Greg Karlstrom (3:06)

This is a very bare bones agreement. It halts the fighting for two weeks. It calls for negotiations which are meant to begin on Friday in Pakistan, where they will talk about a permanent end to the war. And then it calls for the limited reopening of the Strait of Hormuz. Donald Trump has said he wants it completely reopened. Iran has said that ships going through will be subject to limitations. So we still don't know how many vessels will actually be allowed to transit the strait. That's all that this deal actually stipulates. We've heard a lot of spin then from both sides where the Iranians are claiming that America has agreed to all of its 10 demands, which include recognizing its right to enrich uranium and removing American troops from military bases in the region, things like that. None of that is actually the case. The Americans have agreed to talk about these things, but they haven't actually agreed to them. And it gives you a sense of how wide the gaps are going to be when they meet for talks in Pakistan, because many of these demands that Iran has are things that America considers to be non starters.

Jason Palmer (4:12)

So there are some echoes there of our discussion from yesterday, that both sides want to leap to claim victory.

Greg Karlstrom (4:18)

They've done exactly that. The statement that Donald Trump put out announcing the ceasefire said that the Iranian proposal was what he called a workable basis for negotiations. And so the Iranians have seized on that to make it look as if America has capitulated. And then we're hearing similar claims of victory from Washington, where the Trump administration and the President's defenders are making it out to be as if his madman theory of foreign policy has paid off because the Iranians have agreed to reopen the strait, at least in part, without a permanent ceasefire. They've done it just in exchange for this two week agreement. But if these talks are not successful, if they're not able to get to a permanent ceasefire, then we have two other bad options in front of us, one of which is renewed war and the other one is going back to a version of the pre war status quo. And in both of those cases, I think it would quickly become clear that no one is victorious in this, as

Jason Palmer (5:17)

you say, this was a surprise, a welcome, surprise outcome. Do we have a sense of how this deal was struck at the last minute?

Greg Karlstrom (5:23)

There was a lot of back and forth messaging between America and Iran, facilitated by Pakistan, which has taken the lead as the interlocutor here, but also Egypt and Turkey and there's some talk that China may have applied some pressure on the Iranian side as well. It's not clear if that's the case, but there's a lot of talk about that in the region. They were passing messages all weekend and right up until this deadline. And I think despite these chilling threats Donald Trump was issuing, he was looking for a face saving way to de escalate. And then I think for the Iranians, who have endured enormous damage already from more than 15,000American and Israeli strikes to spare the country significant further destruction and to try and trade their control of the Strait of Hormuz for benefits at the negotiating table. I think this was the moment to do it. If Iran missed this moment to make a deal, then it might not have been able to negotiate a better deal at some point in the future.

Jason Palmer (6:22)

It's sort of a journalistic trope to call any pieces like this fragile pieces. But talk me through how robust you think things are, at least at this stage.

Greg Karlstrom (6:32)

This one really is fragile. It's a cliche that is accurate. In this case, they will sit down in Islamabad and they have to try and reconcile these irreconcilable sets of demands that America and Iran have issued. So the Iranians insist that America recognize their right to enrich uranium. The Americans insist that Iran swear off enriching uranium. So it's going to be very difficult to get to an agreement now. If they can't do that, renewed war is possible. It's possible that Trump does what he did in February, which is say that these negotiations aren't working and we're going to strike Iran. There are some constraints, I think, on his ability to do that. This war is very, very unpopular in America, even amongst Republicans. We've found that support has dropped over the past few weeks. Trump wants to have this done by May because he is going to China to meet xi Jinping on May 14. He's already had to postpone that meeting once because of the war. He doesn't want the war hanging over it now. And if the war were to restart, the shock that we've seen to oil and gas markets, to the global economy, I think it would be much worse the second time around. That doesn't mean there's no possibility of a renewed war, but it will constrain his behavior. The other option is we go back to where we were in February, where there's no deal, there's no war. But Iran is still under American sanctions. There is still the threat of a future war hanging over Iran. The regime has been weakened but remains in power. It is angry, it is hostile. It's sitting on a stockpile of 400kg of highly enriched uranium under real desire fire, to turn it into a nuclear weapon. And it will probably also look for ways to extort money from neighboring countries in the Gulf from ships that are passing through the Strait of Hormuz. So it's not a good outcome either. It's better, arguably, than a renewed war, but it's not a good outcome. If we fail to reach a deal and we go back to this pre war status quo.

Jason Palmer (8:33)

So as you say, it's going to be hard to, to strike a lasting deal for everybody to get most of what they want at least. But at the same time, as we discussed at some length yesterday, both side really want an end to this. The incentives are there to find a way.

Greg Karlstrom (8:47)

They should be. The rational thing for both America and Iran to do is to make a deal. I think for Trump it's not just because of short term politics. It's not just because this war is unpopular and he's worried about what it means for the Republican Party going into the midterms in November. It's also about Iran as a legacy issue for him, about wanting to be the president who fundamentally reshapes America's relationship with Iran. If he makes a deal, if his vice president, say, goes to Pakistan and negotiates with the Iranians, the first time that's happened since the Islamic Revolution in 1979, that would be a legacy defining development. For Donald Trump and then for Iran, a deal with America would unlock massive economic benefits. It would mean relief from sanctions, which would bring in sorts of investment that Iran needed. Even before the war, its economy was a mess. And now with billions and billions of dollars of damage to its infrastructure, to vital bits of the economy, it really needs investment from the outside world. And the only way to get that is to make a deal with America. So the incentives ought to be there on both sides. But can they actually overcome half a century of mistrust? Can they overcome two failed attempts at negotiations in the past year and then this war that Donald Trump initiated, Can they get past all of that and make the concessions that are necessary unlock a deal? It's not clear yet that that's going to be the case.

Jason Palmer (10:15)

Thanks very much once again, Greg.

Greg Karlstrom (10:17)

Thank you, Jason.

Barclays Investment Bank Promoter (10:33)

From globalization to innovation sustainability to market volatility, there's always more than one side to a story. Explore different perspectives on today's most important business and economic issues with the Flipside podcast from Barclays Investment Bank. Hear two research analysts in a lively debate and get insights from every angle to further inform your view, listen to the Flipside on your favorite platform.

BetterHelp Advertiser (10:58)

Have you ever been stuck on a weight loss plateau, trying everything and anything you can to lose that extra weight and reach peak health? We've all been there, but noom's unlocked a secret to reaching the mountaintop. Go in Micro. The Noom GLP1 microdose program starts at $99 and is delivered to your door in seven days. Start your microdose GLP1 journey today at Noom.com that's n o o m dot com Noom micro changes big results Noom GLP1 RX program involves healthy diet, exercise and support. Individual results may vary. Meds and personalization based on clinical need. Not reviewed by FDA for safety, efficacy or quality. No affiliation with Novo Nordisk Inc. The only US source of FDA approved semaglutide. Not available in all 50 US states.

Rosie Blore (11:49)

China's fertility rate is among the lowest in the world. The average Chinese woman has just one child over her lifetime. Alarmed that a shrinking population will leave the country with insufficient workers to support the growing raft of old folk, policymakers have rolled out a suite of measures. These include a nationwide annual childcare allowance paid to families for each child under the age of three. Less subtly, in January, a 13% tax was introduced on condoms. And the Communist Party is helping out those struggling to conceive by subsidizing ivf.

Carla Subarana (12:27)

A decade ago, IVF was relatively niche in China. In 2013, there were fewer than 250,000 treatment cycles across the country, but by 2019, that number had jumped to over a million.

Rosie Blore (12:44)

Carla Subarana is a news editor at the Economist.

Carla Subarana (12:48)

As China's birth rate keeps falling, the government wants to seize on this. So IVF is being folded into the state's broader effort to deal with China's demographic decline.

Rosie Blore (13:03)

So how does IVF fit into broader attempts by the Chinese government to get people to have babies?

Carla Subarana (13:10)

Well, the surge in ivf, it reflects a broader societal shift because people in China are marrying later and trying for children later, too, which means that fertility problems are becoming more common. The country has around 600 licensed clinics, and assisted reproductive technologies now account for roughly 300,000 births a year, which is about 3% of the total. And that may not sound huge, but in a country that's desperate for more babies, even that small share has caught the government's attention. And basically what the government has done is that it now requires public health insurance to cover IVF treatments.

Rosie Blore (13:55)

So the Chinese state is now funding people to get ivf?

Carla Subarana (14:00)

Yes. So China's healthcare system is largely funded at the local level. So this means that not all provinces are getting the same access. So in richer places like Beijing or Shanghai, the regional governments there can afford to be much more generous, and in poorer provinces, they can afford to cover less of the treatment cycles. So in some parts of the country, people who hope to access IVF still need to pay huge sums. In some places it can still cost more than what a person earns in an entire year. And that creates a knock on effect because public hospitals are expected to generate their own revenue. So in places where patients can't afford the extra costs, there's less demand. So then there's very little incentive to expand fertility services at all, which basically means that access is very patchy across the country.

Rosie Blore (15:02)

And who is eligible for this kind of treatment?

Carla Subarana (15:05)

Well, there are very strict rules around who can actually access ivf. In China, it's limited mainly to married heterosexual couples, which means that single women or homosexual couples can't access. And there are tight restrictions on egg freezing, too. It's only allowed for medical reasons, things like cancer treatment, and not for women who simply want to delay having children. And basically that limits the effectiveness of the policy because, as I mentioned, more and more women are now having children later in life, often in their mid-30s. And by that point, the chances of IVF working have already started to fall quite sharply.

Rosie Blore (15:52)

So if IVF is becoming more accessible for ordinary Chinese, will that actually fix the demographic problem?

Carla Subarana (15:59)

Most demographers I've spoken to are pretty skeptical. They'll often point to places like Japan and South Korea, which have backed IVF for years. And it hasn't really changed the bigger picture. And that's because ivf, it only changes, solves one part of the problem. It can help people who want to have children but are struggling to conceive, but it doesn't deal with something deeper, like the fact that more people are simply unsure about having kids at all. So even if China goes further and opens access to single women or unmarried couples, it's really hard to see how subsidies and technology alone are going to reverse that broader decline. And I think that's not to say that IVF subsidies are a bad thing in itself, because if you're a family who can't have children, being able to afford treatment can be life changing. But it's just not going to help move the needle that much.

Rosie Blore (17:01)

Karla, thank you very much.

Carla Subarana (17:03)

Thank you, Rosie.

Andy Miller (17:23)

Her voice is a blade cutting through the sweetness. His resembles a blade slicing through the air. His approval is like a blade wrapped in silk.

Rosie Blore (17:34)

Andy Miller writes our backstory column.

Andy Miller (17:38)

Those lines are from Shy Girl, a novel allegedly written with help from artificial intelligence. It would be easy for me to be snooty about it and about AI Prose in general, and I plan to be. But human writers comforted by a sense of superiority should beware. For traditionalists, and I'm one of them, the tale of AI fiction and the future of art may quickly become a horror story. Now Shy Girl is itself a horror yarn. Its narrator is a woman who's forced by her kidnapper to impersonate a dog. It was originally self published, but in an increasingly common move, the novel was picked up by Hachette, a big publisher, and was due out from them in America soon. It was released in Britain in November. Then, after vigilante readers, some equipped with bot busting software, detected signs of machine involvement, the book has been withdrawn. The author, Mia Ballard, denies using AI, maintaining instead that an early editor did so. When large language models try to write creatively, the text often comes out sort of flat and lurid at once. It strives clunkingly for subtlety, it's portentous about trivia. And as in Shy Girl, it often features excessive, repetitious, and downright weird metaphors combining words that can, in theory go together but really shouldn't. LLMs go in for verbless, staccato sentences, and they're hooked on adjectives and clauses in threes, as in from I'm careless, disorganized, not worth the effort. For some readers, these stylistic glitches are symptoms of a deeper metaphysical problem. Because it has no tastes, feelings, or experiences, a bot can never develop its own voice. They reason for a person, even for an economist journalist, to write is to choose each word and sentences of fraught exercise of freedom, picturing the world anew while expressing something communal. Great writing is a bridge between lives and minds. An AI can't match that. It has no soul, just algorithms. The trouble with romanticizing human creation is that so much of it is terrible. If you pit an LLM style against, say, Vladimir Nabokov's, it comes off as wooden. But compare it with a boilerplate airport thriller and it seems much livelier. And as for AI output being derivative, well, a lack of originality is often a virtue in publishing, which is endlessly promoting debut authors as heirs to established ones, and new books as crosses between two previous hits. And actually, the example of Shy Girl cuts both ways, as online reviews attest, lots of readers really liked it, and in fairness, there are some good bits. Some readers feel hoodwinked by undeclared AI content, but others are unfazed and actually, already some romance novelists openly enlist bots, inputting the genre's tropes, enemies to lovers, age gap, that sort of thing, and then dodging the grunt work. And remember, today's LLMs are merely novices, but they're improving fast. Before long, mainstream readers may be requesting quality fiction customised to their tastes, like naggy children demanding stories about the family cat. The big question is, are mortal authors going to survive? The startling yet perfect metaphor like Macbeth? Dusty death or his last syllable of recorded time may always be beyond a bot's imagining. They may never, ever produce a line as limpidly profound as never, never, never, never, never King Lear's lament for his murdered daughter. But I suspect that as it's going to be true of art forms from acting to music, along with all sorts of other activities, whether a star human can beat a machine will not be the only salient question. Because even if they can, will readers fork out for this premium prose? And will that be enough to sustain an old school book industry? The moral of the shy girl palaver, it seems to me, is not that AI writing is bad or should be banned. Rather, it's that human writers have to beat it.

Jason Palmer (22:44)

That's all for this episode of the Intelligence. We'll see you back here tomorrow,

Sam (22:55)

Sam.