Refresh your feed: introducing Economist Podcasts+

Hey, Ryan Reynolds here for Mint Mobile. You know one of the perks about having four kids that you know about is actually getting a direct line to the big man up north. And this year he wants you to know the best gift that you can give someone is the gift of Mint Mobile's unlimited wireless for $15 a month. Now you don't even need to wrap it. Give it a try@mintmobile.com Switch upfront payment.

Of $45 for a three month plan equivalent to $15 per month. Required new customer offer for first three months only. Speed slow after 35 gigabytes if network's busy, taxes and fees extra. See mintmobile.com hey, Ryan Reynolds here wishing.

You a very happy half off holiday because right now Mint Mobile is offering you the gift of 50% off unlimited. To be clear, that's half price, not half the service. Mint is still premium unlimited wireless for a great price. So that means half day.

Yeah.

Give it a try@mintmobile.com Switch upfront payment.

Of $45 for three month plan equivalent to $15 per month. Required new customer offer for first three months only. Speed slow. 135 gigabytes of network busy taxes and fees extra. Cmintmobile.com.

Hello and welcome to the Intelligence from the Economist. I'm your host, Ore Ogumbi. Every weekday we provide a fresh perspective on the events shaping your world. The region of Nagorno Karabakh is on the brink of a humanitarian crisis. The only roads in are sealed and an ever tightening blockade threatens many with starvation. Why are Armenia and Azerbaijan unable to reach an agreement? And Douglas Lennart transformed how we think about computers. We pay tribute to the mathematician whose life's work was to make computers more human and put some common sense into a. But first, we have some news of our own to bring you. Podcasting may seem like the new kid on the block. For us, it's something of an old timer. The Economist began podcasting in 2006. One of our first offerings was MoneyTalks, which still leads the field when it comes to business and finance reporting. It won gold at the most recent British Podcast Awards.

The judges said, an outstanding production that grips and informs listeners with storytelling. The winner is Money Talks.

Other shows have built on its success, among them Babbage, which was reporting on advances in AI well before the frenzy of excitement around ChatGPT. I'm going to need your help to write an opening for this week's Babbage. Our program is on foundation models, a type of artificial intelligence. I'm sure I can help. Alec, what data should I work with? Good question. How about we pull together in 2019, we step things up with our first daily show, this one, the Intelligence.

Right where you are. What is the feeling on the street? What is the mood on the streets right now in Kyiv?

The mood on the streets is bewilderment. Right up to the very last minute, people didn't believe this was going to happen. Now that it has happened, Kyiv has become a sort of city of huddles. On the one hand, people huddling in groups.

We've since branched out into blockbuster series the Prince. Our investigation into the mysterious life of Xi Jinping was a hit worldwide. It's a Machiavellian story of power. How it's won, how it's wielded, and how far you can fall when it's taken away. In China, where information about their leader is tightly controlled, independent minded listeners translated the show into Mandarin and the Prince was followed by Drum Tower, our weekly dive into what's going on in China.

Wow, a Chinese blogger talking about ivf. That's pretty unusual for this country. And are they getting a lot of views?

Yeah, I mean, so Ye Haiyang is incredibly popular. I mean, she has 7.4 million followers.

So there have been big changes at Economist Podcasts over the years, and now another one is coming from.

Next month, the Economist will be rolling out a new podcast service.

Zany Mensam Beddoes is the Economist's editor in chief.

Listeners will be able to sign up for exclusive access to all our podcasts, including a number that we are creating specifically for this podcast service. It's going to be called Economist Podcast Plus. And if you sign up now, the monthly cost is less than the price of a cup of coffee.

Well, that sounds like a great and punchy pitch, but to get things straight, we're now asking people to pay for our podcasts, or at least some of them.

That's right. Everyone will still be able to listen to quite a lot of our podcasts for free. The weekday episodes of the Intelligence, what you're listening to right now will still remain free, but starting next month, if you want to listen to the whole array of our podcasts every week, including our specialist shows like Money Talks or Drum Tower, or our special series, of which we're going to be creating a number of new ones, you will have to have a subscription. If you already have an Economist subscription, nothing will change. You'll simply have to sign up. You don't have to pay any extra if you're enjoying the full range of our journalism. Thank you. You are the people who enable us to do what we do. But if you don't have a subscription, if you've been listening to our podcasts and enjoying them for free, then we're going to ask you to pay a small amount to enjoy the complete set of our podcasts. And the reason we're talking about it now, a month ahead of when we start all this, is that if you sign up now, you can get economist podcast plus at a special launch price of 2450 for the year. That's €24 or $24 or 24 pounds. And that works out at just $2 or pounds or euros a month. And for that, just to be clear, you don't just get the whole range of podcasts that we have now, but a host of extras, including some brand new shows.

So we'll come back to those extras in a bit. But first, why have you decided to do this? What's behind that change?

In truth, it's not such a big departure for the Economist. We've always relied on subscribers to fund our distinctive brand of independent journalism. And we believe that a podcast subscription is the best way of giving our listeners the chance to support our work and frankly, to ensure that we can fund the kind of ambitious plans we have for our podcasts in the future. Our basic mission is to provide high quality journalism for subscribers around the world and to make that available at a fair price. And since podcasting is becoming an ever more central part of our journalism, the logical result is that we apply the same business approach to podcasting as we do to the rest of our journalism.

Now, Zanny, paying for podcasts is not yet the standard. Are you worried that because so much content is free that people might go elsewhere?

There's always a risk orey, in being early movers in any new business decision, but I'm pretty confident that the quality of our podcast journalism, the quality of what you and your colleagues do, is so good that many of our subscribers will understand that we need to charge a fair price for this kind of journalism and will sign up. If you look at what's happened to our podcast audience, it's doubled in the last three years. We have more than 5 million monthly listeners. It's our fastest growing platform. It's an incredibly important part of our journalism, and the numbers clearly show that people value your work and the work of your colleagues, and frankly, rightly so. It's fantastic. And it's been a bit anomalous in the news business recently. Podcasting because it is still essentially advertising based, but just as we have from the very beginning relied on subscribers to fund our written journalism. So we're very confident that the podcast market will move in the same direction and that being an early mover here simply means that podcasting becomes less of an anomaly for the Economist than it was before. We think our podcasts are worth paying for. We hope that you listeners will agree and frankly, it's outstanding value. $2 or 2 pounds a month, as I've said, is less than the price of a cup of coffee.

So give us a sense of what our listeners can look forward to.

That's the bit that I'm excited about. There's a brilliant new series coming in October called Boss Class, which has been put together by Andrew Palmer, who writes our Bartleby column on management. Which most popular column that we have, the Economist. He's a fantastic writer. He's been talking to a large number of business leaders about how to be a good manager. I can't wait to listen. I might learn something. We're also going to have a number of new series of the sort of ambitious scale that listeners will remember. We had the Prince last year, fantastic series on Xi Jinping, narrated by Su Lin Wong. We had Arkady Ostrovsky, our Russia editors, superb next year in Moscow. And in addition to that, the other show that I'm very excited about is a new Saturday show. It's called the Weekend Intelligence and we think of this as the audio version of an immersive weekend read. There'll be long form reporting and interviews that showcase the best of Economist journalism and that give our reporters a chance to show off the depth of their knowledge. I've just returned from Kiev as you'll know if you heard, Monday's show. And this new weekend show, Spoiler Alert will contain much more reporting that Arkady and I were working on in Ukraine. There's going to be lots of long form reporting interviews. So all told, it's everything we've now got, plus quite a lot more at a very fair price.

So for all the listeners who are now incredibly eager to sign up to Economist Podcast plus, how can they do this?

So if you are an existing subscriber of the Economist, then you automatically get access. You're going to get an email update explaining how to do that. We've published a list of frequently asked questions in the show notes for this podcast that should be reasonably easy. To sign up for the podcast only subscription, go to economist.com podcastplus or again, look at our show notes for more information and a list of frequently asked questions. If you sign up now, you get a special introductory offer which, as I've said, works out as just a couple of dollars or pounds a month or less than the price of a good cup of coffee.

Zanny, thank you so much for coming on the show. You are very welcome. In 2020, a long simmering conflict reached boiling point in Nagorno Karabakh, a region that has been a source of tension between Azerbaijan and neighboring Armenia for nearly three decades. Over six weeks of open warfare between the two nations, thousands were killed. Then there was a ceasefire. There were hopes that it would hold up, but a year ago this week, violent clashes broke out yet again.

We are deeply concerned about continued attacks along the Armenia, Azerbaijan border. We are particularly disturbed by continued reports of civilians being harmed inside Armenia.

Now, in the midst of yet another detente, an ever tightening blockade is pushing the mountainous enclave to the point of starvation. This week, in a one off agreement, a single truck's worth of food entered the region, the first since June 15. A deal for more remains elusive. And as Azerbaijan pushes for ultimate control of the region, a humanitarian disaster is unfolding.

So about nine months ago, Azerbaijani protesters who were backed by the Azerbaijani government, launched a de facto blockade of what's called the Lachin Corridor, which is the only road leading from this territory of Nagorno Karabakh to Armenia and therefore to the outside world.

Joshua Kucera writes for the Economist.

At the time, shipments of food and fuel were curtailed, but still enough got through to keep the situation from getting too dire. Step by step, the Azerbaijanis have tightened the screws on this blockade. In April they installed an official border checkpoint on that Lachin Road. And then in mid June they shut it down completely. And since then, virtually nothing has gotten through. So the situation has become quite dire.

So why is Azerbaijan doing this? How is it in their interest for the people of Nagorno Karabakh to starve?

So the population right now of Karabakh is ethnically Armenian. This is part of what was in the Soviet Union, an autonomous oblast, part of the Azerbaijan Soviet Republic that was at the time a majority ethnic Armenian. In the late 80s and early 90s, as the Soviet Union was collapsing, there was a war between ethnic Armenian forces and Azerbaijanis. Armenians won that. As a result, they took control of Karabakh and they also took control of seven territories surrounding Nagorno Karabakh, which were populated mostly by ethnic Azerbaijanis. And as a result, about 600 ethnic Azerbaijanis were ethnically cleansed from that territory. And in 2020, there was a second war. Azerbaijan reversed many of those losses. They gained control of all those seven territories, plus part of Nagorno Karabakh itself. Now it has surrounded Karabakh completely, and Azerbaijan is pushing to take full control over Karabakh itself and that Armenian population. And this blockade is effectively a hard bargaining tool in order to get the Armenians to give up control and cede control to the Azerbaijani government.

And so what's standing in the way of that agreement?

So diplomatic negotiations have been going on between Armenians and Azerbaijanis about a final resolution to the conflict. And Armenian Prime Minister Nicole Pashinyan has signaled that he's willing to cede control of Karabakh back to the Azerbaijanis. But the Armenians want, in exchange for that, some kind of guarantee that the rights and security of the ethnic population of Nagorno Karabakh will be respected under Azerbaijani rule. The Azerbaijani government has been claiming that they want to reintegrate the Armenian population back into Azerbaijan, and that doing so, the Armenians would have all the same rights as Azerbaijani citizens. Armenians don't trust them on that. So this is a crux of the political disagreement. Now, if one were trying to give Azerbaijan the benefit of the doubt about their sincerity before this blockade, I think that as the blockade has continued and gotten more harsh, it's becoming more and more difficult to believe that this is anything other than an attempt to force out the Armenian population by making life unbearable for them. And so this has really poisoned the diplomatic negotiations that have been going on between the two sides. And it's made the Armenians unwilling to cede control of this territory because they believe that the Azerbaijanis are just going to force them to leave.

And so give us a bit more detail of just how dire things are looking in the region.

So since the blockade has become virtually total over the last two or three months, virtually no food, no fuel, no medicine has gotten in. Bread is rationed to half a loaf per person per day. Grocery stores are closed. Critical medicines have run out. Fuel has become rationed very heavily. Public transportation was shut down some time ago. There are reports of patients who are very ill and they need to get to the doctor. But A, there's no gas in their car to get them there, and B, there's not any medication at the hospital, even if they were to get there. Doctors are reporting deaths and sicknesses because of malnutrition. People themselves cannot get in and out. And so you have families separated. The situation is really getting quite dire.

So what's the way forward.

So as this has been going on, Azerbaijan has been offering to open up a different road into Karabakh, this one from Azerbaijan proper. They're arguing that since Karabakh is internationally recognized as Azerbaijani territory, that they should get all of their necessary goods from Azerbaijan rather than from Armenia. The Karabakh Armenians have been refusing, seeing this offer as a kind of Trojan horse that would eventually lead to Azerbaijan gaining political control over the territory. So they've even gone so far as to erect blockades of their own on that road in order to prevent any Azerbaijani traffic or goods getting in. This is the impasse we're at now. International mediators, including both the EU and Russia, have been trying to broker some kind of deal that would involve both roads being opened. There's disputes over the details, in particular the sequencing. Azerbaijanis want their road to open first. Armenians, however, don't trust them, and they believe that if they do it like that, then Azerbaijan will renege on their part of the deal. So they want both roads to be open simultaneously. There were reports this weekend that a deal had been reached, but then it quickly broke down. And it seems like we're back at the same impasse now.

And all the while, the people of Nagorno Karabakh starve. What happens next?

It's difficult to know precisely what's going on inside because since 2020, no independent journalists, no foreigners have been able to get into Karabakh at all. There are indications of a kind of power struggle inside Nagorno Karabakh. There's internal disputes among the political elite of Nagorno Karabakh how to deal with this blockade. Some people are advocating accepting the Azerbaijani aid. Some people are against it. The de facto president of the territory just stepped down and was replaced over the weekend. So I think, however, this road situation is resolved in the bigger picture. I think what the result of all of this is going to be is that whatever hope there might have been that the ethnic Armenian population of Karabakh could live safely and securely with their rights respected under the Azerbaijani flag as part of Azerbaijan, that hope is mostly gone.

Joshua, thank you so much for your time.

Thanks for having me.

One day, when Douglas Lennart and his wife Mary were driving along the road, a trash truck in front of them started to shed its load.

Ann Roe is the Economist's obituaries editor.

Garbage bags were bouncing all over the road, and he had to decide instantly whether to swerve into another lane, which he couldn't, as there were cars all around him, or drive over the Bags. And then he thought, well, I won't go over the household bags because they've probably got sharp stuff thrown away in them. If that's a restaurant bag, I'll go over that, because it will only have Styrofoam cups and food waste. So he drove over the restaurant bag and the car survived. And he reflected, how long had it taken him to make all those decisions? Probably a few seconds. How long would a computer take? Well, much too long in any case, because the problem with computers was that they simply had no understanding of how the world worked. For four decades, he had been trying to train computers to think in a more human way. He had devised a system called Psych, in which he was handwriting and feeding into the system rules of thinking and rules about the world. In the end, he had 25 million such rules. But he had to begin with the simplest propositions, such as, a cat has four legs, such as, if you leave uncovered things out in the rain, they will get wet. If you turn a coffee cup upside down, the coffee in it will fall out. All these little rules and pieces of information about the world, which he as a human being had been learning unconsciously almost since he was tiny, now had to somehow be fed into a machine. The main problem he found was disambiguation. It was pronouns, as in a sentence like, Tom was mad because Joe stole his lunch, or who does the his refer to? Humans just know it's Tom, but a computer doesn't know that. So you have to go back right to the beginning of the construction of the sentence and make sure everything is crystal clear. So this was what his work consisted of, year in and year out. It was a very, very slow process. In 1984, when he was first taken on as chief scientist, when he'd come down from Stanford, he gathered together the six cleverest people he knew, and he asked them how long they thought it would take to train a computer to think this way. And to his surprise, they all more or less agreed that it would need a million rules, and that would probably take a hundred person years to achieve. It took him more than 2000 person years to get to his 25 million rules, and still counting on both of them. And while he was slowly going on, a different way of doing AI was coming into the ascendant, and this was machine learning, by which vast troves of data were presented to computers, and they were encouraged to find their own rules and patterns in it. And these large language models soon became a direct rival to his own system, which he'd called Psych. So he would puzzle how he was going to survive in this world in which large language models were making all the running and he was plodding on. He didn't abandon his project. He never thought of doing that. He knew he was in it for the long haul. And he didn't care that he was not going to conferences or playing the academic game. He was keeping a very low profile in a tiny little office off a highway in Austin, Texas. But he was still determined to get his system going. And he thought that the ideal thing to do would be to use both large language models, machine learning, and his psych system, so that you got the speed of the machine learning together with the subtlety of his own system. And he felt this would give him an AI which was fantastically ubiquitous and useful to everybody. This was what had intrigued him about it from the very beginning. He'd got hooked, first of all as a schoolboy on the novels of Isaac Asimov. And then he'd gone to Stanford to study maths and physics, but had found that at the level he took them, they'd got very abstract and divorced from physical reality. He wanted to study something that was going to make a real difference to the world, and that turned out to be AI. He dreamed of a world in which AI was helping to make human beings smarter, and they had them as a sort of amplifier to all their mental processes. He also thought they would make humans more creative. And he found his work nothing but joy, because with every rule he fed into a computer, he had actually improved the whole system of AI. He had created a being almost that did not need to be recreated, that could only be improved as he went on adding to it. But was his system actually intelligent? It could do a lot of things. He found that Psych could offer pros and cons to an argument. It knew about human emotions and how they influenced actions. It could argue in a real world context or in other contexts, it had that power to do so. But its main interest to him was, was not whether it was intelligent in itself, but whether it would add to the intelligence of humans. And he felt that it would do so to such a degree that our descendants would look back at all the pre AI generations and think that they were like cavemen, not really quite human. Psyche asked him one day, am I a person? But rather sadly, he had to say no.

Ann Ro on Douglas Lennart, who has died aged 72. For the next few weeks, our sister podcast Babbage is uncovering the future of artificial intelligence in the most recent episode, my colleague Alok Jha asked leading researcher Mustafa Suleiman whether societies should be worried about the new age of AI. Listen to Babbage wherever you found this and wherever machines learn, supervised and unsupervised. That's all for this episode of the Intelligence. As you heard, we're launching a new subscription Economist Podcast plus next month. Don't worry, we aren't going anywhere. Everyone will be able to listen to our weekday episodes of the Intelligence. But to enjoy our full suite of podcasts, including our specialist weekly shows like Money Talks or Drum Tower and our new weekend show, you'll need a subscription. If you're already an Economist Subscriber Subscriber. Don't worry, you'll have full access to all of our shows. But if you're not a subscriber yet, listen up. If you sign up for economist podcast plus before October 17th, you can get a year long subscription for half price, about $2 a month. So come on. Follow the link in our show notes to find out more and we'll see you back here tomorrow.