00:00Palantir is the company that really made a name for themselves in the kind of work that all these LLM
00:06companies are being tasked to do for the Department of War.
00:10And look, it's going beyond analytics now.
00:15So what Palantir did was you could look at unstructured data like the GPS coordinates, Thai intelligence in terms of
00:24finding about some source or some entity.
00:28And that was heavily used in a variety of use cases.
00:32Now, I think these LLMs are trained on, we know, 15 trillion tokens.
00:38And they can ingest large amounts of data and do analytics at a level that, you know, we have not
00:45seen.
00:45So in this case, the company Entropic has talked about their AI constitution or, in other words, putting guardrails around
00:54the kind of surveillance activities they want the LLMs to do.
00:57And obviously, from a government perspective, they want to do whatever it takes from a national security perspective.
01:05So I think that's where there is a discussion going around the kind of guardrails that they want to see
01:12and the level of, I guess, work that the government wants them to do.
01:16Okay, so the Department of Defense or Department of War, as it's now known, this relationship with Anthropic, it's not
01:23limited to just using Anthropic's models, is it?
01:26Because it wants to make sure that it's got kind of a wide source of AI large language models to
01:31draw from.
01:32So what is that relationship like with, I don't know, ChatGPT or Gemini or even Grok?
01:37Yeah, all of these companies have a contract with the Department of War.
01:42And like, for example, XAI's Grok has a $200 million contract.
01:48So, you know, the department is using all the possible tools that are available.
01:53And from that perspective, Anthropic is one of them.
01:56But no one else like ChatGPT hasn't talked about, you know, guardrails in such a big way as Anthropic has.
02:03So that's the big distinction.
02:05And obviously, SpaceX is a defense contractor for the department.
02:10And now with XAI merger, I think they can do more work around the drone side.
02:16So at the end of the day, I think you have to think of it from a geopolitical standpoint,
02:22what China is doing, you know, with that infrastructure and similar concepts.
02:26And what I think the U.S. defense is looking to do in terms of surveillance and national security interests.
02:33And these LLMs will play a pivotal role because, as I said, they can help you find that needle in
02:39the haystack that the current technology wasn't able to do.
02:43Who is Anthropic?
02:44Who are these people?
02:45Well, they just raised $30 billion in their latest funding round.
02:49So they're good fundraisers.
02:50They're good fundraisers.
02:52And they'll most likely go public this year.
02:54There is a very high likelihood of both Anthropic and OpenAI going public.
03:00And that's why, you know, they're out there in terms of establishing the use cases and the utilities, not just
03:06a coding agent.
03:07They want to show, you know, that they are addressing a lot more areas in terms of, you know, where
03:14these models are touching and the companies that they are interfacing with.
03:17So there is a very high chance that we'll see IPOs.
03:20So there is a very high chance that we'll see IPOs.
Comments