China: Weekly Summary (September 08-14, 2025)

It was one of those weeks where China felt very big and very close. Like you’re standing at a train crossing and a long freight train slides by, each car different, but still part of the same rumble. To me, it feels like the posts I read circled around two big things: fast hardware and quiet power. Rockets and regulations. Satellites and gangsters. Solar panels and social engineering. I’d say the mood was not breathless. More like eyes-wide-open. Curious, a bit uneasy, sometimes impressed, sometimes annoyed.

The audacity of AI plans, and the mess under the hood

I would describe Jeffrey Ding’s take on China’s “AI Plus” push as a friendly splash of cold water. The plan says AI should be everywhere by 2035. Not just in apps. In factories, hospitals, schools, ports, farms, the whole caboodle. Someone wrote down a number too: 90% penetration in the economy by 2030. That number sticks out like a neon sign. It’s bold, like calling a grand slam before the pitch.

Ding doesn’t just scoff. He pokes at it with history. Tech doesn’t always move like a bullet train. Adoption curves have bumps. Some have potholes. He asks: do we have the talent, the data, the compute, the incentives? He brings up open-source. And this part made me pause. Because we all talk about models and chips, but open-source is the oxygen for a lot of AI progress. In China, it’s alive, but it’s complicated. There’s energy. There’s also barriers. And then, tucked in there, he mentions new facial recognition rules. That’s interesting timing. The state is pushing AI everywhere, and at the same time, it is narrowing how faces get used. Not a full stop. More like “drive, but stay in your lane.”

To me, this pairing tells a little story. You can push for speed and still put cones on the road. If you’ve ever tried to reorganize your garage while the kids keep bringing in new stuff, you know the feel. You want order and you want more at the same time. That’s how the AI Plus plan reads here. It promises a lot, and it also shows China’s habit of making rules while building the machine. If you want the details, go to the post. The charts and adoption history bits are the chewy parts.

Engineers vs. lawyers, at least in one long coffee chat

Then I flipped to a conversation about China that had a different flavor. Homo Ludditus writes about a talk with Dan Wang, who has this on-the-ground vibe. He frames China as a country run by engineers. Not totally, of course, but it’s the mood. Lots of people who can pour concrete, lay fiber, stack lithium cells, and keep a factory humming for 16 hours. He contrasts that with the U.S., which he says feels lawyer-heavy. Rules, hearings, memos, and then, more memos.

I’d say the key here is not the cheap dunk. It’s that this lens explains why bridges open fast in one place and get delayed in another. It also explains why the AI Plus thing sounds plausible to some ears. If your system is set up to mobilize, then big targets don’t feel crazy. They feel like Tuesday. The chat drifts into social engineering too. What happens when the same energy you put into trains goes into people? It’s polite and uneasy at the same time. That’s the tone.

What I liked was the texture. The writer pulls in the sense of movement, the smell of steel and dust, and then flips to political change. Could it bend? Could it break? Nothing too certain. Just the kind of question you jot in the margin and circle twice.

The parade where the metal speaks

There’s also that big Beijing parade, the one marking 80 years since the end of the war. Naked Capitalism ran a piece that reads like someone taking inventory at a serious hardware store. DF-17 hypersonic glide vehicle. DF-41 intercontinental missile. Anti-ship ballistic missiles. New unmanned stuff. The note is blunt: some of this looks ahead of what the U.S. fields, and in a few lanes, China might be in the lead already.

To me, it feels like the post wants to knock us out of old habits. Not with fireworks, but with triangles and range rings. The parade, it says, sends two messages. Inside China, it says, “We’re strong. We’re organized.” Outside, it says, “You’ll have to think different.” Especially in the Indo-Pacific. Ships, carriers, bases—these get harder to use the way folks used to use them. Strategies that once fit like your favorite flannel now feel tight in the shoulders.

If you like nitty-gritty, the piece is jam-packed. The part that stuck with me is how the unmanned systems slot in. It’s not just missiles. It’s sensors, drones, swarms. War by mesh network. If that phrase sounds techie, yeah, but it’s also simple. Many eyes, many hands, moving fast. That’s the pattern these writers keep pointing to.

Satellites everywhere, including ones you don’t talk about

Space was loud this week too. Robert Zimmerman had several notes on launches. Eleven more birds went up for Geely’s constellation. Yep, the car company. That brings it to 52. I would describe them as eyes and ears for positioning, mapping, and who knows what later—probably a lot. It feels like the beginning of a private Chinese space internet. Not in the sci-fi sense. More like truck fleets getting smarter in the sky. Logistics with a halo.

Zimmerman also flagged a classified Chinese satellite launch. Those are the ones where everyone guesses and nobody admits. The tone is matter-of-fact, as usual. “Here’s the rocket, here’s the window, here’s what we know.” The man is consistent. He also noted a Chinese rocket test in another roundup. If you like space as a daily beat, his posts are like diner coffee refills—steady, sometimes a bit strong.

And then there’s the proposed mission to Apophis, the asteroid that will do its close Earth flyby in 2029. China wants in, with a mission aimed at asteroid defense. That’s a very different kind of space flex. Not just “we can lift heavy.” More like “we can do smart planetary chores.” If you think of space as the new shipping lane, doing traffic control is quiet influence. It doesn’t trend on TikTok, but it moves the needle in boardrooms and ministries.

I kept thinking about Geely’s satellites and that AI Plus plan. You can’t do glitzy AI at scale without fat pipes and precise timing. Constellations and compute go hand in hand. It’s not just more data. It’s better data. Cleaner timestamps. Smoother coverage. Like going from a wobbly tape measure to a laser level. The vibe is: build the rails first, then run the trains.

Influence maps, and the surprising soft spots

If the hardware stories smell like fuel and metal, Sam Cooper’s pieces smell like paper and cigarette smoke. Back rooms. Lobby lunches. Contracts with stapled corners. He covered new research from Taiwan’s Doublethink Lab that maps China’s political reach, country by country. The headline is uncomfortable for folks in North America. The report sees the grip tightening, with the U.S. neighborhood feeling more pressure than people like to admit. Canada is name-checked, not with vibes, but with numbers.

One post zooms out to a global view and says Asia is still the core, sure, but the Caribbean and parts of Africa and Latin America are getting encircled too. Encircled is a strong word. It’s also a practical one. It means, you look around and there’s a tie here, a port there, a media deal across the street. Not a single big rope. Lots of small strings that still tug.

Then Cooper zooms in on Canada and it gets sharper. He cites a more than 50% surge in measured influence over two years. Canada ranking 19th on vulnerability to interference in domestic politics, media, tech. There’s a phrase he uses for the Canada story: a triple movement. Increased exposure, sustained pressure, and rising alignment. It’s the alignment part that makes your eyebrow go up. It suggests the debate is not just about foreign activity entering. It’s about institutions shifting their posture to meet it.

Another line jumps out: Canada has seen a spike in election interference incidents but no prosecutions in those cases, while the U.S. has some. That’s a comparison that lands with a thud. It hints at the machinery of response, not just the threat. If you’ve ever dealt with a leak in the basement, you know. The water is pesky, but the pain is the plumber who can’t come for three days. Do you mop or do you fix the pipe? The post asks that, but with democracies and laws.

For more meat on the bone, Cooper’s global piece also points at tech, media, and academia ties. That mix feels familiar. You get a center that manufactures hardware, a software layer that travels easy, and a media layer that shifts what stories get airtime. It’s not mysterious. It’s the same stack people use to sell sneakers, but here it sells narratives and access.

Gangsters, politics, and the line that blurs on purpose

Cooper had another post that reads like a noir film set in Taipei alleys with neon signs buzzing in the rain. He dives into the Bamboo Union and the China Unification Promotion Party, and their links to the CCP. The claim is not just that criminals exist and do crime. It’s that these groups act like Kevlar vests for political influence. They wrap intimidation and money around messages, they attach to parties and local races, and they smooth edges that the official diplomats cannot touch with a ten-foot pole.

It’s a Taiwan case study, but the text nudges other democracies. You can almost hear the subtext: if it can happen there, where people watch for it, it can happen in places where people don’t. The language is careful but not coy. It says infiltration. It says political warfare. If those words feel heavy, that’s the point. Read it if you want the names and the timelines. It’s the kind of piece that makes you check the locks, figuratively.

Here’s where I digress for a second. In a lot of neighborhoods, you’ll find a guy who seems helpful—knows a cousin who can fix your car, can get a permit fast, can “talk to someone.” Everyone shrugs and says, that’s just how things get done. Then, years later, you realize the price. That’s how this post made me feel. It isn’t about spy gadgets. It’s about favors that turn into levers.

The long memory of APT-1

Kim Zetter took me back to 2013, to the report that put “APT-1” into everyday cybersecurity chatter. Mandiant naming names, straight up pointing at a PLA unit. It’s almost folklore now in security circles, but the post strips it down to the human beats. People took risks. They played it close to the vest. They argued inside the room. They were ahead of how government talked about these things at the time.

I’d say it’s useful to revisit because the threads still run through today’s stories. That report helped set the tone for how private firms call out state actors, not just in China but everywhere. It made attribution something you could read over breakfast, not just whisper in a skunkworks. And if you connect it to the Doublethink Lab research, it rounds the picture: networks don’t just influence politics; they hoover up business plans and patents. There’s the megaphone and there’s the magnet. The post hints at how both grew up together.

Also, there’s a quiet lesson in her story. Private companies sometimes move faster than state agencies. They also have their own motives. It’s fine to say both things. This week’s reading had that grown-up feel. Less hero talk, more “this is messy but real.”

When Silicon Valley ships surveillance and calls it innovation

Then I hit Davi Ottenheimer’s piece and it felt like a necessary scolding. The argument is simple enough: American tech keeps sending tools abroad that come back to bite moral and strategic interests. Not new. But the essay makes it modern by pointing at how companies helped build parts of China’s surveillance stack—some knowingly, some by pretending not to know. The tone is not performative outrage. It’s weary. He says the culture prizes cleverness over consequences. It’s a culture note, not just a policy one.

To me, this slots right next to Ding’s mention of facial recognition regulations in China. Think about the picture: the same environment that wraps surveillance in soft language also draws lines on paper for how to use it. That’s not hypocrisy; it’s control. It’s the state saying, we decide how this spreads and where it ends. Meanwhile, some of the source tech got its start under banners like “safety,” “community,” “analytics.” Then, boom, it’s pointed at populations.

I kept thinking of a kitchen tool. The mandoline slicer. In the right hands, it makes beautiful, even slices for your gratin. In the wrong hands, it takes off a fingertip. If you sell a million of them without a guard, you have to ask why. That’s the tone of the post. Less finger-wagging at “them,” more turned inward at “us.” If you care about this subject, it’s worth clicking through. He draws some clean lines between the old military-industrial patterns and the new tech-industrial ones.

Solar panels and the gravity of cheap

Switch lanes. Peter Sinclair writes about solar like a person who has stood on a roof with a drill. He points to an Ember report saying the solar revolution is no longer down the road; it’s pulling into the driveway. And the engine under that hood is Chinese manufacturing. Costs collapsed because China hammered down the supply chain, built massive capacity, and kept shipping. Wind too, but solar is the star.

There’s a political jab in there about Washington’s changing support for clean energy under Trump. You can nod or argue. The main thing is this: when prices plunge, behavior changes. It’s like when data plans stopped counting minutes for calls. Folks didn’t buy only on morals. They bought because it made sense in their wallet. Sinclair basically says, don’t overcomplicate the psychology. Cheap wins.

But cheap isn’t free of strings. Reading this after the influence and surveillance pieces, you taste the risk: dependency. If one country runs the factory of the clean energy future, then policy elsewhere has to speak softly and carry import credits. It’s not doom; it’s just leverage. That’s the real grown-up part of these posts. They don’t deny the bright side. More solar is good. Breathing easier is good. They just keep pointing at the bill stapled to the bottom of the receipt.

I’d describe his tone as practical. He wants people to stop treating the energy shift like a someday thing. For people who like to see numbers, that Ember report is your next stop. The graphs tell their own story, and it’s not subtle.

A syllabus from the past that still whispers

Irwin Collier offers something gentler and oddly grounding. He writes about Henry Rosovsky, the Harvard economist and dean, and an old course on the economic development of China and Japan from 1966–1967. It’s part memory, part archive, part tribute. You learn about Rosovsky’s journey, his leadership, his reforms in the Faculty of Arts and Sciences, and the empathy he carried because of his own refugee past.

If you care about China’s growth story, this is a nice palate cleanser. It reminds you that people have been thinking deeply about the China-Japan lanes for a long time. Not with today’s acronyms, but with rigor. You get the sense of how older frameworks shaped what professors taught and what students took to Washington, to Wall Street, to ministries. The post made me think about how we forget the long conversations that came before our heated threads. It’s not tactical like the other pieces. But it puts a bookshelf behind the headlines. Sometimes that’s what you need.

Zooming out: themes that won’t stop knocking

After marinating in all these posts, a few patterns kept tapping my shoulder. I don’t think any of the authors coordinated, but their notes rhyme.

• Speed meets discipline. From AI Plus to parade gear to satellites, there’s this syncopation: build fast, then regulate. Mobilize, then manage. It’s not laissez-faire. It’s not pure command either. It’s the particular Chinese blend—engineers steering, with legal paperwork catching up.
• Soft power that isn’t soft. Cooper’s reporting makes influence feel like a spreadsheet of relationships, not a movie plot. It’s not one big trick. It’s many small inches. Media buys, student groups, research ties, businesses that suddenly need permits. Little levers add up.
• Tech boomerangs. Zetter and Ottenheimer point to a loop: technology flows out, comes back as pressure. Whether it’s hacking units learning from global exposure or Western firms exporting components for surveillance, the circle closes. People love to say “unintended consequences,” but honestly, a lot of this was predictable if anyone looked past the quarter.
• The climate economy is real, and China is the core machine. Sinclair’s piece doesn’t yell. It states the obvious out loud: cheap solar changes the math. And much of the cheap comes from China’s scale and drive. You can celebrate the emissions trend and still gulp at the supply chain map.
• The map is the point. Space isn’t just bragging rights. Geely’s constellation hints at a world where positioning and sensing are in more private hands, inside China. That will shift how logistics and autonomy work. Also, a mission to Apophis says China wants to join the club that protects the house, not just parks cars in the driveway.

If you stitch these together, you get a picture that’s neither panic nor praise. It feels like standing in a workshop where the machines hum and the foreman keeps moving the chalk line. You can barely hear yourself think. Except you notice the chalk keeps drifting toward the door.

Small disagreements and quiet agreements

Were there fights between the writers? Not loud ones. But I’d say some tension lines show up.

• On capability vs. sustainability. Naked Capitalism sounds confident that China has jumped ahead in several military tech lanes. Jeffrey Ding questions whether grand AI targets are actually feasible by those dates. That’s a subtle disagreement: showpieces vs systems. Both can be true in parts. A missile working well does not mean street-level AI adoption hits 90% on time.
• On “the West” as a steady actor. Davi Ottenheimer is pretty harsh about Silicon Valley’s moral compass. Kim Zetter reminds us the private sector sometimes leads in good ways too, like Mandiant did with APT-1. Those posts tug in different directions—one scolds, one salutes. The tension is healthy. Not all firms are villains, not all are heroes. Some are both on different Tuesdays.
• On optimism about guardrails. Ding mentions facial recognition regulations in China. That looks like a guardrail. Cooper’s pieces about interference suggest that in democracies, the guardrails sometimes don’t bite. Canada’s lack of prosecutions versus the U.S. having some—the contrast muddies any simple story about who is more rule-bound right now in practice.

I also noticed quiet agreements.

• Everyone treats China’s scale as real. No one rolls their eyes at the idea that China can move big. The only question is where the friction hits.
• The information layer is front and center. Whether it’s hacking, influence ops, or the mesh of sensors in modern military kit, the line is the same: information wins fights before they start.
• Risk is moving to the edge. Rockets and satellites from China’s private sector, solar capacity in private factories, lobbying and media buys in democratic backyards—none of this sits safely in a tidy government box. It’s all edges and seams.

Tangents that loop back

I went down a couple small rabbit holes while reading. Nothing crazy, just the kind where you put a sticky note on the monitor.

• If China is embracing open-source for AI, what does that do to the leverage game? Open-source travels. You can’t fence it easily. Ding gestures at this. I kept thinking of a potluck where the best dish gets copied and improved in someone else’s kitchen. Does the state want that? Maybe yes, maybe no. Depends who’s copying.
• That Beijing parade’s dual messaging reminded me of sports intros. The team hypes the crowd and stares down the opponent at the same time. Except here, the fog machines shoot hypersonics.
• Geely’s satellites made me think of ride-hailing. When Uber and Didi started, folks thought “transport.” But the real story was the data off the ride graph. Out in space, a private constellation builds a graph of movement and timing. It’s a quiet advantage if they want it.
• The Bamboo Union piece made me remember old neighborhood advice: if someone offers a deal that avoids hassle, ask whose hassle you’re skipping. Usually, it’s the legal kind. Politics doesn’t change that basic math.
• The solar piece tugged at a very American habit: we preach competition until we meet someone who beats us with it. Then we invent a new word for why it was unfair. Hm. Maybe make more panels, or move fast on storage, or both. The Earth doesn’t care about our press releases.

What feels next, without being a fortune cookie

I’m not making predictions. But I’d say a few doors will probably open soon, based on this week’s notes:

• Watch the enforcement side. In Canada and beyond, people will start asking not just “what is the interference” but “what did we do about it.” That gap will define the next two years of politics in a bunch of places. Cooper’s posts put the pin on the map.
• Keep an eye on private Chinese space firms and constellations. The line between telecom, logistics, and defense is thin. Zimmerman's updates will probably keep catching little tells—a contract here, a test there.
• AI regulation inside China will likely get more specific. If AI Plus is to be more than a poster, someone has to write rules that factories and hospitals actually follow. Ding’s interest in open-source may grow sharper as those rules land.
• The climate build-out will increase the leverage of those who make the stuff. Sinclair’s argument is almost too simple to be ignored: price wins. That means the country that keeps lowering price while raising quality gets to whisper in bigger rooms. If you care about autonomy, that translates into industrial policy, like it or not.
• The tech-ethics fight in Silicon Valley will not be solved by a memo. Ottenheimer’s tone suggests this is cultural. Cultural problems change slowly, unless there’s a shock. We’ve had shocks. Expect more. Zetter’s history shows at least one path where private action forces public clarity.

The part where I nod at the writers and point you their way

If you want to feel the weight of numbers and timelines, go read Jeffrey Ding on AI Plus and facial recognition rules. He handles big promises without being snarky.

If you enjoy long, thoughtful walks through how China gets things done, Homo Ludditus channeling Dan Wang will give you smells and sounds along with policy bones. It’s not fanboy stuff. It’s a careful appreciation of engineering culture with the brakes intact.

If hardware makes your heart beat, Naked Capitalism lays out the military kit like a careful quartermaster. It’s not meant to scare. It’s meant to say: recalibrate.

For space junkies and realists, Robert Zimmerman has your Geely constellation updates, your classified launches, your asteroid detour. He writes like a man who has typed “liftoff” a thousand times and still enjoys it.

If your thing is the invisible strings that move politics, Sam Cooper brings receipts. The Doublethink Lab survey breakdowns, the Canada case, and the Taiwan gangster story are all doors into the same hallway. You don’t have to agree with everything to feel the draft.

For the backstory of modern cyber cat-and-mouse, Kim Zetter on APT-1 is like pulling out an old map that still gets you home. You learn how the field learned to say the quiet part out loud.

If you’re a tech person who wants to argue in good faith about responsibility, Davi Ottenheimer won’t let you hide behind buzzwords. It’s a mirror, not a cudgel.

And if you need a quiet room with books, Irwin Collier will sit you down with Henry Rosovsky and remind you that today’s debates echo older ones, in different shoes.

I didn’t expect solar to tie so neatly into geopolitics in a weekly skim, but Peter Sinclair made it impossible to unsee. The revolution isn’t vague. It’s measured in cents per kilowatt-hour and container ships of panels.

Some weeks you read about China and it’s all thunderclaps. This week felt more like a drumline. Precise, relentless, and marching around corners you didn’t notice before. If you like puzzles, this is a good moment to read slowly, save a few tabs, and let the patterns settle in. I’d say the best parts are in the links. The posts have the specifics, the interviews, the lists of models, the quotes from reports. What I tried to do is sketch the outline you can color in.

Anyway. That’s where my head landed after this batch. Feels like the train is still rolling, and there’s another curve up ahead. I’m going to go back and re-read the bits on open-source and on those island-influence maps. Those two keep tugging at the same thread, and I want to see where it leads.