Hey guys, Non here, coffee in hand. Want to hear some interesting stories?
Rene Haas, the boss of the Cambridge-based chip designer Arm Holdings, told the BBC that artificial intelligence will find a cure for cancer in our lifetime. This matters because it highlights the massive shift in how we think about our most complex biological mysteries. We're also going to look at the first road opening between Russia and North Korea, a three-billion-euro fundraising move by the French AI firm Mistral, the trade war heating up between Canada and the United States. And the life of a woman who spent her years defending the poet Ezra Pound. The reason Haas is making such a bold claim is that he sees a wall we have hit.
Rene Haas, the chief executive of Arm Holdings, recently told the BBC that artificial intelligence will find a cure for cancer within our lifetime. He argues that while human beings currently find the complexity of DNA markers and cancer cell modeling too difficult to solve, computers will eventually bridge that gap as we feed more data into increasingly sophisticated models. Haas, who leads the company that designs the brains for hundreds of billions of phones and cars, also predicts that humanoid robots will become widespread in the next five years for tasks like manufacturing and cleaning. To understand why this is being said now, you have to look at the sheer scale of Arm’s current position.
Earlier this summer, Arm became the most valuable UK-based company in history in terms of cash value. Because they design the architecture for microchips, they are the gatekeepers of the hardware that runs the world's software. When Haas says AI will solve cancer, he is describing a transition from a world where we search for answers to a world where we build machines to calculate them. The popular narrative right now is one of a technological miracle—a "silver bullet" where a smarter computer suddenly makes a disease disappear.
It’s the version of the story that sells subscriptions and pumps stock prices. But there is a nuance here that the headlines often skip. The obvious objection to Haas's optimism is that a cure is only as good as the data it's fed. Someone might say that if we just throw more data at a problem, we might just get faster, more sophisticated mistakes.
They have a point. Professor Chris Bakal from the Institute of Cancer Research actually argues this exact tension. He says the future of medical AI won't belong to whoever builds the biggest computer, which is what the market currently rewards. Instead, he argues it belongs to whoever has the right measurements.
He points out that high-quality data from patient samples is far more valuable than the trillions of words scraped from the internet. Here is the part worth stealing. We are seeing a shift in power from the "builders" to the "collectors." While Arm provides the engines, the real value is moving toward the specific, high-quality biological data that allows those engines to actually function. Arm’s technology is already being used in half of the world's AI data centers. And Meta recently asked them to develop a specific AGI chip, which saw over two billion dollars in demand since March.
But for a cure to actually land, it requires a move away from the "bigger is better" philosophy toward a "better data" philosophy. We aren't just waiting for a smarter machine; we are waiting for a more precise way to describe our own biology.
We have spent the last few minutes talking about how we might use artificial intelligence to map the invisible architecture of our own biology to cure cancer. Now we are moving from the microscopic world of human cells to a very large, very physical piece of infrastructure: a one-kilometer bridge over a river. On Monday, Russia and North Korea officially opened the first road bridge across their shared border. According to a statement from the Russian government, this bridge spans the Tumen River and connects the North Korean city of Rason to the Russian city of Khasan.
It has two lanes and is designed to handle up to 300 vehicles a day. The inauguration happened via video link, with Russian Prime Minister Mikhail Mishustin and North Korean Premier Pak Thae Song attending from their respective countries. To understand why a bridge matters, you have to look at the constraints of geography and the incentives of war. Russia and North Korea already have a railway bridge and air service between them, but those are specialized systems.
A road bridge changes the "friction" of moving goods and people. It turns a logistical hurdle into a commute. This project was agreed upon in 2024 and construction began last year. Here is the part worth stealing: this bridge is not just about moving cars.
It is about the physical hardening of an alliance forged in the crucible of the war in Ukraine. The more important question is what this alliance actually buys for both sides. North Korea has been supplying ammunition and troops to Russia to support its invasion of Ukraine. In exchange, North Korea is receiving economic and military assistance.
By building permanent road infrastructure, Russia is moving from a "transactional" relationship to a "structural" one, where they are building a permanent corridor for trade, tourism, and human exchange.Someone will say that a bridge is just a bridge, and that the opening of a road crossing is a minor diplomatic gesture compared to the intensity of the fighting in Ukraine. You might be right. A bridge doesn't win a battle.
But it does build a dependency. When you build a road that allows 300 vehicles a day to move between two nations, you are creating a circulatory system. You are making it easier for North Korean tourists to enter Russia, and easier for Russian goods to flow into the North. It is a way of knitting two economies together so tightly that they become harder to pull apart later.
The obvious objection is that North Korea is already isolated, and a bridge to Russia doesn't change the fact that it is still cut off from the rest of the global economy. That is a fair point. But look at the numbers. Instead of being isolated, North Korea is now actively choosing a specific neighbor to be its primary artery.
While the rest of the world is watching the "intensifying and stuck" war in Ukraine Russia and North Korea are quietly building the plumbing for a different future.They are creating a closed loop of trade and military support that doesn't require permission from any Western bank or international shipping lane. What survives both of these perspectives is the reality of physical permanence. You can sign a treaty, and you can trade a shipment of missiles for a shipment of grain, but those are fleeting.
A one-kilometer bridge over the Tumen River is a concrete fact. It is an investment in the long game. It suggests that for Russia and North Korea, the current conflict in Ukraine isn't just a temporary emergency to be solved; it is the starting point for a new geographic reality. Russia is betting that by lowering the cost of moving goods across that river, they can make their alliance with the North Korean regime durable enough to outlast the current war.
They are building a permanent path for a temporary alliance. The bridge is now open, and the cars are starting to move.
While a bridge between Russia and North Korea physically connects two nations, the next story is about how a French company is trying to build a digital bridge for the rest of the world. We are moving from the movement of cars across a border to the movement of data across a network. Mistral, the French artificial intelligence company, has just raised three billion euros in a massive funding round. The investment was led by Samsung Electronics, along with Scaleup Europe Fund and PSG Equity.
According to the company, this capital is meant to scale their compute capacity—the raw processing power needed to train models—and expand their international footprint. They already support over one hundred and twenty-five global enterprises, including heavyweights like Airbus and HSBC. To understand why this matters, you have to look at the current split in the AI industry. For the last two years, the goal has been to build the biggest model possible, a race dominated by American labs like OpenAI.
But for a large corporation or a government, "biggest" is often less important than "private." If you are a bank or a defense contractor, sending your sensitive data to a third-party cloud to be processed by a black-box model feels like a massive security risk. Mistral is betting on a different strategy: open-weight models. This means they provide the "weights"—the internal settings that dictate how the AI thinks—so that companies can run the model on their own hardware. This creates what they call a sovereign AI layer.
It means your data stays inside your walls, your compute costs are predictable, and you aren't locked into a single vendor's pricing or roadmap. The obvious pushback is that Mistral isn't "benchmaxxing"—it isn't consistently beating the top American models in every technical test. Some critics argue that if you aren't the strongest, you're irrelevant. But there is a real case that performance is a spectrum.
A company doesn't always need a model that can write poetry in fifty languages. They need a model that can accurately sort through their own internal invoices without leaking those invoices to the internet. By focusing on control and independence, Mistral is trying to make AI a utility you own, rather than a service you rent.
While Mistral is trying to build a sovereign infrastructure for AI, Canada is currently grappling with the messy reality of sovereign trade. We are moving from the theoretical independence of software to the physical friction of moving goods across a border. Canada struck back at the United States on Tuesday with new tariffs on about twenty billion dollars worth of American products. These counter-tariffs, which Prime Minister Mark Carney described as dollar-for-dollar, apply to everything from steel and furniture to cotton t-shirts, with some rates hitting fifty percent.
To put it simply, Canada is trying to match the scale of the pressure it is feeling from its largest trading partner. Here is how the math works. The United States currently has a twenty-five percent tax on Canadian cars and trucks, plus duties on steel, aluminum, and lumber. In late August, President Donald Trump added new fifty percent tariffs on items like dairy, alcohol, and even hockey sticks.
Because the two countries share a bilateral trading relationship worth nearly nine hundred billion dollars in 2025, any tax on a t-shirt or a piece of furniture isn't just a line item in a ledger. It is a direct increase in the cost of living for people who have to buy those things, and a hit to the margins of the companies that ship them. Someone will say that these tariffs are a necessary tool for national sovereignty, a way to force a seat at the table. That is a fair point.
But the reality on the ground is more complicated. We saw this when Canada pulled fresh fish and lobster off its retaliation list after pushback from the seafood industry. The lobster industry in both countries is so intertwined that American-caught lobster is often sent north to be processed before being shipped back south. If Canada had kept them on the list, they would have been punishing their own processors as much as their neighbors.
The more important question is how long this can last. The Canadian Chamber of Commerce has warned that businesses are preparing for a long dispute, but they want a surgical approach rather than endless escalation. If the trade war turns from a disagreement over dairy into a total halt on manufacturing, the cost becomes very real. For instance, Donald Trump threatened to halt all business with Bombardier, a Canadian airplane maker that contributes over seven billion dollars to the annual GDP.
When you move from taxing goods to banning industries, the price of the conflict stops being about trade and starts being about the basic functioning of the economy.
The BBC reports that President Trump is threatening to block the sale of Canadian Bombardier jets in the United States. Over at NPR, we see that Canada has retaliated by slapping tariffs on about twenty billion dollars worth of American goods. Meanwhile, The AI Daily Brief is highlighting a new software called OpenClaw 2.0, which allows human workers to share a digital workspace with AI agents to hand off projects in real-time. To put it simply, these stories are showing us the same friction point in two different rooms.
One is a trade war between two neighbors who are physically touching. The other is a workflow war between humans and software that is increasingly becoming part of our infrastructure. The connection is the struggle over how we distribute and control labor. Canada and the United States are arguing over who gets to profit from the physical movement of goods and heavy machinery.
They are fighting over the "how" of the industrial world. At the same time, the rise of collaborative AI agents like OpenClaw 2.0 shows that the "how" of the digital world is being redesigned to remove the friction of human hand-offs. In both cases, the goal is the same: to find a way to produce more while reducing the cost of coordination. When Canada puts a tariff on twenty billion dollars of goods, they are trying to force a change in the price of cooperation.
When a company uses an AI agent to share context without reconstructing a project from scratch, they are trying to solve the same problem of coordination, just with bits instead of steel. The more important question is whether we are moving toward a world where the "human" part of the work becomes the most expensive and least efficient link in the chain. Whether it is a tariff on a jet or a software patch for an AI agent, we are watching the same pressure build. Watch for the moment when the cost of coordinating people becomes higher than the cost of replacing them with a system that doesn't need to be "managed."
Before the last story: if you are getting something out of this, subscribe. The World in Twenty Minutes is on Spotify, on Apple Podcasts, on iHeartRadio, and anywhere else you already listen. It is free, it lands every morning, and subscribing is the whole reason it keeps finding people.
We just looked at how trade wars can break the machinery of a modern economy, but the next story is about a different kind of construction. The work of building a bridge between a complicated past and a living language. The Economist reports that Mary de Rachewiltz, the poet and translator of the epic poem The Cantos, died on July 24th at the age of 101. To understand why her life matters, you have to look at the scale of the work she undertook. She was the daughter of the American poet Ezra Pound, and she spent her life translating his life’s work into Italian.
This wasn't just a hobby; it was a massive, dual-language project of the 1980s that earned her the prestigious translation award in 1986. She also curated the Ezra Pound Archive and co-edited his letters to his parents, published in 2011. For much of her life, Rachewiltz lived in the shadow of a father who was a polarizing figure, a man who was arrested by U.S. authorities and held for twelve years in a disciplinary training center.
She lived at Brunnenburg castle in the Italian Tyrol, where she taught his literature to students from American universities and wrote her own Italian-language poetry. The core of her life was a refusal to let a complex legacy be simplified into a single narrative. She chose to do the hard, technical work of translation—making sure the specific nuances of her father's words survived the move from one language to another. She spent decades ensuring that the "Cantos" remained accessible, not as a relic of a controversial man. But as a piece of literature that could be read by anyone, in their own tongue.
Thanks for listening today.