is one of the most 00:09:01.600 --> 00:09:05.837 important things Palanteer's done on the battlefield is be able to make up for 00:09:05.920 --> 00:09:09.757 the fact that half your enterprise doesn't even work on the battlefield. It 00:09:09.840 --> 00:09:13.997 does work in a lab on a powerpoint only built by country. 00:09:14.080 --> 00:09:15.997 >> It doesn't work because of machinery or 00:09:16.080 --> 00:09:19.197 human or humans. Well, because the 00:09:19.280 --> 00:09:23.837 conditions of a battlefield are rough, contagious and and like for example like 00:09:23.920 --> 00:09:27.517 the modern if you just take Ukraine as an example, just to make it empirical, 00:09:27.600 --> 00:09:29.437 you know, as a everyone like reads this 00:09:29.520 --> 00:09:31.357 like, okay, well, how hard could it be 00:09:31.440 --> 00:09:35.597 to move a drone from A to B? Uh, well, 00:09:35.680 --> 00:09:38.877 actually, first of all, you're going to need to know where you want to put the 00:09:38.960 --> 00:09:42.477 drone. That's going to require synchronizing all your data. You're 00:09:42.560 --> 00:09:44.797 going to need to do that without transferring that data to your 00:09:44.880 --> 00:09:47.517 adversary, which means you're going to have to know every single person who 00:09:47.600 --> 00:09:50.557 touched it. You're going to have to obuscate it till the final thing. Then 00:09:50.640 --> 00:09:53.917 you're going to want to do it presumably either within the coordinance of your 00:09:54.000 --> 00:09:58.557 strategy or with your ethics. So where does it where does it not go? You're 00:09:58.640 --> 00:10:00.157 going to want to be able to correlate. 00:10:00.240 --> 00:10:04.957 Do you want to put the drone on your asset? No one only two people in Ukraine 00:10:05.040 --> 00:10:08.477 may know that one of the generals is your asset. You can't tell people that's 00:10:08.560 --> 00:10:12.077 your asset. How do you how do you make it look like to your should soldiers 00:10:12.160 --> 00:10:16.557 that you actually were taking people out and you just missed your asset? Um then 00:10:16.640 --> 00:10:20.797 and then the war advances. So the Russians which are off they're often 00:10:20.880 --> 00:10:25.037 very underestimated for reasons I don't understand but like they're 00:10:25.120 --> 00:10:29.197 mathematically arguably the best in the world and things they may not have in 00:10:29.280 --> 00:10:33.117 the beginning they can kind of cobble together and so they began jamming 00:10:33.200 --> 00:10:36.877 electronics. So now you have a completely different problem, but your 00:10:36.960 --> 00:10:40.797 presumably your enterprise has to be the same or developed because now it's not a 00:10:40.880 --> 00:10:44.397 question of going from A to B. It's a question of going through a completely 00:10:44.480 --> 00:10:48.717 jammed environment where you have no connectivity while actually collecting 00:10:48.800 --> 00:10:52.317 data while you go through that environment. And every one of those 00:10:52.400 --> 00:10:55.037 things is like a dynamic challenge. 00:10:55.120 --> 00:10:57.437 >> None of which were foreseen even before 00:10:57.520 --> 00:10:59.917 the Ukrainian. every single battle zone 00:11:00.000 --> 00:11:02.317 in the world has a By the way, the other 00:11:02.400 --> 00:11:04.797 thing is um then people fight 00:11:04.880 --> 00:11:09.597 differently. So like if you look at the big battlefields, I'm sure some people 00:11:09.680 --> 00:11:13.597 love our work here and some people hate it. By the way, we we we welcome all 00:11:13.680 --> 00:11:16.877 opinions at Palanteer. Even at inside Palunteer, we have people who love our 00:11:16.960 --> 00:11:21.037 work and people who are unhappy with the work. Um uh we welcome all. 00:11:21.120 --> 00:11:22.237 >> That's a spirit of dialogue. 00:11:22.320 --> 00:11:24.157 >> It's a spirit of dialogue with a 00:11:24.240 --> 00:11:28.877 somewhat of a leader. uh and uh um but 00:11:28.960 --> 00:11:32.477 you know if you look at you know how uh 00:11:32.560 --> 00:11:35.117 you know when the Ukrainians it's a 00:11:35.200 --> 00:11:37.357 small team of people and very courageous 00:11:37.440 --> 00:11:39.517 soldiers that are very technical they 00:11:39.600 --> 00:11:43.197 have highly technical people that built 00:11:43.280 --> 00:11:47.197 on top of our product things that we don't understand and proprietary ways of 00:11:47.280 --> 00:11:51.517 using the product in Israel they according to rumors use their 00:11:51.600 --> 00:11:54.957 intelligence so most people fight military to military this was int 00:11:55.040 --> 00:11:56.957 intelligence in Iran from what I can 00:11:57.040 --> 00:11:59.197 tell from the papers. Um and then in 00:11:59.280 --> 00:12:03.037 America you just have massive uh forces 00:12:03.120 --> 00:12:06.557 that no other country has but they had to be integrated and the integration 00:12:06.640 --> 00:12:11.357 capacity. So every one of those things is different. Sure. And so the the the 00:12:11.440 --> 00:12:15.437 two-fold role of of of enterprise software on a battlefield is one to make 00:12:15.520 --> 00:12:19.357 sure all the underlying things actually works and then two to raise the level to 00:12:19.440 --> 00:12:21.197 a level no one else has in the world. 00:12:21.280 --> 00:12:25.197 >> Let me translate this now. I mean there's so much technology was created 00:12:25.280 --> 00:12:28.157 from defense whether it was the internet 00:12:28.240 --> 00:12:30.877 uh GPS. 00:12:30.960 --> 00:12:33.277 How do you envision this translating 00:12:33.360 --> 00:12:36.077 from defense and military 00:12:36.160 --> 00:12:39.437 um to corporations to businesses to 00:12:39.520 --> 00:12:43.117 society? Well, first of all, the the 00:12:43.200 --> 00:12:45.917 fact that it's basically a purely raw 00:12:46.000 --> 00:12:50.237 naked environment means that you actually know the ground truth of what 00:12:50.320 --> 00:12:54.237 could work independent of what enterprises think can work. And the 00:12:54.320 --> 00:12:58.557 generaliz the think at a at a at a high level, it's almost one to one 00:12:58.640 --> 00:13:01.197 translatable, but precisely because as 00:13:01.280 --> 00:13:03.357 just to take enterprises in general, not 00:13:03.440 --> 00:13:06.077 all enterprises tend to want to over 00:13:06.160 --> 00:13:08.557 time become like every other enterprise. 00:13:08.640 --> 00:13:12.557 So if you take five, you know, A enterprise and B enterprise and C 00:13:12.640 --> 00:13:14.717 enterprise, they're in the same market. 00:13:14.800 --> 00:13:19.037 Their tech infrastructure is trying to make them into the same enterprise. Uh 00:13:19.120 --> 00:13:20.877 they have the same orchart. >> Sure. 00:13:20.960 --> 00:13:23.597 >> They have presumably roughly the same. 00:13:23.680 --> 00:13:27.277 They don't have the same data infrastructure. And what you learn on 00:13:27.360 --> 00:13:31.357 the battlefield is that and in life is that that's not particularly valuable. 00:13:31.440 --> 00:13:35.037 What is very valuable is an enterprise can do something no enterprise in the 00:13:35.120 --> 00:13:40.157 world can do. And so that is the goal of every single military in intelligence 00:13:40.240 --> 00:13:43.357 service. In fact, all these intelligence service and militaries have their own 00:13:43.440 --> 00:13:48.077 specialization. And so when we went to commerce like what we're saying is how 00:13:48.160 --> 00:13:51.837 can we make your insurance come the way you underwrite? 00:13:51.920 --> 00:13:54.317 How can we make that to your tribal 00:13:54.400 --> 00:13:59.037 knowledge about underwriting? How can we transform this to knowledge everyone has 00:13:59.120 --> 00:14:03.277 to a knowledge only you have and with efficiencies that no one else has? So as 00:14:03.360 --> 00:14:05.917 an example on the battlefield one of the 00:14:06.000 --> 00:14:11.117 most important issues I is how do you acquire data process it and then put it 00:14:11.200 --> 00:14:14.157 into a framework where it can be actioned either from an intelligence 00:14:14.240 --> 00:14:17.437 perspective yeah so just obviously what 00:14:17.520 --> 00:14:21.837 does a business do in the end of the day what is a business doing I mean it is so 00:14:21.920 --> 00:14:26.077 for example especially in anything underwriting banking hospital intakes 00:14:26.160 --> 00:14:28.077 it's information it's information it's 00:14:28.160 --> 00:14:30.077 sorting the information in a way that 00:14:30.160 --> 00:14:33.757 you have a a distinct advantage over 00:14:33.840 --> 00:14:36.477 other people who are similarly situated 00:14:36.560 --> 00:14:41.517 uh and that that advantage can't be eviscerated easily and so de facto for 00:14:41.600 --> 00:14:44.637 what you can't what you're doing on the battlefield with our products without 00:14:44.720 --> 00:14:48.557 getting happy to is like what you're doing with ontology and foundry in some 00:14:48.640 --> 00:14:52.797 cases on the battlefield but definitely in like commercial organizations across 00:14:52.880 --> 00:14:57.517 the world but especially in America is when you approach the underwriter or the 00:14:57.600 --> 00:15:00.717 hospital we power tons and tons of hospitals right they have an intake 00:15:00.800 --> 00:15:02.157 problem they all have an intake problem. 00:15:02.240 --> 00:15:05.677 They all have a shortage of doctors and nurses. They are working in a low margin 00:15:05.760 --> 00:15:09.437 environment, but every single one has a different way of processing their 00:15:09.520 --> 00:15:12.637 patients according to what their specialty is and the kind of patients 00:15:12.720 --> 00:15:16.797 they don't do well with. And how do you manage that? And so the intake flow and 00:15:16.880 --> 00:15:19.517 into your enterprise in a way that you 00:15:19.600 --> 00:15:23.917 can actually process these things 10 15 times faster than you could before. And 00:15:24.000 --> 00:15:25.757 then the other >> that saves lives then 00:15:25.840 --> 00:15:28.957 >> it save what saves a lot of lives. Um it 00:15:29.040 --> 00:15:31.517 also interestingly because you're 00:15:31.600 --> 00:15:36.077 processing the LLM in an ontology you have a structure it all despite what 00:15:36.160 --> 00:15:39.917 people may want to believe it also bolster civil liberties because now you 00:15:40.000 --> 00:15:43.837 can say well I mean just simple questions was someone processed based on 00:15:43.920 --> 00:15:48.877 economic uh considerations or were they based uh were they processed based on 00:15:48.960 --> 00:15:52.957 their background like those things are impossible to see unless you have like 00:15:53.040 --> 00:15:56.717 there's a huge civil liberties betterment side of this that typically 00:15:56.800 --> 00:16:00.637 people don't believe we care about or but it's actually exactly the opposite. 00:16:00.720 --> 00:16:03.117 We do care and you know showing is 00:16:03.200 --> 00:16:08.157 caring. It's like we can granularly show uh why someone came in, why they were 00:16:08.240 --> 00:16:11.517 taken, why they were rejected and we can do it in a way that makes business sense 00:16:11.600 --> 00:16:14.717 for the b the business itself. Um and 00:16:14.800 --> 00:16:17.517 then it it leads to safety efficiencies 00:16:17.600 --> 00:16:19.597 >> and probably brings down costs. Well, 00:16:19.680 --> 00:16:22.077 the if you want to do a a shorter kind 00:16:22.160 --> 00:16:25.037 of financial version, you basically in 00:16:25.120 --> 00:16:29.757 the past to do what we can do in in the full light of a public market, you 00:16:29.840 --> 00:16:33.037 needed a pri you need to take the company private. So, and then you would 00:16:33.120 --> 00:16:37.357 take out the cost structure and you that you probably resell it. Well, now you 00:16:37.440 --> 00:16:41.117 can take out the cost structure, make the workers more important. So, the 00:16:41.200 --> 00:16:46.397 actual workers, not the fat kind of in the middle. And uh um and then you can 00:16:46.480 --> 00:16:49.917 you can change the way they go to market. So the >> hospitals you can process this 00:16:50.000 --> 00:16:52.477 information faster presumably you'll have 00:16:52.560 --> 00:16:57.197 >> much faster and then the nurses the nurses the one of the the nurses and 00:16:57.280 --> 00:17:02.237 doctors are also happier and yeah >> so you you you 00:17:02.320 --> 00:17:05.917 suggested that for companies that rapidly adopt that they have to best way 00:17:06.000 --> 00:17:08.077 to do is go in private then they could restructure but 00:17:08.160 --> 00:17:10.797 >> no no what I'm suggesting is they don't have to do that anymore 00:17:10.880 --> 00:17:15.437 >> they don't have to do that. So what is the basic um 00:17:15.520 --> 00:17:17.677 what is the basic hindrance in the 00:17:17.760 --> 00:17:21.677 adaption of of of the of AI? Uh is it 00:17:21.760 --> 00:17:26.957 just um legacy systems, legacy issues? 00:17:27.040 --> 00:17:29.357 What how do you how do we accelerate the 00:17:29.440 --> 00:17:32.797 adoption that it's good for humanity? 00:17:32.880 --> 00:17:35.357 >> Well, I mean our adoption accelerates 00:17:35.440 --> 00:17:39.917 beyond our capacity. So it it the um I 00:17:40.000 --> 00:17:43.757 think if you just buy large language models off the shelf and try to do any 00:17:43.840 --> 00:17:45.117 of this it won't work. >> It's a commodity. 00:17:45.200 --> 00:17:48.077 >> It's a and well also it's not precise enough. 00:17:48.160 --> 00:17:50.717 >> It like you can't do underwriting with a 00:17:50.800 --> 00:17:52.317 you can't you couldn't do any of these 00:17:52.400 --> 00:17:56.317 things that are regulated. So moving to 00:17:56.400 --> 00:17:59.917 like everybody there's not like the the 00:18:00.000 --> 00:18:02.237 problem with adoption at this point is 00:18:02.320 --> 00:18:07.117 people have tried things that just can never work like buying you borrow a 00:18:07.200 --> 00:18:11.517 large language model you put it on your stack and you wonder why you know it's 00:18:11.600 --> 00:18:13.837 like not working and so what you're 00:18:13.920 --> 00:18:16.717 going to see especially in America is 00:18:16.800 --> 00:18:19.357 people trying to do something like what 00:18:19.440 --> 00:18:24.077 we do withtology maybe by hand because once you build the software layer to 00:18:24.160 --> 00:18:28.397 orchestrate and manage the lounge language models in in a language that 00:18:28.480 --> 00:18:31.597 your enterprise understands, you actually can create value. And [snorts] 00:18:31.680 --> 00:18:34.877 I don't I I think like there's like there's a lot of discussions like are we 00:18:34.960 --> 00:18:36.957 in an AI bubble? What is the meaning of 00:18:37.040 --> 00:18:41.997 the bubble? Well, it's like I I think what if anything we're just in a lag 00:18:42.080 --> 00:18:43.837 where there's a lot of AI. Some of it 00:18:43.920 --> 00:18:49.277 works. Again, if you go back to the battlefield context, like it most pe 00:18:49.360 --> 00:18:53.517 everybody in the world assumed this would not work, but now it does work. 00:18:53.600 --> 00:18:58.157 And so now the question isn't does it work, but how can we get it to work for 00:18:58.240 --> 00:19:01.837 my country? And this is exactly what's happening in 00:19:01.920 --> 00:19:06.157 >> companies. It's like oh this company it worked, mine didn't. What are you doing? 00:19:06.240 --> 00:19:10.157 >> Like for example, I mean if you want to just make it like parochial to us, 00:19:10.240 --> 00:19:12.637 Palanteer barely has a salesforce. In 00:19:12.720 --> 00:19:17.517 fact, it seems to be getting smaller every time I go see them. And it's like 00:19:17.600 --> 00:19:20.637 and it's simply because you're laughing because you know it says Palanteerians 00:19:20.720 --> 00:19:23.117 here, but it's it's getting smaller and 00:19:23.200 --> 00:19:27.117 smaller and smaller. And um and it it's 00:19:27.200 --> 00:19:28.397 not because we're trying to save on the 00:19:28.480 --> 00:19:31.837 unit economics. It's actually because it 00:19:31.920 --> 00:19:36.157 is a low trust environment in AI. People have tried lots of stuff. A lot of it 00:19:36.240 --> 00:19:40.477 hasn't worked. But if you've delivered stuff that does work, why do you need a 00:19:40.560 --> 00:19:43.677 salesforce? like you you you just it sells itself. 00:19:43.760 --> 00:19:46.557 >> Well, you have to say, "Hey, don't talk to us." 00:19:46.640 --> 00:19:48.397 >> And then on that's that's that's that's 00:19:48.480 --> 00:19:50.557 on commercial in government 00:19:50.640 --> 00:19:54.477 >> at this point. It's really we don't like 00:19:54.560 --> 00:19:58.157 we don't we it's it's it's very hard to export simply because we have to train 00:19:58.240 --> 00:20:00.237 the people and we have limited man. 00:20:00.320 --> 00:20:03.997 >> I was going to say that's your limited bandwidth, the training once somebody 00:20:04.080 --> 00:20:06.637 takes on your your software. Well, in 00:20:06.720 --> 00:20:11.117 the government contest, every every country like obviously has the 00:20:11.200 --> 00:20:16.157 equivalent of a security clearance, right? So to use our to build to build 00:20:16.240 --> 00:20:20.237 project Maven or Maven into your architecture, you're going to need 00:20:20.320 --> 00:20:25.117 somebody with the highest level of clearance that also is technical and 00:20:25.200 --> 00:20:28.477 most technical people unfortunately are not going after the highest level of 00:20:28.560 --> 00:20:31.997 clearance. So there are very very few people like that. So the that that 00:20:32.080 --> 00:20:36.477 resource is super scarce. Uh and then they have to be trained and that that 00:20:36.560 --> 00:20:40.397 takes that can take a while and and and then you also have to have like anything 00:20:40.480 --> 00:20:44.637 you really have to believe in this and really think it's important and you know 00:20:44.720 --> 00:20:47.037 not everybody fits in that category. 00:20:47.120 --> 00:20:51.437 >> How many people need to be trained to do this? I mean does it in a corporate 00:20:51.520 --> 00:20:53.117 level does it have to be from the CEO 00:20:53.200 --> 00:20:58.877 down? How does this work as you >> um well you're talking about like 00:20:58.960 --> 00:21:01.837 insurance underwriting? >> Well like you know Yeah. So insurance 00:21:01.920 --> 00:21:05.837 underwriting um I um the the way it 00:21:05.920 --> 00:21:08.317 works is the best case scenario you have 00:21:08.400 --> 00:21:10.157 the best and worst case scenario. Best 00:21:10.240 --> 00:21:15.037 case scenario um CEO is mathematically 00:21:15.120 --> 00:21:17.677 inclined >> even though they they may not know 00:21:17.760 --> 00:21:21.517 nothing about product but you can they 00:21:21.600 --> 00:21:23.197 can impute a product working by looking 00:21:23.280 --> 00:21:28.717 at the math. Um yeah and uh um and uh 00:21:28.800 --> 00:21:34.077 and and has you know and then we probably need to train five or six 00:21:34.160 --> 00:21:38.717 people on there. In the beginning we do all of it and then we transfer that as 00:21:38.800 --> 00:21:41.677 much as we can to them or we're part trying to partner with people who can do 00:21:41.760 --> 00:21:46.557 that with us but you need a small number of people but we still need more than we 00:21:46.640 --> 00:21:50.877 have. you you suggested repeatedly about how 00:21:50.960 --> 00:21:53.517 AI could strengthen foundations for an 00:21:53.600 --> 00:21:58.877 economy, especially we're seeing that in the United States. Um, how rapidly can 00:21:58.960 --> 00:22:02.157 AI change the growth trajectory? Because 00:22:02.240 --> 00:22:04.797 you you you mentioned that earlier about 00:22:04.880 --> 00:22:09.197 how it could improve the e economies, 00:22:09.280 --> 00:22:11.677 the well-being of companies. 00:22:11.760 --> 00:22:15.677 Well, um there in a lot of these things 00:22:15.760 --> 00:22:20.557 there's a speed question, but um 00:22:20.640 --> 00:22:22.237 I think like with a lot of our 00:22:22.320 --> 00:22:25.437 companies, it's like we can take out 00:22:25.520 --> 00:22:29.437 in the area we go in up to 80% of your 00:22:29.520 --> 00:22:33.197 cost and improve your top line dramatically. But it really depends on 00:22:33.280 --> 00:22:36.717 the use case and what we're doing. And then there's a speed function which is 00:22:36.800 --> 00:22:39.997 we can in the past five years ago that would take us a year. 00:22:40.080 --> 00:22:41.277 >> Sure. Now it could take a week. 00:22:41.360 --> 00:22:44.797 >> Week. Yeah. >> Let me further that question though. I'm 00:22:44.880 --> 00:22:48.637 sure it's on the minds of some people here. Is AI going to create jobs or 00:22:48.720 --> 00:22:52.717 destroy jobs overall? >> It Yeah. I think one of the unfortunate 00:22:52.800 --> 00:22:55.597 things of the narrative in the west is 00:22:55.680 --> 00:22:59.837 it it will destroy humanity's 00:22:59.920 --> 00:23:02.477 jobs of like you know you went to an 00:23:02.560 --> 00:23:06.397 elite school and you studied philosophy. 00:23:06.480 --> 00:23:10.237 Use myself as an example. Um, >> I did too. 00:23:10.320 --> 00:23:13.117 >> Yeah, you it hopefully you have some 00:23:13.200 --> 00:23:17.997 other skill. That one is going to be hard to market. Uh, and 00:23:18.080 --> 00:23:18.797 >> we thought it was hard to market. 00:23:18.880 --> 00:23:22.237 >> It was hard to market. Very hard. Uh, >> it was a good education. 00:23:22.320 --> 00:23:25.437 >> A very, very strong education. If you can get a job, you might keep it. But 00:23:25.520 --> 00:23:28.557 the hard That's what I always thought. I was like, if I finally get a job, I'll 00:23:28.640 --> 00:23:30.077 probably keep it and do well, but I'm 00:23:30.160 --> 00:23:31.757 not sure who's going to give me my first 00:23:31.840 --> 00:23:38.317 job. Um and uh um uh uh but like techn 00:23:38.400 --> 00:23:42.157 like technicians if you're a vocational technician 00:23:42.240 --> 00:23:47.277 >> or like like we're building batteries for a battery company and the people who 00:23:47.360 --> 00:23:50.637 are doing it in America are doing roughly the same job that Japanese 00:23:50.720 --> 00:23:53.997 engineers are doing and they went to high school and now they're very 00:23:54.080 --> 00:23:57.037 valuable if not irreplaceable because we 00:23:57.120 --> 00:24:02.157 can make them into something different than what they were very rapidly and 00:24:02.240 --> 00:24:05.037 those jobs are going to become more valuable. 00:24:05.120 --> 00:24:07.517 Um, I mean, you know, I not not to 00:24:07.600 --> 00:24:10.237 diverge into my usual political screeds, 00:24:10.320 --> 00:24:13.277 but it there are will be more than 00:24:13.360 --> 00:24:17.517 enough jobs for the citizens of your nation, especially those with vocational 00:24:17.600 --> 00:24:22.717 training. I do think these these trends really do make it hard to imagine why we 00:24:22.800 --> 00:24:26.957 should have large-scale immigration unless you have a very specialized skill because 00:24:27.040 --> 00:24:31.837 >> what about the foundation for white collar work in 00:24:31.920 --> 00:24:35.277 Europe and the United States has been through the universities but I just 00:24:35.360 --> 00:24:37.357 heard you say we're going to need more 00:24:37.440 --> 00:24:42.557 vocational men and women and they may they're going to be but are you also 00:24:42.640 --> 00:24:47.197 insinuating we're probably going to need less white collar? I think like I think 00:24:47.280 --> 00:24:49.517 what we need to do is yes, but I I think 00:24:49.600 --> 00:24:51.037 we need different ways of testing 00:24:51.120 --> 00:24:54.237 aptitude. You know, it's like um you you 00:24:54.320 --> 00:24:59.037 know there are a lot of people doing X that should be doing Y. Like if you 00:24:59.120 --> 00:25:01.117 could manage one of our system like just 00:25:01.200 --> 00:25:03.917 the person managing our maven system in 00:25:04.000 --> 00:25:06.637 the US Army is a former police officer 00:25:06.720 --> 00:25:09.517 who I think went to a junior college and 00:25:09.600 --> 00:25:12.477 they're doing very very high-end very 00:25:12.560 --> 00:25:16.077 complicated targeting globally and that 00:25:16.160 --> 00:25:20.397 person actually is irreplaceable and I 00:25:20.480 --> 00:25:22.557 think in the past the way we tested for 00:25:22.640 --> 00:25:26.637 aptitude uh would not have fully exposed 00:25:26.720 --> 00:25:30.957 how irreplaceable that person's talents are and would they been as talented if 00:25:31.040 --> 00:25:31.917 they had not [clears throat] gone to 00:25:32.000 --> 00:25:36.157 their college? Yes. Um and but I think 00:25:36.240 --> 00:25:39.357 the the a I tend to even inside Palunteer if you look at inside 00:25:39.440 --> 00:25:41.757 Palunteer what am I really doing all 00:25:41.840 --> 00:25:44.397 day? I'm want walking around figuring 00:25:44.480 --> 00:25:49.357 out what is someone's outlier aptitude and then I'm putting them on that thing 00:25:49.440 --> 00:25:52.957 and trying to get them to stay on that thing and not on the five other things 00:25:53.040 --> 00:25:56.477 they think they're great at like you know >> keeping their 00:25:56.560 --> 00:25:59.837 >> Yeah. It's like well you know everyone at Palanteer every every engineer at 00:25:59.920 --> 00:26:03.117 Palunteer uh it's it's the most wherever 00:26:03.200 --> 00:26:07.437 I go in the world like for for as you know maybe for 18 years everyone thought 00:26:07.520 --> 00:26:10.957 we were like a business joke and now lots of business people want my advice 00:26:11.040 --> 00:26:14.157 you know the only people who don't want my advice at Palanteer about business 00:26:14.240 --> 00:26:17.917 are Palunteer engineers they're like hey Alex I have an idea about how we could 00:26:18.000 --> 00:26:22.477 just be in a much better company and it's it's always like yeah it's like 00:26:22.560 --> 00:26:26.637 it's like literally McDonald's but it's like we should have some titles 00:26:26.720 --> 00:26:28.967 and you should stop speaking in public. 00:26:29.050 --> 00:26:32.157 [laughter] And uh yeah, and then I mean there's probably right about speaking in 00:26:32.240 --> 00:26:34.637 public sometimes. I certainly admit that. 00:26:34.720 --> 00:26:37.597 >> I don't think you uh I don't think you did anything wrong today. 00:26:37.680 --> 00:26:42.637 >> Yeah. So uh yeah, thank you for that high praise. [laughter] 00:26:42.720 --> 00:26:45.277 >> One of the keys to success is setting the bar very low. 00:26:45.360 --> 00:26:48.077 >> Yes. No, I don't believe that's how you 00:26:48.160 --> 00:26:53.400 uh operate Palunteer. Um one last question. 00:26:53.680 --> 00:26:56.877 Where where is this where will the curve 00:26:56.960 --> 00:27:00.397 of AI go in the utilization uh in the 00:27:00.480 --> 00:27:02.797 United States and other developed 00:27:02.880 --> 00:27:05.757 economies? But what about the developing 00:27:05.840 --> 00:27:08.717 economies? How can they participate in 00:27:08.800 --> 00:27:11.517 this? I mean, I read a research report 00:27:11.600 --> 00:27:14.317 yesterday that said the application of 00:27:14.400 --> 00:27:18.637 AI has been so dominant by societies of 00:27:18.720 --> 00:27:20.957 high education or companies of high 00:27:21.040 --> 00:27:24.317 education and they're seeing a very big 00:27:24.400 --> 00:27:27.997 um divergence that is occurring already 00:27:28.080 --> 00:27:32.237 and it's so much based on the application of education and how that is 00:27:32.320 --> 00:27:34.717 being utilized. So, is AI going to 00:27:34.800 --> 00:27:38.397 create more um a more a greater 00:27:38.480 --> 00:27:41.757 imbalance in our in our world in terms of growth? 00:27:41.840 --> 00:27:43.517 >> Well, I think the obvious first 00:27:43.600 --> 00:27:47.117 imbalance is it seems like America and 00:27:47.200 --> 00:27:51.277 China understand versions of making this work and they're different. 00:27:51.360 --> 00:27:53.597 >> Yes. >> But they both work and they work at 00:27:53.680 --> 00:27:59.197 scale and I think that is very likely to accelerate way beyond what most people 00:27:59.280 --> 00:28:02.797 believe is possible. like the discount rate I think not in the short term but 00:28:02.880 --> 00:28:05.917 in the long term is way too high on what 00:28:06.000 --> 00:28:07.917 will be done and how this will impact 00:28:08.000 --> 00:28:10.317 every aspect of our society and I would 00:28:10.400 --> 00:28:13.837 say especially on military and then I I 00:28:13.920 --> 00:28:17.517 tend to be a realist and that I think you know you have wide divergences it's 00:28:17.600 --> 00:28:21.677 going to be hard to have the kind of discussions people want to have where 00:28:21.760 --> 00:28:26.717 two countries are and and with a maybe a third following of Russia on the new on 00:28:26.800 --> 00:28:29.677 like the because they they're so good at fighting 00:28:29.760 --> 00:28:33.197 Um and then and then I I look I spent 00:28:33.280 --> 00:28:35.037 and I'll get to the developing world. I 00:28:35.120 --> 00:28:37.757 spent a lot of my life my most important 00:28:37.840 --> 00:28:41.597 years and my father's family came from a 00:28:41.680 --> 00:28:46.157 part of Germany and I I really care about Europe and especially the germ 00:28:46.240 --> 00:28:48.557 parts of Europe uh where I had many of 00:28:48.640 --> 00:28:50.717 my best years. I still fantasize of 00:28:50.800 --> 00:28:53.677 going back to grad school for not for 00:28:53.760 --> 00:28:57.837 the learning reasons. Um uh and um 00:28:57.920 --> 00:28:58.957 >> you're going to have more fun. 00:28:59.040 --> 00:29:03.357 >> I had so much fun. Oh, we we won't go into that. Uh but uh it's like endless. 00:29:03.440 --> 00:29:04.957 I sometimes when I'm traveling across 00:29:05.040 --> 00:29:06.237 the country, I just think of grad 00:29:06.320 --> 00:29:12.717 school. But um uh it's uh uh um but um 00:29:12.800 --> 00:29:19.277 um uh I I the the tech adoption in in in 00:29:19.360 --> 00:29:22.957 Europe is a serious and very very 00:29:23.040 --> 00:29:25.757 structural problem. And what scares me 00:29:25.840 --> 00:29:30.477 the most is I haven't seen any political leader just stand up and say we have a 00:29:30.560 --> 00:29:32.557 serious and structural problem that we 00:29:32.640 --> 00:29:35.677 are going to fix. So that then you get 00:29:35.760 --> 00:29:38.557 to the developing world. I would imagine 00:29:38.640 --> 00:29:43.757 it also depends what you mean by the developing world. I would imagine with 00:29:43.840 --> 00:29:47.917 not enough knowledge you're just going to find pockets that go very well and 00:29:48.000 --> 00:29:49.917 pockets that go very poorly. As a 00:29:50.000 --> 00:29:52.317 generalization like again if you go back 00:29:52.400 --> 00:29:55.437 to this somewhat n unsuccessful salopy I 00:29:55.520 --> 00:29:57.837 had about the underlying architecture. 00:29:57.920 --> 00:30:00.317 One way to look at at the unfairness of 00:30:00.400 --> 00:30:03.357 AI is it pentests meaning it it 00:30:03.440 --> 00:30:06.077 loadbears on things. So societies that 00:30:06.160 --> 00:30:08.397 can and organizations and companies that 00:30:08.480 --> 00:30:13.357 can bear that load have a huge advantage. The problem is if you can't 00:30:13.440 --> 00:30:17.597 if you've been pretending you're bearing a load you're not it collapses and 00:30:17.680 --> 00:30:22.717 that's where you have to start. And so if you go around and just say okay what 00:30:22.800 --> 00:30:24.477 societies and micro cultures are going 00:30:24.560 --> 00:30:29.837 to be loadbearing here I think you would find that parts of the developing world 00:30:29.920 --> 00:30:31.197 certain communities in that are going to 00:30:31.280 --> 00:30:34.957 do very well you you do need a realistic 00:30:35.040 --> 00:30:37.117 assessment of the loadbearing 00:30:37.200 --> 00:30:39.997 >> and there there's a certain honesty that 00:30:40.080 --> 00:30:42.397 is painful for all of us in in in this 00:30:42.480 --> 00:30:45.677 technology large language models however 00:30:45.760 --> 00:30:48.557 implemented in software it you just 00:30:48.640 --> 00:30:50.877 cannot not obfuscate what can bear the 00:30:50.960 --> 00:30:53.117 load and what can't. And then political 00:30:53.200 --> 00:30:55.677 structures are built to do just that. 00:30:55.760 --> 00:30:59.997 Like, yeah, I can't fix anything, but I can give you some line that you 00:31:00.080 --> 00:31:03.757 want to hear that's going to make you not care about how bad your life is and 00:31:03.840 --> 00:31:05.517 how much worse it's going to be tomorrow. 00:31:05.600 --> 00:31:07.197 >> I can give you that for free. 00:31:07.280 --> 00:31:13.037 >> And those that that stuff uh is um 00:31:13.120 --> 00:31:15.517 that is harder to get away with in this 00:31:15.600 --> 00:31:17.757 culture. And you know, I I still view 00:31:17.840 --> 00:31:20.397 myself as a card carrying progressive. 00:31:20.480 --> 00:31:24.637 And I think it's the single most important thing a progressive could do 00:31:24.720 --> 00:31:28.477 is go around and say, "Yeah, but the revolution that's coming is going to 00:31:28.560 --> 00:31:32.637 expose the actual market value of what you're doing, whether we want it or 00:31:32.720 --> 00:31:36.077 not." Like it's like I don't even want to know the market value of some of this 00:31:36.160 --> 00:31:40.717 stuff. But it is over and over a relative rapid period of time. So next 00:31:40.800 --> 00:31:45.677 three years you're just going to get market value honesty in all sorts of 00:31:45.760 --> 00:31:47.677 character communities and micro 00:31:47.760 --> 00:31:49.517 communities and the best thing you can 00:31:49.600 --> 00:31:52.797 do if you are in a community whether 00:31:52.880 --> 00:31:54.797 that is a large community like Germany 00:31:54.880 --> 00:31:57.677 or a large community larger like America 00:31:57.760 --> 00:32:02.397 is and you really care for the people you're representing is to say yeah but 00:32:02.480 --> 00:32:06.397 let's we have to kind of look closely at what 00:32:06.480 --> 00:32:09.357 what load we can bear. 00:32:09.440 --> 00:32:14.290 Thank you, Alex. >> Thank you, everyone. [applause] 00:32:16.860 --> 00:32:18.880 [applause]