Latest / The Team House / AI: Boom or Doom? | EYES ON GEOPOLITICS
Transcript
- 0:00Speaker 1: Look, I'd rather live in the United States where we
- 0:03Speaker 1: have control over the technology and information that affects Americans.
- 0:08Speaker 1: And when you have tech like that used to be
- 0:11Speaker 1: fine for basically a century because America made the best
- 0:16Speaker 1: technology in most of the technology. Now we're entering a
- 0:19Speaker 1: world where China makes a lot of technology, and in
- 0:23Speaker 1: some cases may make better technology, but they don't want
- 0:27Speaker 1: to abide by our use or our constitution, our rules
- 0:33Speaker 1: around this. So I would just say, fine, like, you
- 0:37Speaker 1: can do whatever you want in your country, but we
- 0:39Speaker 1: are not going to use your technology.
- 0:41Speaker 2: Then, everybody, welcome to another episode of Eyes on Geopolitics.
- 0:48Speaker 2: I'm here today with Jonathan Hacket, and our guest is
- 0:51Speaker 2: Anthony Vincey. He's the author of The Fourth Intelligence Revolution,
- 0:56Speaker 2: The Future of Espionage, and the Battle of Save America.
- 1:00Speaker 2: He's got a pretty amazing resume. He previously served as
- 1:04Speaker 2: the first CTO and Associate Director of Capabilities at the
- 1:07Speaker 2: National Geospatial Intelligence Agency, following earlier work as an intelligence
- 1:12Speaker 2: officer in Iraq, Africa, and Asia. You could actually watch
- 1:15Speaker 2: an episode with Anthony on the Team House that we
- 1:17Speaker 2: did probably a good six months ago. Now, great episode,
- 1:22Speaker 2: check that out. I'll put a link in the description
- 1:23Speaker 2: you can check it out. I'll also put a link
- 1:25Speaker 2: in the descripture for his book. He is now the
- 1:28Speaker 2: co founder and CEO of Vico, an AI company focused
- 1:31Speaker 2: on decision making and finance and national security, and is
- 1:34Speaker 2: an adjunct Senior Fellow at the Center for a New
- 1:38Speaker 2: American Security and to boot, he holds his PhD in
- 1:41Speaker 2: International relations from the London School of Economics. So I
- 1:45Speaker 2: try to get somebody here that's on the level of
- 1:48Speaker 2: smart as Jonathan Hackett is, because Jonathan's got a pretty
- 1:51Speaker 2: good resume too, and I think I did a pretty
- 1:54Speaker 2: good job with the Anthony. So thanks for joining us today.
- 1:56Speaker 1: Yeah, man, thanks for having me.
- 2:00Speaker 2: This whole talk to be about what's going on with AI.
- 2:03Speaker 2: Me and Jonathan have been talking about this for like
- 2:05Speaker 2: a couple of months, like putting together a show and
- 2:06Speaker 2: trying to find a guest that can, like, you know,
- 2:09Speaker 2: give us the lay of the land of what's going on.
- 2:11Speaker 2: And I've joked with Jonathan. I think I joked with
- 2:13Speaker 2: you want an email too, about saying like this literally
- 2:15Speaker 2: could be a three hour conversation because it's like there's
- 2:18Speaker 2: just so many rabbit holes we can go down full disclosure.
- 2:24Speaker 2: I am more on the bearers side of AIS, particularly
- 2:27Speaker 2: when it comes to like just all all the capex
- 2:31Speaker 2: that's going on, and like, you know, wh where's it
- 2:33Speaker 2: really going? The cow expensive the computer is compared to
- 2:36Speaker 2: like how you know enterprises are using them, like they
- 2:39Speaker 2: doesn't really look like there's a big at least for
- 2:42Speaker 2: the frontier model. So it doesn't look like there's a
- 2:44Speaker 2: big profit center right now. But anyway, let's just get
- 2:50Speaker 2: into it. I mean, I don't know how you guys
- 2:53Speaker 2: want to play it, Jonathan, please, like let's just you know,
- 2:57Speaker 2: improve this basically, do we want it explain like what
- 3:00Speaker 2: the frontier models mean, like should we go more of
- 3:03Speaker 2: a basic route for like the Layman. I'm in Layman
- 3:05Speaker 2: obviously too, but I just read this insensitively.
- 3:08Speaker 3: I think it's helpful to have like a general question
- 3:10Speaker 3: of Okay, why should people care about the crossover question
- 3:14Speaker 3: of China and AI? Because there's the China national security
- 3:17Speaker 3: threat that's existed for a long time, especially international security
- 3:20Speaker 3: strategy all the way back to nineteen eighty seven when
- 3:21Speaker 3: the first one came out. Now AI has entered the chat,
- 3:25Speaker 3: and the question for the American listener is why should
- 3:28Speaker 3: they care about that?
- 3:31Speaker 2: Yeah, I love it. Let's go because we can go
- 3:32Speaker 2: in any direction here. So Anthony, what are you track
- 3:35Speaker 2: in basically right now? With AI? I mean, I know
- 3:38Speaker 2: I sent you over that Financial Times article and that
- 3:41Speaker 2: does seem like the big, the big scary.
- 3:45Speaker 1: Yeah, well look China AI, I think is that's the question.
- 3:49Speaker 1: That's like the generational level question because I think about
- 3:54Speaker 1: it like this. AI is not like software. AI does
- 3:58Speaker 1: what people do right, and it communicates information to you.
- 4:03Speaker 1: It doesn't just show you something, but you interact with it, right,
- 4:09Speaker 1: So anything when you get on claud, you get on chat,
- 4:12Speaker 1: GPT and you ask it a question, it can ask
- 4:16Speaker 1: you questions. You can continue to ask it questions. And
- 4:20Speaker 1: the thing is that it can persuade you to do
- 4:24Speaker 1: things right, like it can lean you in a direction
- 4:27Speaker 1: that's different than anything we've had before. Only people were
- 4:30Speaker 1: able to do that previously. So that's one thing. The
- 4:33Speaker 1: second thing is that AI that's kind of LMS like
- 4:37Speaker 1: AI you talk to. There's another form of AI that
- 4:39Speaker 1: you're using for autonomous vehicles that can drive way more
- 4:45Speaker 1: cars around, or it can fly drones around. That's also
- 4:50Speaker 1: super important. There's a third kind of AI that I
- 4:54Speaker 1: think is you were going to see more and more of.
- 4:57Speaker 1: And it's kind of what I do in my company
- 4:59Speaker 1: at Vico, which is using AI to help run businesses
- 5:04Speaker 1: or even to help run the military. It's making strategic choices.
- 5:08Speaker 1: It's allowing you to play through lots of decision lots
- 5:14Speaker 1: of decisions. Right, those three kinds of AI. That's kind
- 5:17Speaker 1: of how the world works. You have people talking to
- 5:21Speaker 1: people now talking AI. You have AI that's able to
- 5:25Speaker 1: drive cars and other vehicles, and you have a system
- 5:28Speaker 1: that can run companies. Now, these are the things that
- 5:31Speaker 1: we obviously are competing with China around. And the question
- 5:36Speaker 1: is going to be are we going to use Chinese
- 5:39Speaker 1: software or AI systems to do these things? Or we're
- 5:44Speaker 1: going to use American systems? And then what about our
- 5:47Speaker 1: allies or just third party countries like that are neither
- 5:52Speaker 1: aligned with the US nor China kind of sit in
- 5:55Speaker 1: the middle and what are they going to use? And
- 5:58Speaker 1: we've seen this situation before or with Huawei with telecommunications equipment,
- 6:04Speaker 1: where the question became, are you going to have your
- 6:06Speaker 1: telecommunication system set up on Western equipment from ericson or something,
- 6:11Speaker 1: or are you going to have it set up on
- 6:12Speaker 1: Chinese equipment. The downside risk there was that Chinese equipment,
- 6:15Speaker 1: they can spy on it and steal all of your data.
- 6:19Speaker 1: It's completely unsecure. Well, we have the same situation with
- 6:23Speaker 1: China right now, is that if you use a Chinese
- 6:26Speaker 1: AI system, they can spy on it, they can access
- 6:30Speaker 1: your data. But it's even worse now because they could
- 6:34Speaker 1: do two other things. One, in theory, they could use
- 6:38Speaker 1: that AI system to control what you're using it for.
- 6:42Speaker 1: So if you're using a piece of AI to run
- 6:45Speaker 1: an autonomous vehicle, in theory, they might be able to
- 6:48Speaker 1: control that autonomous vehicle. That could be really problematic in
- 6:51Speaker 1: war for example. But the even more nefarious possibility is
- 6:55Speaker 1: that Chinese AI remember the part where you can talk
- 6:59Speaker 1: to it, you use chat GPT where you ask it
- 7:02Speaker 1: a question, say about how to vote in this coming midterm.
- 7:07Speaker 1: If it's Chinese AI system, they could mount an information
- 7:10Speaker 1: attack against you and can actually persuade you to do something.
- 7:14Speaker 1: So those are like really serious consequences for like the
- 7:18Speaker 1: average everyday American citizen, Like do you really want your vote,
- 7:24Speaker 1: for example, in the midterm to be biased in some
- 7:27Speaker 1: way because of your interaction with the Chinese system. Do
- 7:30Speaker 1: you really want this waym car that you're driving in
- 7:35Speaker 1: to potentially be controlled by China and say in a
- 7:39Speaker 1: state of war, to be used against you maybe I
- 7:41Speaker 1: don't know, runs people over or something, or American military
- 7:44Speaker 1: equipment to be harmed by a Chinese AI system. And
- 7:48Speaker 1: worst case scenario, if you're running your business and you're
- 7:51Speaker 1: using AI to do this, or the government is being
- 7:53Speaker 1: run by AI, do you really want it to be
- 7:55Speaker 1: compromised by Chinese AI system? And I think these are
- 7:58Speaker 1: like really serious questions to address.
- 8:02Speaker 3: And I think also too, I was actually reading an
- 8:03Speaker 3: article recently, I think this week about open weight models
- 8:06Speaker 3: that China is kind of pioneering and getting out there
- 8:08Speaker 3: a lot of and people are seeing them their lower
- 8:10Speaker 3: costs and all these there are all kinds of benefits
- 8:12Speaker 3: and advantages on one ledger. And as I'm reading more
- 8:15Speaker 3: about what open weight actually means, at my first thought
- 8:19Speaker 3: before I dug into this was that open weight was
- 8:21Speaker 3: the same as open source, and I've learned that that's
- 8:24Speaker 3: not true. Can you talk a little bit about why
- 8:26Speaker 3: that's problematic?
- 8:29Speaker 1: Yeah, the yeah, open weight means you're getting the open
- 8:34Speaker 1: source means it's a piece of software that you can
- 8:36Speaker 1: use and you can modify, and you can do whatever
- 8:38Speaker 1: you want with and build on. So like Linux, for example,
- 8:42Speaker 1: open weight means that they have provided the model, but
- 8:45Speaker 1: not the underlying data in the model. So it does
- 8:48Speaker 1: mean that you can use it now and you can
- 8:51Speaker 1: build that, you can build an app, say on top
- 8:56Speaker 1: of it, but you don't actually know the data that
- 9:00Speaker 1: and into training it, you don't know exactly how it's
- 9:02Speaker 1: going to respond. So the danger is that it could
- 9:06Speaker 1: respond in a way that is biased. So kind of notoriously,
- 9:10Speaker 1: the Chinese open weight models are biased towards the Chinese
- 9:15Speaker 1: Communist Party. Like if you ask a deep Seek model
- 9:18Speaker 1: or Kimmi or choose your Chinese open weight model a
- 9:21Speaker 1: question about tam and Square, it's going to give you
- 9:23Speaker 1: the party line, right, it's going to say what the
- 9:27Speaker 1: CCP wants you to hear, and that that's problematic on
- 9:32Speaker 1: it on its own. And but really what open weight
- 9:37Speaker 1: and open source mean is that it's free. And so
- 9:40Speaker 1: what China's done is they've made these free models. Deep
- 9:45Speaker 1: Seek and a couple of other companies have gone out
- 9:48Speaker 1: and done this, and so what's happening is they're essentially
- 9:51Speaker 1: giving like this low cost solution that US or other
- 9:55Speaker 1: country companies can build on top of. And so it's
- 10:00Speaker 1: as similar to the Huawei thing again, right, Huahwei was
- 10:03Speaker 1: able to produce routers and other telecommunications equipment at a
- 10:08Speaker 1: fraction of the cost of Western equipment, so people started
- 10:11Speaker 1: using that kind of equipment, and these these systems were
- 10:14Speaker 1: compromised in a way or could be compromised by Chinese
- 10:18Speaker 1: intelligence and so forth. Well, same thing with AI. They're
- 10:21Speaker 1: sort of giving away these open source models that are
- 10:25Speaker 1: effectively free so that people will build their own software
- 10:29Speaker 1: and apps and so forth on top of them. But
- 10:32Speaker 1: they're compromised.
- 10:35Speaker 3: And actually that brings up a good point. I was
- 10:37Speaker 3: working somewhere over the summer where we got fable from
- 10:39Speaker 3: Anthropic for like two days, and then the government shut
- 10:43Speaker 3: it down, and there was this kind of disruptive moment
- 10:45Speaker 3: where the government was trying at the US government was
- 10:46Speaker 3: trying to figure out, how do we regulate this, do
- 10:49Speaker 3: we put export controls on it? How dangerous is it?
- 10:51Speaker 3: And all these questions. But at the same time China
- 10:53Speaker 3: is looking at this in the opposite way. They're trying
- 10:55Speaker 3: to get as much access to people as possible, whereas
- 10:58Speaker 3: the US seems to be trying to be more careful.
- 11:00Speaker 3: What's kind of the right way to balance these two
- 11:03Speaker 3: competing philosophies.
- 11:06Speaker 1: Yeah, so there's a downside to letting AI models out right,
- 11:10Speaker 1: which is that they can be used to do dangerous things. Right,
- 11:15Speaker 1: you could take an AI model and in the case
- 11:17Speaker 1: of Fable and Mythos the Claude models, that they could
- 11:23Speaker 1: be used to hack to perform an offensive cyber attack,
- 11:29Speaker 1: and that could be really dangerous. Imagine you just sort
- 11:31Speaker 1: of let everybody in the world have that, and now
- 11:33Speaker 1: these criminals, criminal hackers were able to use the stuff,
- 11:37Speaker 1: and then they can go and pack into banks or
- 11:40Speaker 1: into healthcare systems and so forth. Could be super dangerous.
- 11:45Speaker 1: So there is a debate just on that, like when
- 11:49Speaker 1: and how should we release these models. And it gets
- 11:51Speaker 1: worse by the way than just hacking. As these models
- 11:55Speaker 1: get more powerful, it's possible you could use them to
- 11:57Speaker 1: create like a bioweapon, right, that a person with little
- 12:02Speaker 1: to no biological science experience could actually create a bioweapon
- 12:07Speaker 1: by following the instructions like you would with you know,
- 12:11Speaker 1: a cooking recipe from an AI system, and that like
- 12:14Speaker 1: that's like terrifying to even think about. So we want
- 12:17Speaker 1: to have some control. China like does have some control
- 12:23Speaker 1: even on the the open weight models, right. The CCP
- 12:27Speaker 1: has established guidelines and controls to those companies like to
- 12:31Speaker 1: Deep Seek, for example, around information that they want certain
- 12:34Speaker 1: information about Tenement Square or the Wigers, which is a
- 12:38Speaker 1: minority group that's that's persecuted in China. They want to
- 12:42Speaker 1: control that information, so they are putting in those controls.
- 12:45Speaker 1: But then these open weight models are being released to
- 12:49Speaker 1: everybody and they don't have there's no way for the
- 12:53Speaker 1: US government to put its own controls, like say, around
- 12:56Speaker 1: cybersecurity issues. Now, this wasn't a problem until recently because
- 13:01Speaker 1: the US models like Claude and Chat, GPT and GROC
- 13:05Speaker 1: and so forth, were by far the most powerful models
- 13:07Speaker 1: and could do the most powerful things. And these open
- 13:10Speaker 1: weight models were, you know, a generation or two generations behind,
- 13:15Speaker 1: and so they weren't as dangerous. But now they're sort
- 13:17Speaker 1: of catching up, and as they catch up, they might
- 13:19Speaker 1: get as dangerous as a Claude model for hacking and
- 13:24Speaker 1: for other purposes, and there's no way for the US
- 13:26Speaker 1: government to regulate it or control it in any way.
- 13:29Speaker 1: And in theory, China might let people, you know, allow
- 13:33Speaker 1: a model out that could be used for cyber hacking.
- 13:37Speaker 2: What do you guys take what's your take on? You know,
- 13:40Speaker 2: Microsoft announced maybe a few weeks ago, also Microsoft being
- 13:44Speaker 2: a big investor into open ai, that they're going to
- 13:47Speaker 2: start using deep seek or at least look into using
- 13:49Speaker 2: deep seek for their some of their enterprise stuff. I
- 13:52Speaker 2: mean that's going to ring alarm bells for everybody, know, Yeah.
- 13:57Speaker 1: I mean, look, my personal take is that just like Huawei,
- 14:03Speaker 1: it should effectively be made illegal or at least very
- 14:08Speaker 1: problematic for US business to use these things because for
- 14:11Speaker 1: all the reasons I mentioned, like these things, they're compromised
- 14:15Speaker 1: in that sense, and there's a danger in that they're
- 14:17Speaker 1: already large American companies using Chinese open weight models. Airbnb
- 14:23Speaker 1: uses them, there's a few other big ones, and then
- 14:27Speaker 1: there's lots of small startups that are using them. Microsoft
- 14:30Speaker 1: saying they're going to use it is not great. I
- 14:33Speaker 1: suspect that really they would be using it for something
- 14:37Speaker 1: that they consider to be low security risk, right, And
- 14:41Speaker 1: that's fine, But then who's making that call about what's
- 14:44Speaker 1: low security risk? Right? And we are in effect creating
- 14:50Speaker 1: the same situation as Huawei, where we're going to kind
- 14:53Speaker 1: of build up their open weight model companies over time
- 14:57Speaker 1: and then we won't be able to compete.
- 15:00Speaker 3: You mentioned something really interesting, you said, like Airbnb and
- 15:02Speaker 3: these smaller companies are using deepseek. You know, I used
- 15:05Speaker 3: to teach at our Human Intelligence Collection course and we
- 15:09Speaker 3: taught about ubiquitous technical surveillance and how to do things
- 15:12Speaker 3: under a uts threat and so on. That was before AI,
- 15:16Speaker 3: And now I'm very curious about how it's not just
- 15:20Speaker 3: a vector. I mean, in ubiquitous technical surveillance, you have
- 15:23Speaker 3: multiple vectors that overlap that make the thing a threat
- 15:26Speaker 3: or a problem for you. And it seems that it
- 15:29Speaker 3: goes beyond the vector problem when you're using AI. Like
- 15:32Speaker 3: let's say you're booking on Airbnb, and there's certain behavioral
- 15:34Speaker 3: patterns that you the user use on Airbnb that are
- 15:37Speaker 3: almost exactly the same behaviors or exactly the same behaviors
- 15:40Speaker 3: when you're undercover under a different persona on some other platform,
- 15:43Speaker 3: and AI is able to see that and correlate you
- 15:45Speaker 3: immediately way faster than a human analyst could. I already mentioned,
- 15:49Speaker 3: you know, what would you say to the average American.
- 15:51Speaker 3: But I'm curious also to the average case officer who
- 15:53Speaker 3: might be listening, because I know there are some who
- 15:54Speaker 3: do listen to the show. What would you kind of
- 15:56Speaker 3: caution them on?
- 15:59Speaker 1: Yeah, I mean, well, look, even for the average American,
- 16:02Speaker 1: by the way, I don't think the average American wants
- 16:05Speaker 1: a Chinese company or a Chinese security service or the
- 16:10Speaker 1: Chinese government to know what they're doing and to be
- 16:14Speaker 1: able to match the records from Airbnb to other things
- 16:17Speaker 1: in their life. We don't want our government doing that, right,
- 16:20Speaker 1: Like your average American like me, is like at least
- 16:24Speaker 1: a little bit libertarian leading in that sense. We don't
- 16:27Speaker 1: want our government to spy on us. Why would we
- 16:29Speaker 1: want the Chinese government to spy on us?
- 16:31Speaker 2: Right?
- 16:31Speaker 1: So I think everybody, But yeah, for case officers. Look,
- 16:34Speaker 1: I was a case officer, you know, and by the
- 16:38Speaker 1: time I stopped doing that, the UTS issue is just starting, right,
- 16:43Speaker 1: and so I only really saw a little bit of it,
- 16:46Speaker 1: and I can't imagine, starting from scratch today what it
- 16:50Speaker 1: might must be like. And the you know, even without AI,
- 16:56Speaker 1: you can match people's patterns right, just literally, like old
- 17:00Speaker 1: school big data analytics, I crunch enough data about you,
- 17:04Speaker 1: I can detect patterns. The AI makes it easier and
- 17:10Speaker 1: more ubiquitous, right, Like I don't even need to you know,
- 17:14Speaker 1: I can. I can do it like that example you
- 17:16Speaker 1: brought up as a great example. Actually, like my habits.
- 17:20Speaker 1: It might not, by the way, even be what I'm
- 17:23Speaker 1: doing on Airbnb. It could even be like the way
- 17:26Speaker 1: I move my mouse or the way I type. Right,
- 17:30Speaker 1: there are signatures that are everybody has. These patterns can
- 17:35Speaker 1: be taken from one place and then and then applied
- 17:39Speaker 1: to another. So as a case officer, you're trying to
- 17:41Speaker 1: act and cover somewhere. And you know, even if you
- 17:47Speaker 1: even if you use a false name for everything you
- 17:50Speaker 1: did in work, and in fact, even if you used
- 17:53Speaker 1: a false name for everything you did in your commercial life,
- 17:57Speaker 1: well your pattern still might be there. Right, Like maybe
- 18:01Speaker 1: you're using your wife's account to log into Airbnb, but
- 18:04Speaker 1: you're the one on the mouse and making the decisions.
- 18:08Speaker 1: It would still be able to recognize these patterns. So
- 18:12Speaker 1: it's like becomes nearly impossible to kind of like keep
- 18:17Speaker 1: yourself out of there now. I actually wrote a whole
- 18:19Speaker 1: chapter about this in my book because like I think,
- 18:22Speaker 1: what where it goes to be a case officer. I
- 18:25Speaker 1: think you just have to sort of assume that anything,
- 18:29Speaker 1: any pattern is or could be tracked, and then you
- 18:32Speaker 1: I think we will enter a world where you just
- 18:34Speaker 1: have to sort of live in pattern and what it
- 18:37Speaker 1: is to kind of be and cover and so forth
- 18:39Speaker 1: completely changes. Like there's no more doc swapping and all
- 18:42Speaker 1: that stuff we saw in spy movies, Like that's over
- 18:45Speaker 1: and like now now you just have to sort of
- 18:48Speaker 1: live your life as that person, like kind of closer
- 18:50Speaker 1: to what it was like to be a source back
- 18:52Speaker 1: in the day.
- 18:54Speaker 2: Hello everyone, if you're hearing this you are not a
- 18:57Speaker 2: member of the Patreon you can get your hands on
- 19:02Speaker 2: ad free episodes that come out early and you can
- 19:05Speaker 2: ask us questions maybe we'll answer them. You can also
- 19:07Speaker 2: watch the team House podcast completely ad free and early,
- 19:11Speaker 2: and you can also watch it live as we interview
- 19:13Speaker 2: our guests, and it's the best way to support the channel.
- 19:17Speaker 2: So if you can, if you're open to it, you
- 19:19Speaker 2: can go to Patreon dot com slash the Teamhouse. That
- 19:22Speaker 2: link is in the description. We appreciate the supporters and
- 19:26Speaker 2: we hope you would consider being one. So it's Patreon
- 19:29Speaker 2: dot com slash the Teamhouse. Thanks guys.
- 19:33Speaker 3: Yeah, Actually, I was reading about when the strikes are
- 19:36Speaker 3: going on February twenty eighth and forward in Iran at
- 19:38Speaker 3: the very beginning of the war, how Israel used some
- 19:41Speaker 3: security cameras and some uts access they had inside of
- 19:44Speaker 3: Iran to better triangulate targets and then predict follow on
- 19:47Speaker 3: targets very rapidly, which is fantastic. But then the question
- 19:50Speaker 3: is that's with the AI that we have right now.
- 19:53Speaker 3: What happens when it's an order of magnitude either more
- 19:55Speaker 3: intelligent or faster or in the wrong hands, you know what,
- 20:00Speaker 3: how do we develop safeguards for that in a military context?
- 20:05Speaker 1: Yeah, Or what happens if it's applied to us here
- 20:09Speaker 1: right where every you know, every factory in America has
- 20:14Speaker 1: some electronic device in it that might be, you know,
- 20:17Speaker 1: could be compromised. And China has an AI system that's
- 20:21Speaker 1: able to sort of assess and analyze all of those factories,
- 20:25Speaker 1: all the people working on them, and then do do
- 20:29Speaker 1: something about it. It could be you're right, it could
- 20:32Speaker 1: be using that for targeting information, right. It could be
- 20:36Speaker 1: they can figure out just by these assessment features, you know,
- 20:41Speaker 1: which ones are building key you know, components in a missile.
- 20:46Speaker 1: It also could be well, why not just target the
- 20:48Speaker 1: people working at the factories non kinetic means, right, Like
- 20:53Speaker 1: why not use an information attack against you know, sometimes
- 20:59Speaker 1: in these factories for exams, you have like one or
- 21:01Speaker 1: two guys who are the key guy, like they know
- 21:03Speaker 1: how to do everything. Why don't I just target that guy?
- 21:06Speaker 1: Why don't I implant some criminal information on that guy's
- 21:10Speaker 1: computer and then put him in prison? And now that
- 21:14Speaker 1: factory is fifty percent as effective because we've lost that guy, right, Like,
- 21:19Speaker 1: there's all sorts of things that can be done if
- 21:22Speaker 1: you you know, if you're sneaky thinker with this stuff.
- 21:27Speaker 2: I question into going back a little bit to uh,
- 21:30Speaker 2: you know, being worried about American companies and American startups
- 21:34Speaker 2: using the Chinese models that are some places I've read
- 21:38Speaker 2: it's like one third cheaper and stuff like that. How
- 21:40Speaker 2: do we stop? And there was like how do we
- 21:42Speaker 2: stop basically it being so expensive for American companies to
- 21:47Speaker 2: use frontier models?
- 21:49Speaker 1: Yeah, yeah, the the first of all I would start
- 21:56Speaker 1: to question some of the like costs different is right,
- 22:00Speaker 1: just like like with anything open source, like the actual
- 22:05Speaker 1: cost to buy it versus like the core cost to implement.
- 22:08Speaker 1: It might be totally different, but let's just assume from it.
- 22:11Speaker 1: And I think I think it is true. It is
- 22:12Speaker 1: going to be it is going to be cheaper. I
- 22:15Speaker 1: think we need basically cheaper models. And I have this
- 22:19Speaker 1: idea which I'll throw out right here, which is the
- 22:24Speaker 1: Trump administration has made for its own reasons, has has
- 22:30Speaker 1: said to Anthropic that there's a problem. They don't want
- 22:34Speaker 1: it to be used in Department of War. We are
- 22:38Speaker 1: where we are, well. I think at some point we
- 22:42Speaker 1: should make a deal with Anthropic, get them back into
- 22:46Speaker 1: the into the Department of War. But I think we
- 22:50Speaker 1: should say in exchange, one of the things that we
- 22:52Speaker 1: want is you should make an open source model that
- 22:56Speaker 1: is or or just like a really cheap model that's
- 22:58Speaker 1: really good.
- 23:00Speaker 2: Right.
- 23:00Speaker 1: There are ways that we can work with these companies
- 23:05Speaker 1: and make deals so that it's good for the nation
- 23:08Speaker 1: and it's good for the government, it's good for the company.
- 23:10Speaker 1: And I think that's how we're going to have to
- 23:12Speaker 1: do it. And I see a little bit of that happening,
- 23:14Speaker 1: like if you sort of read between the lines. After
- 23:17Speaker 1: all this stuff came out about Chinese open weight models,
- 23:21Speaker 1: and I can assure you, as somebody is sitting inside
- 23:24Speaker 1: the Washington DC Beltway right now, that people here read
- 23:28Speaker 1: those articles and got really really concerned about it. You
- 23:31Speaker 1: could see a few of these companies started offering their
- 23:35Speaker 1: models for cheaper so like, they're not dumb, they're listening,
- 23:39Speaker 1: and I just think that that's how we're going to
- 23:41Speaker 1: have to deal with this. I also think that frankly,
- 23:44Speaker 1: like I said, I think that the US government should
- 23:47Speaker 1: make it illegal in fact to use Chinese models, that
- 23:51Speaker 1: you just cannot use them here, that they compromise their
- 23:55Speaker 1: security threats for cybersecurity and various things. They're informational threat
- 24:00Speaker 1: to the American people, and we should just not be
- 24:02Speaker 1: able to use them. And you know, we could use
- 24:05Speaker 1: them for research purposes. And if you're an individual, I
- 24:09Speaker 1: don't think we should be telling individuals what they can
- 24:11Speaker 1: and cannot use. But if you're a business you want
- 24:14Speaker 1: to sell a product, you should not be able to
- 24:17Speaker 1: use it. And we've done this before with Huawei, where
- 24:21Speaker 1: we have effectively gone to American telecom operators and been
- 24:24Speaker 1: like no, like we're done not here, and I think
- 24:28Speaker 1: we need to do the same thing with open source models.
- 24:31Speaker 3: We actually did something really similar with Huawei with our
- 24:33Speaker 3: foreign partners too. I remember I was overseas in working
- 24:35Speaker 3: in the embassies and during the Trump one administration, and
- 24:39Speaker 3: there was a huge push throughout Europe and other partners
- 24:41Speaker 3: as well to hey, please abandon Huawei, please use US technology.
- 24:45Speaker 3: We'll create incentives for you to do that, But it
- 24:48Speaker 3: also requires buy in from the partner because they do
- 24:50Speaker 3: lose a little bit. It costs them more money to
- 24:52Speaker 3: adopt this despite the other non monetary incentives that the
- 24:56Speaker 3: US can offer them. What do you think about those
- 24:59Speaker 3: kind of incentives?
- 24:59Speaker 1: For Ai agreed, I'd run the exact same play I would.
- 25:04Speaker 2: I would.
- 25:05Speaker 1: I would literally send out, you know, US Department of
- 25:08Speaker 1: Commerce and used Department of State state folks to go
- 25:11Speaker 1: to these countries and been like, you should not use
- 25:14Speaker 1: this stuff, and we have other levers. By the way,
- 25:17Speaker 1: there's carrots and sticks that can be applied. But part
- 25:21Speaker 1: of that would also be, hey, well, you know here
- 25:24Speaker 1: we have some less expense. You know, here are some
- 25:27Speaker 1: less expensive models that are American that you can use.
- 25:30Speaker 1: And I think we work with the US companies. By
- 25:32Speaker 1: the way, it's it's in the interest of these big
- 25:36Speaker 1: companies like Anthropic and Open Ai and and Xai and
- 25:40Speaker 1: so forth, right two to offer these models for less money.
- 25:46Speaker 1: They it's a business issue for them too, because if
- 25:49Speaker 1: all of these startups in the US and Europe or
- 25:51Speaker 1: wherever start using Chinese open weight models, now they're and
- 25:56Speaker 1: those companies, assumably over time will get bigger and bigger.
- 25:59Speaker 1: Now they're not going to be customers of Anthropic and
- 26:02Speaker 1: open Ai and so forth, So those companies are incentivized
- 26:05Speaker 1: as well, and like, we just need to get everybody
- 26:08Speaker 1: on the same page.
- 26:10Speaker 3: It actually reminds me a lot of when the iPhone
- 26:11Speaker 3: first came out. I think it was two thousand and
- 26:13Speaker 3: seven when I had the very first model and there
- 26:16Speaker 3: was no app store at the time, and you could
- 26:18Speaker 3: just jail break your phone and get all these apps
- 26:20Speaker 3: on your phone. And then Apple saw that and they
- 26:22Speaker 3: were like, hold on a minute, this is our device,
- 26:24Speaker 3: and they create came up with the concept of the
- 26:26Speaker 3: wald Garden, where in order to play on an Apple device,
- 26:29Speaker 3: you had to be registered on Apple device and all
- 26:31Speaker 3: these special safeguards. And I wonder if there's like some
- 26:33Speaker 3: way to do that with AI, at least in the
- 26:35Speaker 3: United States, like you were mentioning with commerce, there might
- 26:37Speaker 3: be some way to create that walled garden where the
- 26:40Speaker 3: incentives inside the wall are much higher than outside the wall.
- 26:43Speaker 3: But you do have to exchange something of your sovereignty
- 26:46Speaker 3: a little bit to be able to play inside there.
- 26:49Speaker 1: Yeah, I mean these companies I suspect over time will
- 26:53Speaker 1: naturally create wald gardens like a great way to make money.
- 26:57Speaker 1: But you're right, I mean, in a sense, we did
- 26:58Speaker 1: this with foreign government where we said, you know, I
- 27:01Speaker 1: remember that time period in Trump one where it's like, okay, well,
- 27:05Speaker 1: if you want to have Huawei equipment on your system,
- 27:07Speaker 1: we're not going to share any classified information with you.
- 27:11Speaker 1: So if you want to play in our walled garden,
- 27:14Speaker 1: then you've got to go out and you've got to
- 27:17Speaker 1: put in like real routers and so forth, because we're
- 27:20Speaker 1: not going to lose our information. And it could be
- 27:22Speaker 1: the same. You could imagine it in terms of other
- 27:27Speaker 1: systems like banking systems and other commercial systems and saying
- 27:31Speaker 1: that's great, you're using an open source model. I guess
- 27:34Speaker 1: we can't legally stop you from doing it. But you
- 27:37Speaker 1: don't get to play in swift. You don't get to
- 27:40Speaker 1: play in like the banking system because we're not going
- 27:42Speaker 1: to accept the additional risk of you using apps that
- 27:45Speaker 1: were developed by a Chinese system. That might be a
- 27:48Speaker 1: way to do it.
- 27:49Speaker 3: And I'm also with you too, like I'm a less
- 27:51Speaker 3: regulations guy, but I also look at, especially in the
- 27:54Speaker 3: international community, when you have in the international relations theory
- 27:57Speaker 3: of how governments are black boxes and so on, how
- 28:00Speaker 3: do you regulate these states? Maybe regulation's wrong, Maybe there's
- 28:04Speaker 3: some other way at that higher level where an individual
- 28:07Speaker 3: company alone cannot be the one on the frontier changing
- 28:10Speaker 3: things for the good. You can't depend on them to
- 28:13Speaker 3: do that. How can states jump in there besides the
- 28:16Speaker 3: United States because it has to be all nations. Basically
- 28:18Speaker 3: this is a this is a global question. How can
- 28:21Speaker 3: we create incentives for states to monitor or keep safe
- 28:26Speaker 3: this development?
- 28:29Speaker 1: Yeah, this is this is like we're in this sort
- 28:32Speaker 1: of weird new world, right where like states almost now
- 28:35Speaker 1: have to think like companies and vice versa. Right, these
- 28:39Speaker 1: like anthropic is or any hyperscaler right, like Google almost
- 28:46Speaker 1: has to act like a state and deal with state threats.
- 28:49Speaker 1: And at the same time, our state, the government is
- 28:53Speaker 1: now thinking in terms of these like commercial motivations, and
- 28:57Speaker 1: there's no like I did a PhD in international relations,
- 29:01Speaker 1: and this when I did my PhD twenty years ago,
- 29:06Speaker 1: twenty something, you're like, none of this existed. There's no
- 29:08Speaker 1: theory behind this, there's no like playbook to go by,
- 29:12Speaker 1: and so everybody's winging it. And I think like Trump won,
- 29:16Speaker 1: Biden now Trump two, they've all sort of tried to
- 29:19Speaker 1: muddle through of how do you actually do this? And
- 29:22Speaker 1: it does seem to be like there's some hard edged
- 29:27Speaker 1: level where the government goes and says US government goes
- 29:31Speaker 1: and says you must do this, like we are not
- 29:33Speaker 1: going to share classified information with you unless you fix this.
- 29:37Speaker 1: So there is like a you know, hard edge old
- 29:40Speaker 1: school power level. Then there's like this sort of middle
- 29:46Speaker 1: kind of like softer edged power, like incentive structures like hey,
- 29:51Speaker 1: we'll subsidize some of what we're doing, or hey, I
- 29:55Speaker 1: you know, we see you want to refinance your you know,
- 29:58Speaker 1: your your World Bank debt or you know some you
- 30:01Speaker 1: know you want something, you want some diplomatic action from US. Okay, well,
- 30:06Speaker 1: like now you know Huawe's on the table or AI
- 30:10Speaker 1: is on the table, right, so we're going to come
- 30:12Speaker 1: up with like a negotiation that you get what you want,
- 30:16Speaker 1: We'll get what we want. And then there's like this
- 30:19Speaker 1: commercial level where these companies I think are in the
- 30:23Speaker 1: US they're being regulated and they want to play nice
- 30:27Speaker 1: with the government too, and where you know, the government
- 30:30Speaker 1: can kind of say, look, look, we we don't want
- 30:34Speaker 1: to regulate your business here, but we need a deal.
- 30:37Speaker 1: And part of that deal is we think you should
- 30:39Speaker 1: offer to these allies a lower cost version of what
- 30:42Speaker 1: you're doing, for example, So there's like a negotiation with
- 30:45Speaker 1: the companies as well. I think there's like those are
- 30:47Speaker 1: sort of the levels and this is what now, you know,
- 30:51Speaker 1: a national security advisor or president's chief of staff has
- 30:55Speaker 1: to kind of like fix, you know, muddle through.
- 30:58Speaker 3: And earlier in the show, D asked about frontier models.
- 31:01Speaker 3: I'm curious how do those fit into this conversation we're
- 31:03Speaker 3: having here, because it seems that frontier means that it's changing,
- 31:06Speaker 3: you know, by the minute. How do how do we
- 31:08Speaker 3: deal with that?
- 31:11Speaker 1: Yeah, there's like this added danger in essence of the
- 31:15Speaker 1: frontier models is that they're so capable, right like if
- 31:19Speaker 1: you use fable or something, it's so capable that it
- 31:22Speaker 1: can do things that are inherently dangerous. And there's this
- 31:27Speaker 1: question of one you sort of it seems like kind
- 31:30Speaker 1: of have to inoculate yourself. Like that was sort of
- 31:34Speaker 1: what was going on with mythos, and they were sort
- 31:36Speaker 1: of like, Okay, let's release this to only a few people,
- 31:42Speaker 1: some big companies and the government so they can kind
- 31:44Speaker 1: of inoculate the cybersecurity systems so that they don't immediately
- 31:48Speaker 1: get hacked, right, So that you have to sort of
- 31:50Speaker 1: do that, and then you also have to control them,
- 31:55Speaker 1: lock them down at some level. And that was also
- 31:59Speaker 1: what happened with fable right it got released, like you said,
- 32:04Speaker 1: it was released for like two days. Trump administration saw
- 32:07Speaker 1: that like, wait, people are using this in a bad way.
- 32:10Speaker 1: You need to get that back out, fix it before
- 32:13Speaker 1: you send it out. So you have to like lock
- 32:15Speaker 1: them down a little bit. And that, like, I think
- 32:20Speaker 1: that that is the future of these things. I think
- 32:22Speaker 1: these models are so dangerous. It'd be like imagine if
- 32:28Speaker 1: you know, when we first developed you know, you know,
- 32:33Speaker 1: the American car industry in the fifties, and we had
- 32:37Speaker 1: nuclear power plants on these things, and it's like some
- 32:41Speaker 1: people figured out how to blow up these cars and
- 32:43Speaker 1: blow up a city block. You'd be like wait, wait, wait, wait, wait,
- 32:46Speaker 1: you can't sell it like that, right. It's sort of
- 32:49Speaker 1: like it's like we've created these products that are truly dangerous,
- 32:53Speaker 1: and for better or worse, we regulate firearms and things
- 32:57Speaker 1: like this as well, for like the same reason, like
- 32:59Speaker 1: they're dangerous, and so I think we're going to have
- 33:02Speaker 1: to do it with AI. And then and then you
- 33:05Speaker 1: come up with a question, well, then what happens if
- 33:07Speaker 1: some other country offers a product that is non regulated,
- 33:10Speaker 1: like deep Seek is giving these open weight models. Well,
- 33:14Speaker 1: you know, if you ask me, they you're selling a
- 33:16Speaker 1: product in America, it should be regulated like any other
- 33:19Speaker 1: American product, even if you're giving it away for free,
- 33:23Speaker 1: like you you still have to go through the same
- 33:27Speaker 1: you know, the same process.
- 33:29Speaker 2: Yeah.
- 33:29Speaker 3: And actually the nuclear analogy, I've been thinking about this too,
- 33:32Speaker 3: because in the twenties and thirties when nuclear technology generally
- 33:36Speaker 3: was being discovered, essentially there was the weaponization of it,
- 33:40Speaker 3: but there was also the innovation of it, you know,
- 33:42Speaker 3: and one leads to potentially unlimited energy with fusion reaction,
- 33:46Speaker 3: and one leads to the destruction of the human race
- 33:48Speaker 3: and the entire planet. But when you're first starting out,
- 33:51Speaker 3: you know, splitting that first atom that's not really what's
- 33:53Speaker 3: on your mind. Instead, you're just fascinated by this new
- 33:55Speaker 3: development and how far and positive it could go. And
- 33:58Speaker 3: there's a book called Irrational exuber It's that really talks
- 34:01Speaker 3: about this in the market, where people just have this idealistic,
- 34:04Speaker 3: rosy eyed view of everything in the future that they're
- 34:07Speaker 3: touching and it actually doesn't work out that way usually.
- 34:10Speaker 3: How far do you think the gap is between what
- 34:12Speaker 3: will probably happen versus this irrational exuberance about how AI
- 34:15Speaker 3: will change everything for the better.
- 34:19Speaker 1: Yeah, I mean I'm a bit of an optimist pessimist
- 34:22Speaker 1: in one, right, Like I believe I just think about
- 34:26Speaker 1: in terms of the optimism, Like, clearly AI is going
- 34:31Speaker 1: to create and do amazing things. Right, We're going to
- 34:35Speaker 1: use it for medicine, and it's going to make better medicine.
- 34:38Speaker 1: We're going to use it just literally to make companies
- 34:42Speaker 1: in the government more efficient, which will save us all
- 34:45Speaker 1: sorts of money. Right, Like they're like wonderful things that are
- 34:49Speaker 1: gonna happen. It's also gonna you know, it's also going
- 34:52Speaker 1: to be really really bad, right, Like people could hack
- 34:56Speaker 1: into things, people might make bioweapons. It's probably going to
- 35:01Speaker 1: put all sorts of people out of work, right, and
- 35:04Speaker 1: so it's like any other technology. It's like saying the internet.
- 35:07Speaker 1: The Internet was great and people got to communicate and
- 35:09Speaker 1: we got all sorts of memes and fun stuff out
- 35:12Speaker 1: of it, but also at the same time we got
- 35:14Speaker 1: cyber hacking and we got cyber bullying and like all
- 35:18Speaker 1: of these bad things. So I think it will be
- 35:20Speaker 1: the same. And the I think the question we're going
- 35:24Speaker 1: to face is how do we control something like AI
- 35:29Speaker 1: And like even the Internet, which can seem pretty uncontrolled,
- 35:34Speaker 1: like does have some controls right, Like there there there
- 35:37Speaker 1: are some minimal controls. We usually go through corporations to
- 35:41Speaker 1: make those controls. Like the actual laws governing are pretty weak.
- 35:46Speaker 1: But historically we've done this again and again. In Japan
- 35:51Speaker 1: for two hundred years, the Japanese made firearms illegal, and
- 35:57Speaker 1: you know, think of that, like for two hundred years,
- 35:59Speaker 1: they have the technology. They fought wars with firearms. Not
- 36:03Speaker 1: a lot of people realize this, but like Samurais were
- 36:06Speaker 1: fighting with with rifles and then and then basically this
- 36:11Speaker 1: this one you know, Samurai leader Won Tokugawa and he
- 36:16Speaker 1: said that's it, like we're not using firearms and you
- 36:19Speaker 1: can only use them for hunting. That's all they used
- 36:21Speaker 1: them for, was like hunting for two hundred years. So
- 36:25Speaker 1: it can be done. Societies can control technologies or think
- 36:30Speaker 1: of like the printing press, like there's been all sorts
- 36:32Speaker 1: of attempts. Sometimes you try to control technologies and it
- 36:35Speaker 1: doesn't work right. The you know, we used to send
- 36:38Speaker 1: copy machines to the Soviet Union to like undermine, you know,
- 36:42Speaker 1: the Soviet government because we wanted people to pass around
- 36:46Speaker 1: secret information, not secret like information they didn't want passed
- 36:51Speaker 1: around like political information. So I think we're gonna, you know,
- 36:55Speaker 1: we're sort of entering a period of history in America
- 36:58Speaker 1: in particular where we're gonna have to figure out how
- 37:00Speaker 1: to control this technology before you know, the downside outweighs
- 37:05Speaker 1: a good side.
- 37:06Speaker 2: Antony, I got a question, what do you make of
- 37:08Speaker 2: the people that are saying that all the talk about
- 37:11Speaker 2: mythos and fable and then I think open Ai came
- 37:13Speaker 2: out with their version in terms of saying, oh, yeah,
- 37:16Speaker 2: it's very dangerous that being all a marketing kind of
- 37:21Speaker 2: hype thing to keep juice in their valuations.
- 37:25Speaker 1: I mean, if you ask me personally, like I think
- 37:29Speaker 1: probably both things are true, like that it is absolutely
- 37:33Speaker 1: a marketing hype thing. The timing was not accidental on
- 37:38Speaker 1: the release of you know, the Mythos model to you know,
- 37:41Speaker 1: as Anthropic was going through some stressful times with the
- 37:46Speaker 1: government to remind the government that this was like a
- 37:49Speaker 1: really powerful technology and they wanted access to it, right,
- 37:52Speaker 1: And then it wasn't accidental that chat cheet you know
- 37:55Speaker 1: that that open Ai released its model that did the
- 37:58Speaker 1: same thing not too long afterwards to remind everybody, hey,
- 38:01Speaker 1: we also have a powerful model, right, Like I think
- 38:04Speaker 1: that's probably true just me reading you know, the news
- 38:08Speaker 1: about these things, but also it's real, Like I mean,
- 38:12Speaker 1: I think you can just hop on these models right
- 38:14Speaker 1: now and see how powerful they are. Like it feels like,
- 38:18Speaker 1: you know agi and we just don't want to say it,
- 38:21Speaker 1: Like it feels like they've created models now that can
- 38:25Speaker 1: do almost anything that a person can do other than
- 38:29Speaker 1: maybe the most talented super geniuses in the world.
- 38:33Speaker 2: And so.
- 38:35Speaker 1: I do I do think like even if it was,
- 38:38Speaker 1: for example, let's say Mythos wasn't the best cybersecurity operator
- 38:45Speaker 1: our nation has ever produced, like you know, the best
- 38:47Speaker 1: guy at NSA or something, but it was just like
- 38:50Speaker 1: a pretty good guy, like it was a pretty good hacker,
- 38:53Speaker 1: like somebody who had ten years experience. Well, just that
- 38:56Speaker 1: you can do it at scale, for example, that you
- 38:58Speaker 1: can do it doesn't sleep, it doesn't rest, you can
- 39:01Speaker 1: spin up as many agents and instances as you want.
- 39:05Speaker 1: And like, just that alone makes it dangerous. And as
- 39:10Speaker 1: it gets better and better, and I do think it will,
- 39:12Speaker 1: you know, these things are probably closing in on being
- 39:15Speaker 1: as good as any person is, then you know it
- 39:20Speaker 1: is truly powerful.
- 39:22Speaker 3: We talked a lot about the threat the product might pose.
- 39:25Speaker 3: But I'm curious from your perspective. You know, you've worked
- 39:27Speaker 3: in finance. You we're a managing director, You've been at
- 39:29Speaker 3: Bridgewatery've been at Service. There's a lot of circular financing
- 39:32Speaker 3: going on that a lot of a lot of articles
- 39:35Speaker 3: like Financial Times, Wall Street Journal, like this is actually
- 39:36Speaker 3: like a weekly thing that many articles about this, a
- 39:39Speaker 3: lot of huh. And I'm curious what your view is
- 39:42Speaker 3: of that, you know, analytically, from what you what you
- 39:45Speaker 3: know from the market as far as value circularity, especially
- 39:49Speaker 3: from Nvidia investing in a lot of these other firms
- 39:51Speaker 3: that have very significant interests in Nvidia and all these
- 39:54Speaker 3: kind of things.
- 39:55Speaker 1: Yeah, I mean, how do you put together a trillion
- 39:59Speaker 1: dollar as a capex or four trillion? I mean I've
- 40:02Speaker 1: heard different numbers like how are you going to finance that?
- 40:06Speaker 1: And in a sense, you know, just taking a step back,
- 40:10Speaker 1: it's actually one of America's advantages versus China is that
- 40:15Speaker 1: we have the ability to finance just about anything right privately.
- 40:21Speaker 1: Like think of that, all this finance is happening without
- 40:25Speaker 1: really any tax dollars, just happening privately, and we're building it.
- 40:30Speaker 1: And so how do you you know, you can look
- 40:32Speaker 1: back at other times we've financed like huge capex. So
- 40:36Speaker 1: the railroads, like in the eighteen hundreds, that was probably
- 40:39Speaker 1: an equivalent like how do you build railroads across all
- 40:42Speaker 1: of America which are eventually going to unlock an immense
- 40:46Speaker 1: amount of value? Well, like railroads, like they had to
- 40:49Speaker 1: use bonds and things like this to do that. Then
- 40:52Speaker 1: you look at the interstate highway system, right and like
- 40:56Speaker 1: partially private, privately funded, but like the majority of it
- 41:01Speaker 1: was government funded. And by the way, Eisenhower like part
- 41:05Speaker 1: of how we got that through Congress is he turned
- 41:08Speaker 1: it into a national security issue. Not a lot of
- 41:11Speaker 1: people have heard this, but like the reason bridges over
- 41:15Speaker 1: interstates or the height that they are is that they
- 41:18Speaker 1: had to be able to fit a truck holding a
- 41:21Speaker 1: nuclear missile under them, like that they're high enough. That's
- 41:26Speaker 1: how they determine that height. And so he was like, well,
- 41:29Speaker 1: in the event of nuclear war, we might have to
- 41:31Speaker 1: drive these missiles around. And so this is a new
- 41:34Speaker 1: you know, this is like part of our you know,
- 41:36Speaker 1: warfare logistics system. And that's how that's partially how we
- 41:39Speaker 1: got that money together. And then you look at like Telecom,
- 41:42Speaker 1: like at and T. We let them be a monopoly
- 41:45Speaker 1: for like almost a century because we knew that they
- 41:49Speaker 1: had so much capex. So now I think we're in
- 41:51Speaker 1: another situation where we have an immense amount of capex
- 41:55Speaker 1: that we require to have this big, big future payoff, right,
- 42:00Speaker 1: And how do you put that money together? Nobody wants
- 42:03Speaker 1: to use the government to do it, and so they're
- 42:05Speaker 1: finding whatever financing means that they can do, and like
- 42:09Speaker 1: they've happened on this like it is a circular financing system.
- 42:14Speaker 1: And then you have to ask yourself, Okay, do we
- 42:17Speaker 1: really believe that it will be as big as you
- 42:20Speaker 1: say it will at the end? And if so, this
- 42:21Speaker 1: is a great deal, right, Like you kind of look
- 42:24Speaker 1: at it and you're like, well, actually, I think, like
- 42:27Speaker 1: you know, Anthropic is a steal at two trillion dollars,
- 42:32Speaker 1: like and this is going to be a twenty trillion
- 42:33Speaker 1: dollar company and they should spend any amount of money
- 42:36Speaker 1: they want on Capex and it doesn't matter. That is
- 42:41Speaker 1: like the personal assessment of everybody invests in these companies
- 42:44Speaker 1: to make that call.
- 42:47Speaker 3: Right. Actually, the railroad example is excellent because a lot
- 42:50Speaker 3: of the Panics in the nineteenth century after the railroads
- 42:53Speaker 3: started was because of these bondishuances and speculation on future
- 42:57Speaker 3: value that could really it was hard to calculate. And
- 43:00Speaker 3: like the Panic of nineteen oh three, for example, when
- 43:02Speaker 3: JP Morgan stepped in and basically save the US banking
- 43:04Speaker 3: system voluntarily as a private citizen. I mean, that's fantastic.
- 43:08Speaker 3: I wonder is it possible now, especially with things like
- 43:11Speaker 3: Blue Owl and other private credit runs on private credit,
- 43:14Speaker 3: which are very interesting derivatives of the two thousand and
- 43:17Speaker 3: seven financial crisis of you know, actual banks doing this
- 43:19Speaker 3: and now it's non banks that are doing it. I'm
- 43:22Speaker 3: curious if that did happen. Let's say doomsday happens, there's
- 43:24Speaker 3: a panic market failure. Are there private individuals like the Carnegie,
- 43:29Speaker 3: like Jay Gould, like JP Morgan that could step in
- 43:32Speaker 3: and actually help.
- 43:36Speaker 1: H Well, I don't know if they. Yeah, that's a
- 43:40Speaker 1: great question. That's an interesting one to think about. I meane,
- 43:45Speaker 1: one interesting thing is that the amount of wealth in
- 43:48Speaker 1: private hands right now is huge. Like that you know,
- 43:52Speaker 1: we had an individual who was for a brief minute
- 43:54Speaker 1: a trillionaire, is like incredible to even comprehend. And so
- 44:00Speaker 1: you know, in some sense you could kind of argue
- 44:04Speaker 1: that these circular financing deals are a little bit of
- 44:07Speaker 1: this right where they're like, hey, we know that this
- 44:12Speaker 1: infrastructure has to be out there. Our personal valuations are
- 44:17Speaker 1: based on this, you know, whether I'm Larry Ellison or
- 44:21Speaker 1: Elon Musk or like choose your choose your billionaire. So
- 44:26Speaker 1: I'm going to We're going to figure out some way
- 44:29Speaker 1: to finance these things and take like huge amounts of
- 44:34Speaker 1: risk if it all comes tumbling down, Like what what
- 44:39Speaker 1: will happen? I don't know. I think like the real
- 44:42Speaker 1: danger today that's different from back then is like everybody
- 44:45Speaker 1: who has a pension is now exposed to these companies
- 44:50Speaker 1: and so in the case that you know, there was
- 44:53Speaker 1: a tumbling in stock market valuation, like that would really
- 44:56Speaker 1: take a toll, especially on like retirees, right like could
- 45:00Speaker 1: and then we've been in a high inflation environment for
- 45:03Speaker 1: a while and so things are just like literally more expensive,
- 45:07Speaker 1: which is okay if your pension is doing really well,
- 45:10Speaker 1: but if your pension takes a tumble, and so you know, look,
- 45:14Speaker 1: I'll tell you this. I think that and maybe we
- 45:17Speaker 1: can get into it. I think that it's inevitable that
- 45:22Speaker 1: these companies are at least partially nationalized because of all
- 45:26Speaker 1: of this, because there are such a possible threat to
- 45:29Speaker 1: the economy, but also that the upside risk is potentially
- 45:34Speaker 1: so high that these companies may be worth so much
- 45:38Speaker 1: that the American people are going to want a piece
- 45:41Speaker 1: of that action, and the government's going to want a
- 45:43Speaker 1: piece of that action, and like taxes are not really
- 45:45Speaker 1: capturing it. And that there is this feeling which I
- 45:49Speaker 1: think is rightfully so that these AI models are built
- 45:51Speaker 1: upon all of our knowledge, like as individuals and historically
- 45:57Speaker 1: going back centuries, so sort of you in a sense
- 46:00Speaker 1: belong to some of us. So I think it's inevitable
- 46:02Speaker 1: these things to get nationalized.
- 46:04Speaker 3: That's actually an interesting point because at and T that
- 46:07Speaker 3: you brought up earlier, At and T and the National
- 46:09Speaker 3: Security Agency had a very close relationship since the nineteen
- 46:11Speaker 3: forty seven National Security Act that formed the National Security Agency,
- 46:14Speaker 3: and it only grew closer over time. And I immediately
- 46:18Speaker 3: think about with AI, if it was nationalized, probably a
- 46:21Speaker 3: very similar activity would happen where there is an embedded
- 46:23Speaker 3: way where the US intelligence community has some kind of
- 46:26Speaker 3: very special access to this, you know, amazing technology that
- 46:31Speaker 3: the average person does not and other states do not,
- 46:33Speaker 3: and it would create these whole new counterintelligence problems for
- 46:36Speaker 3: the United States obviously, but of course there'd benefits at
- 46:39Speaker 3: the same time. It's just kind of an interesting thought
- 46:40Speaker 3: experiment to see, like, where does this go, the symbiosis
- 46:43Speaker 3: between national security and a nationalized asset that is so powerful.
- 46:49Speaker 1: And you might even read the you know what's kind
- 46:52Speaker 1: of happened with AI over the last year and a
- 46:55Speaker 1: half or so or two years and say, well, it's
- 46:58Speaker 1: already kind of happening, right, Like the all of the
- 47:02Speaker 1: big model companies were tripping over themselves to do a
- 47:06Speaker 1: deal with the Department of War, right even though as
- 47:10Speaker 1: like a market. It's not going to be their biggest market.
- 47:14Speaker 1: Like everybody thinks the Department of Ward trillion dollar budget
- 47:18Speaker 1: is big, Like a trillion dollar budget is like you know,
- 47:22Speaker 1: I don't know how much you Americans spent on you know,
- 47:25Speaker 1: snack food per year. I'm sure it's more than a
- 47:28Speaker 1: trillion dollars, right, Like it's it's like, it's not actually
- 47:32Speaker 1: that big. And by the way, most of it is
- 47:34Speaker 1: spent on people, so and and and then of the
- 47:39Speaker 1: remainder after people, most of that is spent on like
- 47:42Speaker 1: facilities and upkeeping, ships and stuff like that. Like the
- 47:45Speaker 1: actual it budget, it's big, but it's not it's not
- 47:49Speaker 1: that big. And so in some sense, these companies have
- 47:53Speaker 1: already done what you're saying and have like created this
- 47:57Speaker 1: relationship with the government. And I suspect again I'm just
- 48:01Speaker 1: going this is just my like personal analytical read like
- 48:05Speaker 1: I have no knowledge, but I suspect it's because they
- 48:08Speaker 1: realized that they need to play nice with national security
- 48:13Speaker 1: and be helpful if they don't, if they don't want
- 48:16Speaker 1: to face the threat of regulation.
- 48:19Speaker 2: What does helpful mean? Exactly, like just trampling over the
- 48:21Speaker 2: Fourth Amendment exactly, Like what does that mean, like, what
- 48:25Speaker 2: is helpful me because like you know, the whole beef
- 48:27Speaker 2: against the d OD and Tropic was supposedly and I
- 48:31Speaker 2: could see that could be also seen as like a
- 48:33Speaker 2: marketing ploy too, that they had issues with Fourth Amendment
- 48:37Speaker 2: stuff like a surveillance on American citizens and stuff like that.
- 48:41Speaker 2: That was a big part of it. And then open air,
- 48:42Speaker 2: I just jumped in there, happy to do it, you
- 48:45Speaker 2: know what I mean? Uh? And I get that because
- 48:47Speaker 2: you know, even Sam Altman said over the last couple
- 48:50Speaker 2: of years, like at some point it's maybe that'd be
- 48:53Speaker 2: some financial uh some governmental like money coming in and
- 48:58Speaker 2: stuff like that, right, because you know at some point
- 49:00Speaker 2: private DEC's gonna run out and uh so, yeah, what
- 49:05Speaker 2: its back to my point is like the Fourth Amendment
- 49:08Speaker 2: does that count for anything?
- 49:09Speaker 3: Or no?
- 49:11Speaker 1: Well, look here here's how I here's how I look
- 49:14Speaker 1: at it. And it again, I'm reading the I'm reading
- 49:19Speaker 1: the uh the same news right as everybody else. And
- 49:26Speaker 1: the sticking point seemed to be about whether the Department
- 49:29Speaker 1: of War could use this for all legal means, which
- 49:35Speaker 1: at one level is like fair, yes, like it's the
- 49:38Speaker 1: American people who determine through policymakers what's legal, Like, we
- 49:42Speaker 1: pass laws, we have judges who you know, adjudicate those laws,
- 49:47Speaker 1: and then we say this is legal, this is a
- 49:48Speaker 1: legal thing to do. And so what counts as surveillance. Well,
- 49:53Speaker 1: I don't know. Some things like we have, we've had
- 49:56Speaker 1: wires happening since we've had phones. It's I think if
- 49:58Speaker 1: you ask to average America can like should the FBI
- 50:01Speaker 1: be able to wiretap you know, of Mafia Mafia Don's phone,
- 50:05Speaker 1: they'd probably be like, yeah, that seems like I've seen
- 50:07Speaker 1: that with a warrant. With a warrant, right, so we's
- 50:11Speaker 1: established some rules around it, right, So but we accept
- 50:13Speaker 1: that there's some level of surveillance that's acceptable and legal,
- 50:17Speaker 1: and then there's some level of surveillance that's not where
- 50:20Speaker 1: we're just like, no, you can't just like listen to
- 50:23Speaker 1: everything we're doing without a warrant and without any reason
- 50:27Speaker 1: to believe they would do this. This is like completely
- 50:30Speaker 1: breaking the constitution. And you know, I think that Edward
- 50:34Speaker 1: Snowden like revolutions like really got people very nervous about
- 50:38Speaker 1: what would be possible. So there's so if you just
- 50:42Speaker 1: kind of break apart this question, like, yes, it seems
- 50:47Speaker 1: reasonable that the Department of War or intelligence community or
- 50:53Speaker 1: law enforcement community in the US government should be able
- 50:55Speaker 1: to do anything legal that it wants, like that seems
- 50:59Speaker 1: reasonable with what it's buying. Then this question of well,
- 51:02Speaker 1: what should it be able to do legally is now
- 51:05Speaker 1: a policy question and a politics question and an oversight question, right,
- 51:11Speaker 1: because you know, you could argue that, like it gets
- 51:16Speaker 1: really subtle with Edward Snowden revelations, but like some of
- 51:20Speaker 1: some of what he was telling the public in these
- 51:25Speaker 1: documents was legal, Like it was perfectly legal, like the
- 51:29Speaker 1: government said, this is how we do it. We have
- 51:31Speaker 1: and then in some cases he's saying, you know, and
- 51:35Speaker 1: I'm not a legal expert on this, but it appears
- 51:37Speaker 1: that he's saying, well, there are these cases where it
- 51:40Speaker 1: wasn't legal that people cross the chalk line, that they
- 51:44Speaker 1: were doing things they shouldn't do. Now that's oversight. So
- 51:48Speaker 1: even for this AI situation, the Department of War is
- 51:51Speaker 1: not saying we're going to do illegal things. They're saying
- 51:55Speaker 1: we're just going to do legal things, and they should
- 51:57Speaker 1: be held to oversight so that they do those things.
- 52:00Speaker 1: And so you'll get the AI company. And look, I
- 52:04Speaker 1: do think that at a moral level. Somebody works on AI.
- 52:08Speaker 1: You don't want your AI to be used to do
- 52:10Speaker 1: illegal things or even like bend the rules, right, Like
- 52:15Speaker 1: you don't even want to get close to the chalk line.
- 52:18Speaker 1: And I get that, and I think that that's a
- 52:20Speaker 1: good ethical place to sit. So, like it's a complicated debate, right,
- 52:26Speaker 1: like do you do the legal things? I don't think
- 52:29Speaker 1: that you know, I you know, to to to to
- 52:35Speaker 1: to to like close it out. I think that you
- 52:40Speaker 1: have to decide between you want the American people to
- 52:43Speaker 1: decide what's legal. You want the government to have the
- 52:46Speaker 1: oversight to determine and ensure that it's done legally, And
- 52:50Speaker 1: you don't necessarily want a situation where companies get to
- 52:54Speaker 1: determine what is legal or not usage like that's not
- 52:58Speaker 1: really up to the company. But also at the same time,
- 53:03Speaker 1: some of these things don't have a legal precedent, and
- 53:06Speaker 1: so you do want companies to be careful, and like
- 53:09Speaker 1: I kind of applaud that cautiousness.
- 53:12Speaker 3: The surveillance issue is very interesting because I think a
- 53:14Speaker 3: lot of Americans don't know what is legal, and these
- 53:18Speaker 3: laws are created by people they elected, Like the Patriot Act,
- 53:21Speaker 3: for example, allows warrantless, Why are tapping in some situations?
- 53:25Speaker 3: That wasn't just made up. That was a drafted law
- 53:28Speaker 3: by elected officials that people in America elected, and those
- 53:31Speaker 3: officials enacted the policy that those people gave them mandate
- 53:34Speaker 3: to an act. Which is very interesting because then people
- 53:35Speaker 3: complain afterwards like, oh, this was unconstitutional. Well, you elected
- 53:39Speaker 3: these people and they passed this law, and you keep
- 53:41Speaker 3: voting for the same people, And there's a disconnect there
- 53:44Speaker 3: between the source of the vote, which is the person,
- 53:47Speaker 3: and then the outcome of that vote, which is a
- 53:49Speaker 3: policy thing or a law that says that things can happen.
- 53:52Speaker 3: And a great example of that is the biometrics that
- 53:54Speaker 3: everybody has on their smartphone and their laptop. I'm not
- 53:57Speaker 3: sure if Americans are aware generally that there was a
- 54:00Speaker 3: Supreme Court case that recently said that using someone's biometrics
- 54:04Speaker 3: to access a device is not a search, does not
- 54:07Speaker 3: require a warrant, it's not a Fourth Amendment problem.
- 54:10Speaker 2: Right.
- 54:11Speaker 3: Does the average American know that? I don't know, But
- 54:13Speaker 3: that's important because Americans need to inform themselves and educate
- 54:17Speaker 3: themselves about what are these things that they have allowed
- 54:20Speaker 3: to become laws? And if they don't like it, then
- 54:23Speaker 3: to do something about it instead of just complaining that he, well,
- 54:26Speaker 3: that's my right. Well is it because you kind of
- 54:27Speaker 3: surrendered that by electing that official who made that law,
- 54:30Speaker 3: that you continue to keep an office.
- 54:33Speaker 1: Yeah. And by the way, we typically make these laws
- 54:38Speaker 1: under a period of what feels like you know, distress, right, right,
- 54:43Speaker 1: That happen after nine eleven, and there was general, you know,
- 54:47Speaker 1: cultural consensus that we were afraid and we needed a
- 54:51Speaker 1: law like that, and it passed. But now it's twenty
- 54:54Speaker 1: something years later, we still have that law in place.
- 54:57Speaker 1: People don't really feel that way, and the technology has
- 55:00Speaker 1: radically change. But then also this is where elected officials
- 55:04Speaker 1: come in, because you can re elect people and tell
- 55:08Speaker 1: them change this law, right, and we don't want that anymore.
- 55:11Speaker 1: Or you can you can elect a president and say,
- 55:14Speaker 1: don't you know, enforce that law, like, don't use it
- 55:17Speaker 1: this way. Even though you can, law is still on
- 55:20Speaker 1: the books, we don't want you to use it that way.
- 55:22Speaker 1: So like people also forget that you kind of have
- 55:26Speaker 1: a say on this stuff. And I think what happens
- 55:29Speaker 1: is sometimes you know, when it comes to election time,
- 55:32Speaker 1: a lot of people, myself included frankly elect based on
- 55:36Speaker 1: economics or social issues or something like this, and forget
- 55:40Speaker 1: about all the national security stuff because none of these
- 55:44Speaker 1: no politician wants to talk about national security stuff in
- 55:47Speaker 1: the lead up to election day. They don't want to
- 55:49Speaker 1: remind people that, like the Patriot Act exists, and so
- 55:54Speaker 1: so it doesn't you know, it doesn't change.
- 55:57Speaker 3: So this is a good question. Then if you are
- 55:58Speaker 3: king for a day right now, what would you do
- 56:01Speaker 3: about AI policy in the United States.
- 56:03Speaker 1: Look, I'd rather live in a United States where like
- 56:06Speaker 1: we have control over the technology and information that affects Americans.
- 56:11Speaker 1: And when you have like that used to be fine
- 56:15Speaker 1: for basically a century because America made the best technology
- 56:21Speaker 1: in most of the technology. Now we're entering a world
- 56:24Speaker 1: where China makes a lot of technology and in some
- 56:28Speaker 1: cases may make better technology, but they don't want to
- 56:32Speaker 1: abide by our use or our our constitution our rules
- 56:37Speaker 1: around this. So I would just say, fine, like you
- 56:42Speaker 1: could do whatever you want in your country, but we're
- 56:44Speaker 1: not going to use your technology then, and so I
- 56:47Speaker 1: would immediately make any information, data or electronic technology from
- 56:53Speaker 1: China basically illegal or like highly highly restraint and just
- 56:59Speaker 1: say you're are going to make TikTok and you're going
- 57:02Speaker 1: to use it to collect information on American citizens, including minors,
- 57:06Speaker 1: and you're going to use it to mount effectively information attacks.
- 57:11Speaker 1: You don't get to have that anymore, like and in fact,
- 57:14Speaker 1: you don't get to have any more apps in America.
- 57:17Speaker 1: We're done, right, And I would do the same with AI.
- 57:19Speaker 1: You're going to censor your your you know, deep seek
- 57:24Speaker 1: and kimmy for your own political purposes and enact government
- 57:29Speaker 1: control and use it to collect data on American citizens
- 57:32Speaker 1: or to persuade American citizens. You don't get to sell
- 57:36Speaker 1: that here. You can sell it elsewhere. We'll battle you
- 57:39Speaker 1: on those fronts too, but you don't get to do
- 57:41Speaker 1: it here. So that's that's where I would come down.
- 57:45Speaker 2: Are these AI companies over valued?
- 57:49Speaker 1: So hard to say, right, Like, you know, what's the
- 57:51Speaker 1: trillion dollars anymore? You know, I guess it's like, look,
- 57:56Speaker 1: I run an AI company right nowviously not you know,
- 58:02Speaker 1: valued at a trillion dollars. And I look at how
- 58:05Speaker 1: powerful these models are, like it literally can feel like magic.
- 58:10Speaker 1: It does things that only people could do even three
- 58:14Speaker 1: years ago. And so I kind of look at the
- 58:18Speaker 1: whole US economy, the whole global economy. And I'm like, yeah,
- 58:23Speaker 1: we're going to use these models to do everything we
- 58:25Speaker 1: used to do, to draft legal contracts, to do accounting,
- 58:30Speaker 1: like to do human resources, to pick whatever subject that
- 58:36Speaker 1: you want. They're going to do it. So what is
- 58:40Speaker 1: the value of that? I mean, it's like you're at
- 58:42Speaker 1: this point you're sort of saying it's a fraction of
- 58:44Speaker 1: the total economy. You're like kind of like, okay, what's
- 58:48Speaker 1: the US economy fourteen trillions something like this, you know
- 58:53Speaker 1: per year GDP. I think like you're talking about fraction,
- 58:56Speaker 1: like it probably is worth that. And if you look
- 59:01Speaker 1: to the future, you know, you look at like Google,
- 59:05Speaker 1: you know, they were doing advertising and so forth. Now
- 59:09Speaker 1: they made Waimo. I don't know if you ever have
- 59:11Speaker 1: ridden one of these weaymos. I was in one for
- 59:14Speaker 1: the first time a few weeks ago in Los Angeles.
- 59:18Speaker 1: They don't have them in DC yet, and I you know,
- 59:21Speaker 1: I was a little nervous getting into this thing. And
- 59:23Speaker 1: I'm like, you know, I've tried the autopilot on you know,
- 59:26Speaker 1: Tesla's and stuff, and I'm like, it still feels shaky.
- 59:28Speaker 1: I got into this WAYMO, and it was the safest
- 59:32Speaker 1: car ride I've ever felt like I was in none
- 59:34Speaker 1: of the like human sloppiness of driving. There was no
- 59:39Speaker 1: weird person in the front of the car. Like I
- 59:41Speaker 1: got to be in my own room and I walked
- 59:43Speaker 1: out of this thing, and I was like, I would
- 59:46Speaker 1: would never take a taxicab again in my life if
- 59:49Speaker 1: I didn't have to, right, And so I'm looking at
- 59:52Speaker 1: like Google, and I'm like, man, whatever whatever valuation it is,
- 59:59Speaker 1: like at a true dollars because every car in America
- 1:00:02Speaker 1: is going to be driven by Waymo or one of
- 1:00:03Speaker 1: its competitors, right, So, like these things are going to
- 1:00:08Speaker 1: be worth so much it's hard every I think every
- 1:00:10Speaker 1: aircraft is going to be flown by an AI. Every
- 1:00:13Speaker 1: boat is going to be piloted by an AI. Every
- 1:00:16Speaker 1: lawn mower is going to be driven by an AI, right,
- 1:00:19Speaker 1: like the guy you know, you're just going to take
- 1:00:22Speaker 1: it out of the garage and send it on its way.
- 1:00:24Speaker 1: Already my sprinkler system is run by AI off of
- 1:00:27Speaker 1: an app, right, So you know, it's like the value
- 1:00:32Speaker 1: of these companies is probably worth it. Like I bought
- 1:00:35Speaker 1: SpaceX stock even you know, I lost money immediately, But
- 1:00:39Speaker 1: I'm like it's going to be worth more. I can
- 1:00:41Speaker 1: almost guarantee that.
- 1:00:43Speaker 2: How do we get to compute cost down?
- 1:00:47Speaker 1: Yeah, there's a lot of there's a lot of companies
- 1:00:49Speaker 1: working on lower energy essentially lower energy you know, compute
- 1:00:52Speaker 1: like that's like the real problem is that the the
- 1:00:55Speaker 1: literal amount of power that you have to put into them.
- 1:00:59Speaker 1: And then you know, you can also cut off like architecture,
- 1:01:03Speaker 1: you can change and so forth to kind of make
- 1:01:05Speaker 1: these things more efficient and optimal. I think those costs
- 1:01:09Speaker 1: are actually going to come down. The problem is that
- 1:01:13Speaker 1: we're going to want more smarter and smarter systems, right,
- 1:01:20Speaker 1: and so I think the average cost per call it
- 1:01:23Speaker 1: like IQ point will come down. But we're going to
- 1:01:27Speaker 1: want more and more IQ And remember, we're going to
- 1:01:31Speaker 1: get more and more data soon, right, It's not just
- 1:01:35Speaker 1: that all the Internet data or all the books or
- 1:01:37Speaker 1: all the things written down. People are going to be
- 1:01:40Speaker 1: wearing glasses and things like this that are going to
- 1:01:43Speaker 1: be recording the world. We're going to be recording all
- 1:01:46Speaker 1: of our conversations. There's going to be a lot more
- 1:01:47Speaker 1: data we're going to need to process that too.
- 1:01:49Speaker 2: Got didn't think genre everything that's it. We get it
- 1:01:53Speaker 2: all out. Yeah, Wow, that was easy in an hour.
- 1:01:56Speaker 2: Why can't they do this and watchington and so the
- 1:02:00Speaker 2: value you know, they'll.
- 1:02:02Speaker 3: Well, they'll feed our video into AI and they'll use that.
- 1:02:05Speaker 2: They'll figure it out. Yeah.
- 1:02:07Speaker 1: I mean we should have had you know, we should
- 1:02:08Speaker 1: have had you know, one of the AI models as
- 1:02:11Speaker 1: the other guest and could just ask that the same questions.
- 1:02:16Speaker 2: Funny.
- 1:02:16Speaker 1: Yeah, could have gone meta.
- 1:02:18Speaker 2: Give me more, give me more power, ye, give me
- 1:02:22Speaker 2: more information. I want to I want to get those books,
- 1:02:25Speaker 2: upload them and burn them.
- 1:02:27Speaker 3: Yes, unnecessary removing.
- 1:02:29Speaker 1: Yeah, I think I don't understand the book thing. I
- 1:02:32Speaker 1: assumed it was because they have to chop the spine
- 1:02:35Speaker 1: off the book to put it through the scanner, and
- 1:02:37Speaker 1: then what do you do with a book that has
- 1:02:39Speaker 1: no spine. It's like, well, yeah, we could throw it away,
- 1:02:42Speaker 1: or we could burn it. It's paper, we may as
- 1:02:43Speaker 1: well burn it. So sometimes I see these news stories
- 1:02:47Speaker 1: and I'm like, it seems pretty reasonable that, Like I
- 1:02:50Speaker 1: don't know if you've been to a used bookstore recently,
- 1:02:53Speaker 1: but like there's a lot of use books in the world.
- 1:02:55Speaker 1: Like I don't think these like I don't think our
- 1:02:58Speaker 1: problem is like you know that there aren't enough used books,
- 1:03:02Speaker 1: and like if somebody wants to take a used book
- 1:03:04Speaker 1: and chop the spine off and load it into a system, fine,
- 1:03:07Speaker 1: there's like a million other copies of Tolstoy out there, right.
- 1:03:12Speaker 2: I think there was some talk about this, some first
- 1:03:14Speaker 2: editions possibly, but I don't know. I mean, if they're
- 1:03:17Speaker 2: worth anything, someone's selling them, right, They're not going to
- 1:03:20Speaker 2: just be throwing them away.
- 1:03:21Speaker 1: You would think the economy would like figure that out,
- 1:03:24Speaker 1: like why would somebody burn a book that was worth
- 1:03:26Speaker 1: more than like five dollars? Like, I can't imagine they would.
- 1:03:31Speaker 1: And I'm sure, okay, have like some slip through the
- 1:03:34Speaker 1: cracks and yeah, but also you know what, some slip
- 1:03:38Speaker 1: through the cracks because like you know, somebody has a
- 1:03:42Speaker 1: first edition book and they just literally throw it in
- 1:03:44Speaker 1: the garbage when they're doing spring cleaning.
- 1:03:48Speaker 3: Well, not only that, we have digitization of like you know,
- 1:03:50Speaker 3: the Internet archive. There are Anna's Archive, which are you know,
- 1:03:54Speaker 3: not so legal versions.
- 1:03:55Speaker 2: Of the books.
- 1:03:56Speaker 3: There's there's a lot of electronic versions of some of
- 1:03:58Speaker 3: these books, especially rare books, because as you mentioned, the
- 1:04:00Speaker 3: market is intelligent and people immediately know that book is
- 1:04:04Speaker 3: valuable as soon as they see like, oh, this is
- 1:04:06Speaker 3: a first edition from the fourteenth century, handwritten by some monk.
- 1:04:09Speaker 3: We're going to digitize that, and then we can just
- 1:04:12Speaker 3: take the digital copy of that and use that to
- 1:04:13Speaker 3: train the model. We don't need the original physical book.
- 1:04:16Speaker 3: It's completely irrelevant. All that matters is do we have
- 1:04:18Speaker 3: any version of the book so we can put it
- 1:04:20Speaker 3: into this model. And I think that there's a lot
- 1:04:23Speaker 3: of I think unnecessary hooha about the physical book problem
- 1:04:26Speaker 3: in this Yeah.
- 1:04:28Speaker 1: And like also I think Google did scan all the
- 1:04:32Speaker 1: books like ten or twenty years ago, right, and then
- 1:04:35Speaker 1: they couldn't release it because of copyright issues, and so
- 1:04:39Speaker 1: the issue is like more about copyrights. And I I'm
- 1:04:43Speaker 1: again I don't know, but I assume now the AI
- 1:04:46Speaker 1: companies have enough money where they can pay off the
- 1:04:50Speaker 1: publishing companies to get rights to all these books, right,
- 1:04:55Speaker 1: And so you know, authors aren't seeing that, like I
- 1:04:59Speaker 1: didn't get, you know, some AI payment for my book.
- 1:05:04Speaker 1: But I assume the publishing companies too.
- 1:05:06Speaker 3: And there's even any non monetary incentive for the author
- 1:05:09Speaker 3: to be included in this training because it means that
- 1:05:12Speaker 3: their work will persist beyond and it'll be used for
- 1:05:16Speaker 3: something you know, beyond what they wrote it for. Whereas
- 1:05:19Speaker 3: if you opt out, it's like your book is dying
- 1:05:21Speaker 3: on the vine because you refuse to include it in
- 1:05:23Speaker 3: this global knowledge base that will affect all of humanity
- 1:05:26Speaker 3: from now until forever.
- 1:05:28Speaker 2: Yeah.
- 1:05:28Speaker 1: Also, by the way, what I've learned now publishing book
- 1:05:33Speaker 1: like almost no authors actually are being paid, like to
- 1:05:38Speaker 1: work as full time authors, like like that. There are
- 1:05:42Speaker 1: some people, you know, Stephen King, you know, makes a
- 1:05:45Speaker 1: great living doing it, and like people like that, and
- 1:05:48Speaker 1: Noah Harari on the nonfiction side or something. I'm sure
- 1:05:51Speaker 1: you make a rowling billionaire. I'm sure these people make
- 1:05:54Speaker 1: like reasonable money. But like most people publishing books these days,
- 1:06:00Speaker 1: I think you're making your money actually from well you're
- 1:06:04Speaker 1: not getting paid by podcasts, that's for sure, but like
- 1:06:06Speaker 1: you know, you're getting paid to speak, or you're getting
- 1:06:09Speaker 1: paid like to consult, or you're doing other things if
- 1:06:11Speaker 1: you're on the nonfiction side, or if you're a fiction author,
- 1:06:14Speaker 1: maybe you're teaching and doing things like that. And like
- 1:06:18Speaker 1: the the impression that like you know, it's all about
- 1:06:23Speaker 1: just getting paid is not really how the economy works anymore.
- 1:06:26Speaker 1: It's like maybe you have a sub stack and you
- 1:06:28Speaker 1: get paid through the SUBSEAC. There's so many other ways
- 1:06:30Speaker 1: to make money that's like totally new and I'm like
- 1:06:32Speaker 1: learning how to do it now. And so to your point, like, actually,
- 1:06:35Speaker 1: you kind of want your book on AI, and if anything,
- 1:06:40Speaker 1: I want AI systems recommending me more like, if somebody
- 1:06:44Speaker 1: asks a question about intelligence, I want them to reference
- 1:06:47Speaker 1: my book and the like, you know, get my name
- 1:06:50Speaker 1: out that way. It's almost like becomes like another path
- 1:06:53Speaker 1: to get your name out there.
- 1:06:55Speaker 2: And the book's called The Fourth Intelligence Oh, the Fourth
- 1:06:58Speaker 2: Intelligence Revolution, The Future of Esmina in the Battle of
- 1:07:00Speaker 2: Save America. That link is in the description. You can
- 1:07:03Speaker 2: grab it right now if dey tell us what Vico does.
- 1:07:09Speaker 1: Yeah, we you know, look, we built a system to
- 1:07:12Speaker 1: forecast events, right I used to be an intelligence officer
- 1:07:15Speaker 1: at some level. Intelligence is about figuring out what's going
- 1:07:19Speaker 1: to happen next in the world. So we built an
- 1:07:22Speaker 1: AI system at forecasts political, economic, and geopolitical futures better
- 1:07:28Speaker 1: than people, better than prediction markets. So it you know,
- 1:07:32Speaker 1: we forecasted the strikes on a run. We forecasted the
- 1:07:37Speaker 1: enclosure of the straight or hermus days and weeks before
- 1:07:41Speaker 1: prediction markets like cal shem polymarket and or you know
- 1:07:45Speaker 1: before kind of the consensus opinion by analysts. We did
- 1:07:49Speaker 1: the same on inflation and economic questions. We did the
- 1:07:52Speaker 1: same on political questions, like we forecasted really non consensus
- 1:07:58Speaker 1: things like that that Virginia would overturn the redistricting, for example,
- 1:08:03Speaker 1: or that grand Platner would not win the main race,
- 1:08:07Speaker 1: like we were way ahead on that, And so we
- 1:08:09Speaker 1: forecast events. And then on top of that, we have
- 1:08:12Speaker 1: built a whole system of tools from helping them make
- 1:08:16Speaker 1: decisions to unpack you know, how do I figure out
- 1:08:21Speaker 1: which courses of action to create the future that I want?
- 1:08:26Speaker 1: Right say, I'm a military commander and I want to
- 1:08:28Speaker 1: know which which you know, which which military moves I
- 1:08:34Speaker 1: want to make. Or I'm a lobbyist working with a
- 1:08:38Speaker 1: company and I want to understand what way that I
- 1:08:41Speaker 1: can help get a law enacted. Or I'm a trader
- 1:08:46Speaker 1: and I want to kind of figure out how to
- 1:08:48Speaker 1: balance my portfolio in a better way based on all
- 1:08:51Speaker 1: these future risks. So we built a whole series of
- 1:08:54Speaker 1: tools around that and by the way, for regular people
- 1:08:58Speaker 1: to use as well. Like, I think we're sort of
- 1:09:00Speaker 1: entering a world that's so complicated right now in economics
- 1:09:04Speaker 1: and geopolitics, Like for anybody, even people with like PhDs,
- 1:09:09Speaker 1: it can be hard to navigate. And so just anybody
- 1:09:13Speaker 1: can use the system as well, and you can try
- 1:09:15Speaker 1: it out. If you go to Vigo dot Io, Viico
- 1:09:18Speaker 1: dot io, you can try it out. But we primarily
- 1:09:20Speaker 1: work with, you know, people for whom decisions matter. So
- 1:09:23Speaker 1: people in finance were launched on the Bloomberg terminal. We
- 1:09:26Speaker 1: work with hedge funds and banks and insurance companies, and
- 1:09:30Speaker 1: then people in national security, like I still want to
- 1:09:32Speaker 1: do the mission. We work with the Department of War
- 1:09:34Speaker 1: in places like that.
- 1:09:37Speaker 2: Cool. Yeah, check that out. That link will be in
- 1:09:39Speaker 2: the description as well. Anthony, let's do this again. This
- 1:09:42Speaker 2: is awesome.
- 1:09:43Speaker 1: Yeah, this is great. Thanks guys, Thank you.
- 1:09:47Speaker 3: Thanks for answering my softball questions.
- 1:09:51Speaker 2: Yeah, I would have came at you a little bit
- 1:09:52Speaker 2: harder because I'm very bearish on AI. But not I
- 1:09:56Speaker 2: think it's going to be like Cisco. I think it's
- 1:09:57Speaker 2: going to be overvaluated. It's going to go down twenty
- 1:10:00Speaker 2: five years to get back to the evaluation, even though
- 1:10:02Speaker 2: it will do things that will change the world.
- 1:10:05Speaker 1: Yeah, I like, yeah, Well, first of all, like every
- 1:10:08Speaker 1: question you guys asked was like an op ed piece
- 1:10:11Speaker 1: that could have been hit something. Yeah, I'm like, oh man,
- 1:10:14Speaker 1: this is a lot. Yeah, but look, I don't know,
- 1:10:19Speaker 1: like the do you really think at the end of
- 1:10:24Speaker 1: the day, we're not going to use A We're going
- 1:10:27Speaker 1: to use AI less in five years than we use
- 1:10:30Speaker 1: it today.
- 1:10:32Speaker 2: No, I think we'll use it more.
- 1:10:34Speaker 1: Yeah, and then like I think we're going to use
- 1:10:38Speaker 1: it like one hundred times more because as you get
- 1:10:41Speaker 1: closer to this stuff, like I we run our whole
- 1:10:45Speaker 1: company on Claude, Like every meeting has transcribed, all the
- 1:10:50Speaker 1: code is in the repository. It reads all my emails,
- 1:10:53Speaker 1: and now I'm just like, hey, write this memo for
- 1:10:56Speaker 1: me or put together this deck, right, like all it
- 1:11:00Speaker 1: doesn't mean we're not using people, but like sure, yeah,
- 1:11:03Speaker 1: I think the whole world's going to be like that,
- 1:11:05Speaker 1: Like like the corner store is going to be run
- 1:11:08Speaker 1: on AI. Like you know, I don't know, We'll see,
- 1:11:13Speaker 1: we'll see. So you're not you're not you were not
- 1:11:16Speaker 1: a SpaceX byer.
- 1:11:17Speaker 2: No, no, no, I'm a SpaceX shorter. Nice Yeah, yeah,
- 1:11:22Speaker 2: I mean SpaceX particularly, I mean they're not even in
- 1:11:24Speaker 2: it was it was over valid coming to market. I
- 1:11:26Speaker 2: agree with you. Yeah.
- 1:11:28Speaker 3: Also, the prospectus was a great read.
- 1:11:31Speaker 2: It's crazy. It's the craziest read you'll ever see. I mean,
- 1:11:34Speaker 2: it was unbelievable to me reading that thing, uh first,
- 1:11:37Speaker 2: getting through the first fifteen pages of rockets. And don't
- 1:11:41Speaker 2: get me wrong, Starlink's an incredible company. Like I think
- 1:11:43Speaker 2: starlink is itself would be an unbelievable company that I
- 1:11:48Speaker 2: would invest in. Right, they actually have Yeah, they have
- 1:11:51Speaker 2: a you know, they have the monopoly essentially on it.
- 1:11:54Speaker 2: You know, at least they're at least five or ten
- 1:11:56Speaker 2: years ahead of people. So I love that one. But
- 1:12:00Speaker 2: the rest of it's a total ship box.
- 1:12:02Speaker 1: I was. I just took my son camping and hiking
- 1:12:05Speaker 1: and Olympic National Park and I had a star like
- 1:12:07Speaker 1: mini and like could still check email. It's like, it's
- 1:12:12Speaker 1: incredible we live in that world where you do that. Now,
- 1:12:15Speaker 1: It's you're right that part of the businesses is good.
- 1:12:18Speaker 1: But yeah, there was some high I mean, Elon is
- 1:12:20Speaker 1: like one of the best salesmen. You know. It's like
- 1:12:23Speaker 1: it's like P. T. Barnum or something like. This guy's
- 1:12:25Speaker 1: like incredible.
- 1:12:28Speaker 2: Interesting. Another fact. Fun fact, because you mentioned about like
- 1:12:32Speaker 2: the money being so you know, inequality being so crazy.
- 1:12:36Speaker 2: Back in the Gilded Age, Rose not Roosevelt, what's his name?
- 1:12:39Speaker 2: Rockefeller had his net worth was one point five percent
- 1:12:43Speaker 2: of our GDP at the time. Elon's now is like
- 1:12:46Speaker 2: over three percent of the GDP.
- 1:12:49Speaker 3: So it's like interesting, it's in different equity structures than Rockfeller.
- 1:12:53Speaker 2: Said, So right, John, I'm so sorry. Yeah, very important. Yeah.
- 1:12:58Speaker 1: Yeah, I bet Elon shows up as a global I
- 1:13:03Speaker 1: wonder what the global percentage is. I wouldn't be surprised
- 1:13:06Speaker 1: if it's like around one percent global, Yeah, which is.
- 1:13:09Speaker 2: Insane, it's crazy. So we'll see what happens. I mean, listen,
- 1:13:13Speaker 2: I use AI too. I use it for the podcast,
- 1:13:16Speaker 2: like to get time stamps, you know. I use it
- 1:13:18Speaker 2: as a helpful tool, you know, for sure. Yeah, and
- 1:13:22Speaker 2: I listen. I hope it cures cancer. I really do. Like,
- 1:13:24Speaker 2: I hope it changes humanity for the better. And you know, I.
- 1:13:28Speaker 1: Think you we'll see when these autonomous cars, when these
- 1:13:31Speaker 1: waymo cars and stuff, whoever's going to do, you'll see
- 1:13:34Speaker 1: an immediate like benefit in that a certain percentage of
- 1:13:39Speaker 1: people just won't die. Yeah, because they just drive better
- 1:13:42Speaker 1: than people. Like all drunk drivers could go away right
- 1:13:45Speaker 1: like every like you could live in a world where
- 1:13:47Speaker 1: like cars just don't crash, right that, Like if you're
- 1:13:51Speaker 1: drunk and you fall asleep, the car just slows you
- 1:13:54Speaker 1: down and puts you on the side of the road.
- 1:13:57Speaker 1: Like stuff like that is real, you know, and you know.
- 1:14:02Speaker 2: We'll see that's awesome, man, uh anything, let's do it
- 1:14:06Speaker 2: again for real.
- 1:14:07Speaker 1: This is great cool, Thanks guys, this is fun.
- 1:14:10Speaker 2: See you all right.