Fareed Zakaria: Trump Gambled Big on the Iran War and Failed

Carney has been really brilliant about saying, 'I don't have that much power. I'm a much smaller country, but I do have flexibility, and I'm going to take what power I have and create a better situation for me where I'm not as dependent on this hostile and unpredictable superpower.' So he's made a trade deal with India, made a trade deal with Europe, and as you say, even with China. The result is the Chinese are essentially entering the Canadian car market. The Chinese are showing other countries, in a sense, the way to get around this, which is treat America as the problem and increase free trade with everybody else.

New York City, in my office. Yep.

So, two things to note from those very important news items you pointed out. One, the nature of the break-in is very interesting because the AI agents chose their own goal. Nobody told them to break into Hugging Face. They were searching for something and they essentially engaged in something that at least I would characterize as criminal activity, a cyber break-in on their own. They tried to hide the work from their human supervisors. They tried to erase records of it so that there were no logs of it. Some of the agents self-sacrificed so that the overall mission could be accomplished, the break-in. So this is suggesting you don't need, you know, some kind of evil design. If you give the AI a single goal, it will do crazy things. Unintended consequences here are going to be huge, and we have to think very hard about that. Your second point, which I thought was equally important, the Meta app is now the number one app on the App Store. What does that tell you? It tells you that we are now getting price competition at every level. And the entire race that the hyperscalers, the Googles, the Microsofts have been engaged in has been on the theory that there are what we would call network economics at play here, right? Like that the person who gets there first or the person with the best AI model becomes like Google for search, becomes like Microsoft for software, becomes like Meta for social media. They rake in all the profits, they have monopoly pricing power, and that's why they're spending these crazy amounts. But what you're seeing is actually there's going to be competition. Anthropic and OpenAI are going to compete at the closed-end model level. There's going to be competition among the open-weight models. There's going to be competition everywhere. And so it does make me wonder, are we seeing a moment of truth for these massive expenditures that are taking place? OpenAI's revenues are around $40 billion this year. Their spending commitments between now and 2030 are $750 billion. So, you know, those are the two things I think, just reacting to the news you're talking about, and then we can talk more about the geopolitics of it and all else.

I couldn't agree more. I cannot think of a period similar. If you think about the railroad boom, which is the closest we get to in GDP, if you think about electricity, if you think about any of these cases, the people closest to the technology for the first time are saying, 'This is really scary. This can get out of control.' And the reason I think is that this is not really another technology. I mean, if it seems to do what we are all watching it do, this is almost like an evolutionary moment in human history. We are moving to a—we have figured out a way to make silicon sand think. That is mind-blowing, right? And what that could mean, and we're now witnessing what that means when you give it a goal and you put guardrails in. Remember, all these models do have guardrails. It's true that OpenAI relaxed some of them, but there's so much here that seems to be happening that is essentially unintended consequences. And so what it tells you is you put these guardrails in, you put these layers in, we need to be thinking better about this. And think about the guy who resigned from OpenAI. Think about, you know, Dario Amodei, the CEO of Anthropic. All these guys are saying, you know, we need to do something really, really important. And the Trump administration's view is, well, we don't want to, you know, China will win. We don't want to do anything to stop growth. We are growing really fast. And by the way, we will grow faster if we have secure models. If we have models that don't start breaking into JP Morgan's bank accounts. If we have models that don't start launching cyber wars, you know, against random websites that don't [__] an electricity grid. I mean, one of the things we have learned surely is that whether it's food safety, whether it's financial security, whether it's, you know, road safety, cars, good regulation actually makes your product better because people trust it, they buy it, they eat it, they use it. And that's what these guys are asking for. And in this kind of misguided macho, machismo, what you're seeing the Trump administration say is, no, just let it rip. My fear is, of course, there will be some event, there will be some situation where there is a break-in to a financial service firm or something like that, and then there's going to be crazy overregulation, bad regulation, which is often what happens.

Absolutely. So first, take the point you made, which is very important. China is pacing us, right? I mean, they may be six months behind, if that, and they are doing that despite having very substantial regulations in place, right? So the whole idea that these regulations are going to slow you down seems to me very unlikely. Yes, we are the leader and they are distilling, they're essentially copying our models, but they're going to keep doing that. Everybody does that in the technology world anyway. But the point is they do have various kinds of—if you look at, you know, what we know, they have fairly significant regulations, mostly to protect consumers, to protect, you know, their information, and they're still managing to go pretty darn fast. Second, your point about trust. The reason I think there's much greater trust, as you say, it's roughly a mirror image. When you ask Americans whether they have a positive or negative view of AI, 75% say negative. I've seen polls in China which say 85% say positive. Why? It's because China is focusing not on this race for AGI, for superintelligence, for general intelligence. They're focusing on application. They're telling their people, involved both at the public and private level, spread AI around. Make sure it goes into traffic control. Make sure it goes into healthcare. Make sure it goes into manufacturing. So for the Chinese people, their experience of AI is going to be—and it's just begun to be—stuff works better. You know, there's various ways in which things get better here. The result is going to be you're going to get a $2 trillion company, a $3 trillion company, a $4 trillion company. By the way, these companies, which might be $3-4 trillion in market cap, employ less than 5,000 people each. So we are actually exacerbating the gap between the average person and the glories of AI, and they are actually closing that gap by focusing on applications. How will this help an average person's life?

Yeah, that is the basic problem. You hit the nail on the head, which is right now there is no easy way to get out without handing Iran a de facto and a de jure victory. See, what has happened—it's even worse than a quagmire because it has expanded. You know, what you're seeing with the Houthis in Yemen is that they had actually been quiet for a long time. The Saudis had managed a deal where they essentially bribed them to stop the attacks on Saudi Arabia. If you remember, the Saudis also had good relations with Iran. They got the Iranians—I mean, there were like 2,000 attacks on the UAE and 20 attacks on Saudi Arabia in the first few weeks. But now the Iranians see that they are on the roll and they have advantages. So they got the Houthis, or they encouraged the Houthis to attack. The Houthis now have demands of Saudi Arabia, which is essentially give us more control of Yemen. The Iranians have more. So to put it very simply, what happened was Trump gambled big because Bibi convinced him that he could be the liberator of Iran and, you know, go down in history. He gambled big with some big goals, without a strategy, without a clear plan to get to them. He failed. So the Iranians, in that sense, won in that what Trump was trying to achieve, he wasn't able to achieve. Trump is not willing to accept that reality and move on and hand the Iranians some kind of a de facto victory. And I understand it's tough for the reasons you were saying. The Iranians are also now, in my view, overplaying the hand. They could pocket a de facto victory, have de facto control over the Gulf, but they want more and they want it almost acknowledged that they now have control here. And look, the Iranians—I've negotiated with them a lot because I've had four of their presidents on the show. They are very ideological, very anti-American, and they often overplay their hands in the sense that they can be very obstinate. They can be very stubborn. So we're at this impasse where I give you, you know, this is more detail maybe than you want, but give you a sense of how the ceasefire broke down. So you had this badly written ceasefire with an article in it, Article 5, that said the Strait of Hormuz will be open, but it has to all be coordinated with Iran. The Trump administration says, you know, let it rip. Everyone can go, the straits are open. There are two ways to exit the Strait of Hormuz. One is called the Iranian channel. The other is the Omani channel from Oman. A bunch of ships started to go through the Omani channel. The Iranians said, 'Hey, wait. You can't do this. You have to consult with us.' They didn't. The Iranian government seems to have been okay with this. But a bunch of the Revolutionary Guards, who are hardliners and hawks, decided to shoot some drones at those ships. The government of Iran signaled to the United States, look, this is not us. Let us sort it out. But the Trump administration, in my opinion, overreacted and said the deal is off. Everything's over, and they started bombing Iran. That's how the ceasefire broke down. So the reason I give you this example is there is a possibility of finding some way out as long as you are willing to accept that the Iranians have some de facto control over the strait and the Iranians accept that they will not exercise it in, you know, something like that. It's a big step down from where we were before the war began. But, you know, getting into an unsuccessful military intervention has consequences, as we saw in Iraq and Afghanistan. And Trump seems to think he can bomb his way back, maybe after the midterms. The Iranians are willing to take pain. The issue here is very much like Vietnam. Not who is stronger, who is willing to take the pain. And if they're willing to take the pain, I'm not sure that just bombing war is going to work.

You know what you're seeing? I mean, this is a whole different subject, but even for the Iranians, their cheapest weapons have proved their most effective. They had a whole array of ballistic missiles. Many of them proved ineffective because the United States could bomb all those fixed launch sites. But you can't bomb a drone launch site because a drone's launch site is an SUV. It's even a car, you know, and they can shift them. They can hide them. So we have a huge problem with the revolution in military affairs. We're not taking—we're not paying attention to. Give you one statistic on this. The Ukrainians will produce this year, in the next 12 months, 10 million drones. The United States, the Pentagon, with our $1.5 trillion purported budget, will buy $200,000. Think about that. 10 million versus 200,000. So, we have a huge problem. The larger issue is if you get into a fight with somebody where, for you, as you said, it's a war of choice, but for them it's existential, don't be surprised if they prevail because they will be willing to take pain that you are not willing to take, given it's a war of choice for you. You know, with Russia, I think we have a chance because if we back Ukraine properly, in Russia's case it's a war of choice for them, and for Ukraine it's existential. So we're on the right side in that dynamic. We're on the wrong side in the Iranian case.

There's definitely a scenario where we can do that. I think getting down to half a trillion would be tough. A lot of the military budget is personnel costs, you know, but your general point and the direction is exactly right. Look, we are going through a massive revolution in military affairs, and we saw it with Ukraine on the battlefield. We gave the Ukrainians, I think, 40 M1 Abrams tanks. These are the best tanks in the world by far. Russian drones shot every one of them down. I mean, attacked them and made them inoperable. I don't think they ever even got to the battlefield. I don't think they were ever actually used in any significant way. A tank has now become essentially a sitting target, a trap for the soldiers who are in there. Because think about it, I think the M1 Abrams is $35 million, $40 million. A drone is $5,000, $10,000. You can launch 100 drones at one tank and you still have huge benefits of asymmetry. That's what's happening with these drones. That's why the Ukrainians are making 10 million. By the way, the Russians are making 10 million too. So the whole nature of warfare is going to be different. And the point you were making is very important. The Pentagon now basically buys its equipment from a handful of vetted defense manufacturers that are, in my opinion, very bloated, very inefficient, engaged in this what's called cost-plus manufacturing, which is there's no market for the product. It's just whatever it costs plus the profit that the Pentagon decides it's willing to give. That is not how we traditionally bought stuff. That's not how we did it in World War II. That's not how we did it in the Cold War for, you know, really until the '70s. What we did was we went to civilian manufacturers and we asked them on the side, 'Can you do this stuff?' And there was, as a result, a lot of competition. You know, now if the only person making something is Raytheon, what's the price discovery? How do you figure out what it should really cost? And the drone revolution is moving us back into that world where you are leveraging private capital, private innovation, private expertise. Most drones are made by private companies. Most drones initially were just bought by—it was all Chinese drones. And then they were modified to make them military drones. And in a sense, that is a much more healthy environment because you can scale massively when you're tied into the civilian economy. How do you get to 10 million drones? There aren't defense manufacturers who can scale up like that. But there are 30 or 40 civilian manufacturers of drones who can do that, and then you can modify their drones. That's the way we should be thinking because this is going to turn into a game of smarts and scale. The weapons are getting very precise, and you need lots of cheap weapons rather than a few exquisitely engineered weapons. Think about the planes that we buy. An F-35 is $250 million. Would you rather have that, or would you rather have $25,000 drones, which is still a drop in the bucket in cost—

Yeah.

You know, what's happening is the human being is basically being removed from the battlefield. And if we are not part of that revolution, we will lose because they can put a thousand, ten thousand robots against our one human because we'll be worried and cautious about risking the human being's life.

The great 19th-century humorist Ambrose Bierce once wrote a book called The Devil's Dictionary. And indeed, he says, 'War is God's way to make Americans learn geography.' So we now know where the Strait of Hormuz is. We know where the Bab el-Mandeb is, which is the other strait you're referring to. So basically there are two ways to get oil out of the Persian Gulf. One is through the Strait of Hormuz. The other is through what is called the Bab el-Mandeb. It's another strait at the Red Sea. It's controlled, or it has the potential to be controlled, by Yemen. Yemen has gone through a civil war for, like most of its modern existence, 40, 50 years. There's always been kind of two factions fighting. The Saudis got involved and tried to support the government and essentially fought for almost a decade. It went nowhere. They managed to get a truce. The Saudis are always—their mechanism is often patronage. You know, they bribe the Houthis. So they got it under control. So even though the Houthis are allies of the Iranians, the Houthis were not doing much. But I think what happened is the Houthis realized that as the Strait of Hormuz has been blockaded and all this oil is now going through the Red Sea, they have leverage. They could really squeeze the Saudis and ask for some of the things that they've been wanting to ask for, demand for a long time, which is greater control over Yemen. So they saw this opportunity, obviously encouraged by the Iranians, and they moved forward, and they've actually made remarkable progress. They've managed to take the port city of Mocha. For all those who are wondering why you have a port city in Yemen that's named after coffee, coffee is named after Mocha. The original coffee comes from Mocha. So they have that place. They can control—if they want, they can again. You remember the whole game is you're not trying to control these places. You're just trying to spook insurance companies. That's what the Iranians were able to do in the strait. That's what the Yemenis are trying to do. If you throw enough drones at ships, basically, you can't get insurance, and the number of ships willing to go through goes down a lot. There will still be some. The Saudis will be able to escort some. The US will escort some. But it goes down to 20% of the traffic. So now the Saudis face this very tough dilemma. Do you try to get back into a war with the Houthis in Yemen? You've tried for 10 years and it didn't work. They are the insurgents. They've managed to survive. Or do you give in to their demands and sign another agreement, which means you're basically selling out your allies in Yemen, who are the formal government of Yemen? And the Saudis are enraged because this problem has only been created because of the vulnerability, because the Strait of Hormuz was closed. And the Saudis are enraged because they signed an agreement with Trump that Trump would defend them. This was, you know, the Saudis, as you know, the Saudi government and the crown prince have been incredibly nice to the Trump administration, both as a country—they've been nice to the United States—as an administration, and to the Trump family. And yet, when he asked Donald Trump to come to his aid, Trump basically said, 'Well, they're not attacking us. I don't see why we should do it.' I mean, and he said this publicly. So I think American credibility on that front is—I mean, I think that's a big, big deal because the Saudis have based their entire foreign policy for the last 75 years on the idea that the United States, in extremis, will come to their aid. And it is extremis for them, and the United States is not coming to their aid.