Vai al lettorePassa al contenuto principale
  • 2 giorni fa

Categoria

📺
TV
Trascrizione
00:04We are joined by co-founder and CEO Alex Karp of Palantir.
00:07Alex, thank you for having us here at your Artificial Intelligence Platform conference.
00:12Delighted to be here. Thank you for coming.
00:15You know, normally I would start off by showing off about how wonderful and strong and splendorous we are,
00:21but we are in the middle of a war, and one of the reasons I'm very happy we're having NDC
00:27is,
00:29obviously, Palantir's core mission, and I would hope the core mission of everybody in corporate America,
00:34is to dedicate all of our resources to making sure our soldiers come home healthy and well,
00:41and we are very, very focused on that right now.
00:44Of course. What can you tell us about Palantir's role in the Middle East War?
00:47How is your technology being used?
00:49Well, you know, we're not—well, first of all, again, just to repeat,
00:53the most important thing for Americans to understand is our brave men and women are in harm's way,
01:00and in my view, and I think in the view of most people in America, and certainly in the world,
01:05they're doing really important noble work.
01:08You know, I've read in the papers that we are able to engage and fight in war in the way
01:16we haven't been able to in the past,
01:18that we've regained our deterrent capabilities, and partly, obviously, it's the men and women on the front line
01:25that deserve most of the credit, but, you know, the fact that you can now target more precisely,
01:31more accurately, more quickly, and that you, meaning America, can do these things,
01:36organize the total power of our fleet and all of our resources,
01:41and bring it to bear against adversaries and enemies has shifted the way in which war is fought,
01:48and I have read that Palantir's Project Maven is the core backbone of that,
01:54and then I've also read that all the allies, Arab and non-Arab in the Middle East,
01:59may or may not be users of our platform as well, and that's expanding rapidly.
02:03I think the most important thing, leaving aside the heroism of our troops,
02:07which is the most important thing, is our adversaries and enemies are witnessing an ability to fight
02:13that they don't have, and they are going to find it very hard to acquire for a couple reasons.
02:18We have, for better or worse, as many know, I've been essentially an anti-neocon.
02:23I don't really believe in the wars we fought in the past,
02:26because I don't believe in regime change, and that's one of the reasons I'm supportive of this policy
02:31that we currently have, but what you want is to have a capacity that no one else can replicate,
02:40and America is the center of the AI revolution.
02:44The large language models on the battlefield are only useful if powered inside an ontology,
02:48and that's our ontology, and yeah, I guess I've read we're at the core of everything,
02:52and maybe that's exaggerated, but...
02:55Project Maven used to identify Iranian targets, including the Supreme Leader.
02:59Well, you know, I can't obviously go into specifics,
03:03but there are two things that are very important about this in general.
03:06One, you have to take infrastructure, both commercial and in government,
03:11that was built for an industrial age,
03:13and you have to transform that into infrastructure that takes advantage of the industrial capacities,
03:19and then puts them in a position where they can use post-industrial technologies, AI.
03:25What Palantir's products do in commercial and in government is take latent infrastructure,
03:31hospitals, planes, engines, missiles, and allows you to coordinate in a way as if it was built for today,
03:40very, very rapidly.
03:42And then we layer things on top of it that did not even exist a year ago,
03:46like being able to do 1,000 munitions that would have taken two years, five years ago,
03:51being able to understand what hit, how it hit, under what conditions,
03:55what's actually...
03:56Also coordinating between allies.
03:57Like if you, as a matter of theory,
03:59you're coordinating between allies that do not have full access to our data.
04:03And how do you do that?
04:04Well, that's a software Palantir uniquely driven problem that currently only we can do.
04:08So Palantir is working with our allies in the Middle East that are currently being attacked by Iran?
04:13Well, if you were attacked and you needed to coordinate,
04:16you would have to have a coordinating function.
04:18There's only one product that can actually do that for security.
04:21And that has to do with how we pipeline and do things in Foundry.
04:24So the short answer is, without answering your question,
04:27were this to work, there's only one way you can do it.
04:30And by the way, it's exactly what's happening in U.S. commercial.
04:34It's like if you have an airline or you have an engine,
04:38you essentially have 50 companies.
04:40It's not one company.
04:41And now you have to make those 50 companies work as one company
04:44because otherwise the company that buys our product is going to do that.
04:47And commercial is growing 137% year over year.
04:51Two big deals announced just in the last hour.
04:53Excuse me.
04:54The U.S. Navy as well as GE Aerospace.
04:56Yeah, look, you know, there's a lot we could share.
04:59Again, if you look at the financials,
05:01the bigger question in all of this, honestly, is what creates value?
05:05What gives you outsized alpha that can't be replicated?
05:08On the battlefield, on the commercial battlefield.
05:11So at large companies, you know, GE Aerospace, NVIDIA, very technical companies,
05:17and or essentially our ability to target and take out adversaries and enemies in a way no one else can.
05:23That's it's I mean, from a non-moral perspective, they're exactly the same.
05:27And why would you have to do this?
05:30It's because what makes you lethal on the battlefield and what makes you commercially viable,
05:35like in the current phase we're in, an AI-enhanced phase,
05:39you have physical assets that are not being used optimally,
05:43meaning the integration is not optimal because your data systems can't work,
05:47and then they have to be enhanced in a way that makes them truly unique, an N of one.
05:52So what makes America special right now is our lethal capacities, our ability to fight war,
05:59both because we've been doing it for 20 years, because we have meritocracy in our military,
06:03and because we finance it in a way no other country has,
06:06and because the AI revolution is uniquely American, something we, it's like every company,
06:12all the model providers that are relevant and the ontology that makes them valuable
06:16and the chips that they run on are built in this country.
06:21And that's happening.
06:22It's like, well, yeah, it's obviously happening.
06:25And then the question is, what does it mean for society?
06:28And those are really big questions because, you know, like, what does it mean for our job force?
06:33What does it mean for our fighting force?
06:35And a lot of, I mean, right now I'm 100, as focused as you can be on bringing,
06:39making sure we're delivering the magical technology to our warfighters.
06:43But then, you know, many people are watching this and figuring out, well, where should I invest?
06:48It basically comes down to, is this company going to be able to create value in a way no other
06:55company can?
06:56Can the military create value in a way?
06:58Or you're going to have a huge problem, unlike any problem you had in the past,
07:02because this is not we all do well kind of world.
07:05This is a winner-takes-all world on the battlefield, commercial,
07:09and that's why the U.S. is positioned so strongly on these.
07:12Then talk to us about Anthropik, the use of this technology, Claude,
07:17given that the Department of Defense has blacklisted Anthropik.
07:19Is Palantir still using Claude?
07:22We were in the middle of all these discussions.
07:24And so both for respect to the Department of War and Dario,
07:28I obviously can't go into specifics.
07:30And I, okay, but I can tell you what's, first of all, what's in the papers.
07:36The Department of War is planning to phase out Anthropik.
07:39Currently, it's not phased out.
07:41Our products are integrated with Anthropik.
07:43And in the future, it will probably be integrated with other large language models because of this dispute.
07:52The core question, though, is, and where I can give you my opinion,
07:57is in this country, on the battlefield, not domestic surveillance.
08:02No one believes it, but Palantir is the most important protector of the Fourth Amendment,
08:07or Fourth Amendment, meaning the right of privacy, in this country because of the way our product works.
08:11And I'm deeply committed to that, as are most Americans.
08:14The Fourth Amendment, in my view, or extrapolation, does not apply to adversaries on the battlefield.
08:21On the battlefield, where we are confronting people who want to kill us,
08:25and where men and women who are the most noble people in this country,
08:29and, by the way, the only institution in this country that's widely respected by every demographic,
08:34every demographic in this country respects the U.S. military.
08:37They deserve the best and most lethal technology in the world,
08:40and we at Palantir are going to make for effing sure that they get it,
08:45whether we get it from Anthropik, whether we get it from OpenAI,
08:49whether we get it from Elon and X, or whether we get it from NVIDIA or Google.
08:54The thing for us is we are committed to the warfighter getting the best technology.
08:59And, by the way, I'm bifurcating that from domestic issues.
09:02I see lots of reasons to restrict these technologies in law enforcement and in other use cases
09:09because they are so powerful they could proxy eviscerate,
09:12even if they're within the context of the law,
09:15the rights that were given to us by a higher being in the First, Second, Fourth, and Fifth Amendments.
09:19Okay, so if it's not Claude, there's another option available.
09:23For more information, please visit www.fema.org.
Commenti