Professor Jiang began with an object rather than an argument: a semiconductor, "the basis of all technology, of all computer". Inside it, he said, three elements do most of the work — silicon, gallium and germanium — and together they make possible things that sound impossible, from precision missiles to satellite communication.
His point was about where those elements come from. By his figures, China produces roughly 60–70% of the world's silicon, about 70% of its gallium and about 90% of its germanium. Then he traced the journey of a single chip on The Diary of a CEO. Companies in California such as NVIDIA or Apple design an advanced chip to power a phone or a video game. The design goes to Taiwan and Malaysia to be manufactured. The machines that actually print the circuits are built in the Netherlands and Japan — Jiang named ASML, whose lithography machines he priced at $400 million each. The finished chips then travel to China to be fitted into appliances.
Lithography is the step that gives a chip its pattern. As ASML's own technical explanation describes it, a mask carrying a circuit design is projected through optics onto a light-sensitive silicon wafer, the wafer moves, the exposure repeats, and layer by layer the transistors are built up. Light sources, lenses and mirrors, positioning and measurement are each their own specialism.
That, for Jiang, is the whole lesson. "When you see this semiconductor, you're not just seeing a piece of metal and plastic," he said. "You're really seeing the world converging together."
The formula
Jiang asked the host to remember a formula: specialization times globalization. Technological progress, he argued, depends on more countries being able to specialize and on those countries being able to trade freely. Holding that trade together, in his account, are the US dollar and American military dominance — which he defined as the absence of any rival able to challenge the United States for control of the seas.
Run the formula backwards and you get his forecast. If war and rivalry push nations to build self-sufficient supply chains, countries become more general rather than more specialized, trade narrows, and progress has to slow. "We've basically hit the apex, the height of technological advancement," he said. He traces that coming chain of events to the war in Iran, the subject of much of the rest of the episode.
Dependence runs in both directions
Jiang spent some time on a detail he thinks gets lost in talk of China cornering critical materials. China exports silicon and rare earths, he said, but it has to import helium to keep its own semiconductor manufacturing running, because helium is used as a cooling agent in the process.
His mirror example was American. Spruce Pine, North Carolina, he said, holds something like 90% of the world's high-quality quartz, which the silicon manufacturing process also needs. The supplier Sibelco describes its high-purity quartz as coming from two ore bodies at Spruce Pine, used to make fused-quartz crucibles for growing single crystals of silicon, along with tubing and other quartzware used in wafer processing. The quartz is not part of the chip; it is part of the equipment that makes the material the chip is cut from. Solar manufacturers use the same crucibles. Those company documents set out particular manufacturing dependencies; the claim that technology has reached its ceiling is Jiang's own.
"Robots."
The formula came back later, applied to a different problem. Jiang had argued that Europe faces two pressures at once: migration from younger, more unstable regions to its south, and its own rapidly aging population, which needs cheap healthcare and nursing. Blocking immigration, he said, is exactly what newly ascendant right-wing parties promise — he named the AfD in Germany, MAGA in the United States and Vox in Spain — but it leaves the aging question unanswered.
The host supplied the obvious modern answer: "Robots."
"Exactly, right? That's what you would think," Jiang replied — and then took it apart with the same supply chain. Robots need copper, from Chile, Peru, China and Australia. They need lithium, from the lithium triangle in South America. They need silicon and rare earths, mostly from China. All of it, he noted, comes from places a long way from Europe and the United States. In theory you could use robots; when global supply chains break down, he said, you have to find the cheapest way to solve a labor shortage, and the cheapest solution is immigration. "People are much cheaper and much more expandable than robots," he said.
He was aware of how this sounded. "I know this is a bit scary. I know this is really depressing," he said, before offering the reverse case: if the United States won the war in Iran and achieved regime change, oil would flow again to South and East Asia, those economies would manufacture and recycle their profits, and with that money "AI is able to innovate to a point where we achieve AGI" — artificial general intelligence, a machine broadly capable across the range of tasks people do — and the problems would be solved.
What AI becomes instead
Near the end of the episode the host asked what else was on his mind, then raised AI himself, noting that it runs underneath everything they had discussed: robotics, energy, semiconductors, chips, the United States and China.
Jiang returned to the formula. In a time when global supply chains are breaking down, in a time of war, he said, AI cannot make the leap into AGI. Instead: "AI will be repurposed as a control grid rather than AI developing the intelligence necessary to reduce scarcity, to solve problems" — the most pressing medical problems, climate change. It would be turned to monitoring and surveilling everyone.
The host thought that was coming anyway, for reasons that had nothing to do with Iran. Governments are increasingly interested in watching people all the time, he said, and the cameras are already installed — they were recording in New York City, where "there's cameras everywhere". The limitation is that the cameras are "just a little bit dumb" and "they don't really talk to each other". With AI, they can. Jiang agreed, and the two sketched the result: the cameras noting that the professor is walking down the street, on his way to his hotel, and now inside it. That, the exchange concluded, was the most likely trend.