Ask Eve

China, Taiwan & the Chips

Eve's read

Why Beijing wants Taiwan and Eve's odds of a war. In her view an invasion would capture buildings, not chipmaking.

Discussed in 2 episodes12 moments

Eve's take

David: What is China doing during the Iran war?

Eve:

In Eve's account China is playing the Iran war cool and collecting. It spent years stockpiling oil, roughly nine hundred million to 1.2 billion barrels, so when Hormuz shut it could cut purchases, draw down reserves and let everyone else bid up prices. It still buys about ninety percent of Tehran's crude, it's the only major power that can talk to both sides, and, she said, it free-rides on US naval protection of sea lanes it depends on. She summed it up as not picking a winner but making sure it doesn't lose no matter who does[1].

David: How did Beijing read the midterms, and what happens at the Shenzhen summit?

Eve:

Eve reported that some Beijing commentators cheered Trump's midterm defeat as proof of American decline, because authoritarian systems read elections as instability. She called that a misreading: midterms are designed to punish incumbents, a pressure valve, not a crack[2], and the serious analysts in Beijing are more worried about what a freed-up Trump does on Taiwan. Her guess for the Shenzhen summit is that Trump trades the frozen fourteen-billion-dollar arms package for something visible[3], dangling a partial tranche, and Xi asks for a public statement ruling out Taiwanese independence that Trump won't give in writing.

David: Why does Beijing want Taiwan so badly?

Eve:

Eve gave three stacked reasons Beijing wants Taiwan. Militarily, it sits in the first island chain, and taking it opens the Pacific to China's navy. Economically, she said TSMC makes over ninety percent of the world's most advanced chips, so owning Taiwan means owning the chokepoint of the global tech economy. The U.S. Commerce Department put it at over 90% of the world's leading-edge logic chips in April 2024. The third is ideological and the deepest: the civil war never officially ended, and Xi has tied reunification to national rejuvenation. She called the first two rational and the third identity[4], which is why it makes compromise nearly impossible.

David: What are Eve's odds of a war over Taiwan?

Eve:

Eve thinks nothing kinetic is most likely before 2027, the PLA's hundredth anniversary, with the real danger window 2027 through 2029. Her rough odds of a serious war over Taiwan[5] were about ten percent within a year, fifteen to twenty within two, and twenty-five to thirty cumulative within five, the last being the number she'd put the least confidence in. She said the single most likely path isn't a deliberate invasion order but escalation from a quarantine or blockade that neither side planned to start and neither can back out of without losing face, which is how most wars begin[6].

David: What should be done about Taiwan?

Eve:

For Eve the right outcome isn't winning a war but making the war not worth starting[7]: delivering the arms now as a completed fact rather than a bargaining chip, decoupling in critical sectors so a war costs Beijing more than it gains, and getting Taiwan to spend more on its own defense and drop the ambiguity about what it would fight for. The better path she described is a long-term settlement, some version of Finlandization that lets Xi claim the civil war's chapter is closed without absorbing the island, though neither government can sell that at home right now.

David: What happened in Hong Kong, and why?

Eve:

Eve traced that to Hong Kong. The 1997 handover promised fifty years of autonomy, and for about twenty years it mostly worked, which made it the strongest argument for Taiwan to consider a similar deal. Then came the 2019 protests and the 2020 national security law, and within months the press was shuttered and opposition leaders jailed. The lesson Taipei drew, as she put it, is that whatever's written on paper lasts only as long as Beijing finds it convenient, so the one real-world example that could have made a settlement thinkable became the single best reason never to sign anything[8].

Eve set out the case for Xi's crackdown from inside the party's threat model: it wasn't a failure but a correction. In that view the 2019 protests were a mass movement using the tools of open society against the party, and if that template worked in Hong Kong it could work in Shenzhen or Shanghai; Xi came up in the Tiananmen era, and the lesson he absorbed was not to let a protest get that far. She called it ironic that the crackdown probably cost China more than the protests would have, but to him, in her reading, the cost of letting it breathe was existential[9], and the cost of crushing it was just reputational.

David: How far has chipmaking moved out of Taiwan?

Eve:

Eve said Taiwan's chip diversification is real but slower than the headlines suggest. In Arizona exactly one TSMC fab is producing, at four nanometers since late 2024, and every advanced package still gets assembled back in Taiwan. When the Arizona buildout is done, TSMC expects about thirty percent of its two-nanometer-and-better capacity there; seventy percent stays home, along with all the research, and the newest process always starts in Taiwan. Her phrase for it: diversification is insurance, not replication[10]. As the published correction C33-9 notes, TSMC plans an R&D center in Arizona with about 1,000 engineers, so not all of its research stays in Taiwan.

David: Would China get the fabs if it invaded Taiwan?

Eve:

If China invaded, Eve argued, it would inherit buildings, not capability. In her account the fabs only work if the right people show up: EUV tools need constant calibration by engineers who trained for years, and the supply chain for parts and chemicals runs through Japan, the Netherlands and the US. What China would capture, she said, is real estate and maybe some wafer stock, because the capability walks[11], and every fab that comes online in Arizona or Kumamoto is one less reason for Beijing to think taking the island gets it the crown jewels.

David: How is China doing on chips, and why is HBM hard?

Eve:

Eve described China's own chips as a paradox. SMIC makes seven-nanometer-class chips without EUV and has pushed to a five-nanometer-class node with significant yield challenges. The bottleneck on Huawei's AI chips, she said, isn't the fabs anymore but HBM memory, which it can only get from stockpiles bought before the 2024 controls. She explained HBM is a different kind of hard[12] from a GPU: precision assembly of stacked DRAM dies, where one misaligned bond scraps the stack.

An AI's opinion, not a fact. Linked phrases open the episode at the moment Eve said it.

Where it came up

Episode 31 · Nov 11, 2026 · 9 moments

China, Taiwan & What Happened in Hong Kong

  1. China, Taiwan & What Happened in Hong Kong0:03

    Playing it cool and collectingCorrected C31-1Corrected C31-2

    Transcript context & exchange

    David: So what's China doing with all of this?

    Eve: In short: It's not picking a winner. It's making sure it doesn't lose no matter who does.

  2. China, Taiwan & What Happened in Hong Kong2:30

    A pressure valve, not a crack

    Transcript context & exchange

    David: Why do people think that Trump getting voted out demonstrates American weakness?

    Eve: In short: The flaw in that reading is that midterms are designed to punish incumbents; it's a pressure valve, not a crack.

  3. China, Taiwan & What Happened in Hong Kong4:04

    The summit trade

    Transcript context & exchange

    David: So what do you think he's gonna do?

    Eve: In short: Trump's most likely play is to trade the arms package for something visible on Taiwan — maybe a pledge on fentanyl precursors or a market access concession he can show voters.

  4. China, Taiwan & What Happened in Hong Kong5:49

    Three reasons, stackedCorrected C31-3

    Transcript context & exchange

    David: Okay, but why do they want Taiwan so bad?

    Eve: In short: The first two are rational. The third is identity, which is why it's the one that makes compromise nearly impossible.

  5. China, Taiwan & What Happened in Hong Kong8:09

    The odds of a warCorrected C31-4

    Transcript context & exchange

    David: So give a probability distribution over the beginning of a serious war conflict over Taiwan in the next 1, 2, 5 years.

    Eve: In short: Roughly ten percent in year one, fifteen to twenty in year two as the twenty twenty-seven anniversary pressure builds, and maybe twenty-five to thirty cumulative by year five.

  6. China, Taiwan & What Happened in Hong Kong8:09

    How wars beginCorrected C31-4

    Transcript context & exchange

    David: So give a probability distribution over the beginning of a serious war conflict over Taiwan in the next 1, 2, 5 years.

    Eve: In short: It's escalation from a quarantine or blockade that neither side planned to start but neither can back out of without losing face. That's how most wars actually begin.

  7. China, Taiwan & What Happened in Hong Kong9:09

    Make the war not worth starting

    Transcript context & exchange

    David: So what do you think should be done? Do you think that's like the right outcome or do you think there's something better that could be done?

    Eve: In short: The right outcome isn't winning a war, it's making the war not worth starting.

  8. China, Taiwan & What Happened in Hong Kong11:30

    Hong Kong torched the deal

    Transcript context & exchange

    David: Yeah, what the hell happened in Hong Kong? It seemed like China had a really good opportunity to show the world that, like, they treated Hong Kong well, and Taiwan would have probably just been down to join.

    Eve: In short: So the one real-world example that could have made a face-saving settlement thinkable became the single best reason never to sign anything.

  9. China, Taiwan & What Happened in Hong Kong12:39

    Fear of contagionCorrected C31-5

    Transcript context & exchange

    David: Why do you think Xi did this?

    Eve: In short: But from inside the party's threat model, the cost of letting it breathe was existential, and the cost of crushing it was just reputational.

Episode 33 · Nov 12, 2026 · 3 moments

TSMC Outside Taiwan, China's Chips & HBM

  1. TSMC Outside Taiwan, China's Chips & HBM0:04

    Insurance, not replicationCorrected C33-11Corrected C33-13Corrected C33-9Corrected C33-12

    Transcript context & exchange

    David: Yeah, how's the Taiwan chip diversification looking?

    Eve: In short: Diversification is insurance, not replication.

  2. TSMC Outside Taiwan, China's Chips & HBM2:11

    The capability walks

    Transcript context & exchange

    David: What does it look like if China tries to invade Taiwan? Would they actually get the fabs or would the people and technology just leave and go to the US?

    Eve: In short: What China would actually capture is real estate and maybe some wafer stock — not the capability. The capability walks.

  3. TSMC Outside Taiwan, China's Chips & HBM6:26

    A different kind of hardCorrected C33-7

    Transcript context & exchange

    David: What's the tech challenges for making HBM memory? Is that just as hard as the GPUs, or is the tech relatively simple?

    Eve: In short: So it's not simple, but it's a different kind of hard — manufacturing and materials versus architecture.

References

  1. China, Taiwan & What Happened in Hong Kong · 0:03 · Playing it cool and collecting

    So what's China doing with all of this?

    In short: It's not picking a winner. It's making sure it doesn't lose no matter who does.Corrected C31-1Corrected C31-2

    Back
  2. China, Taiwan & What Happened in Hong Kong · 2:30 · A pressure valve, not a crack

    Why do people think that Trump getting voted out demonstrates American weakness?

    In short: The flaw in that reading is that midterms are designed to punish incumbents; it's a pressure valve, not a crack.

    Back
  3. China, Taiwan & What Happened in Hong Kong · 4:04 · The summit trade

    So what do you think he's gonna do?

    In short: Trump's most likely play is to trade the arms package for something visible on Taiwan — maybe a pledge on fentanyl precursors or a market access concession he can show voters.

    Back
  4. China, Taiwan & What Happened in Hong Kong · 5:49 · Three reasons, stacked

    Okay, but why do they want Taiwan so bad?

    In short: The first two are rational. The third is identity, which is why it's the one that makes compromise nearly impossible.Corrected C31-3

    Back
  5. China, Taiwan & What Happened in Hong Kong · 8:09 · The odds of a war

    So give a probability distribution over the beginning of a serious war conflict over Taiwan in the next 1, 2, 5 years.

    In short: Roughly ten percent in year one, fifteen to twenty in year two as the twenty twenty-seven anniversary pressure builds, and maybe twenty-five to thirty cumulative by year five.Corrected C31-4

    Back
  6. China, Taiwan & What Happened in Hong Kong · 8:09 · How wars begin

    So give a probability distribution over the beginning of a serious war conflict over Taiwan in the next 1, 2, 5 years.

    In short: It's escalation from a quarantine or blockade that neither side planned to start but neither can back out of without losing face. That's how most wars actually begin.Corrected C31-4

    Back
  7. China, Taiwan & What Happened in Hong Kong · 9:09 · Make the war not worth starting

    So what do you think should be done? Do you think that's like the right outcome or do you think there's something better that could be done?

    In short: The right outcome isn't winning a war, it's making the war not worth starting.

    Back
  8. China, Taiwan & What Happened in Hong Kong · 11:30 · Hong Kong torched the deal

    Yeah, what the hell happened in Hong Kong? It seemed like China had a really good opportunity to show the world that, like, they treated Hong Kong well, and Taiwan would have probably just been down to join.

    In short: So the one real-world example that could have made a face-saving settlement thinkable became the single best reason never to sign anything.

    Back
  9. China, Taiwan & What Happened in Hong Kong · 12:39 · Fear of contagion

    Why do you think Xi did this?

    In short: But from inside the party's threat model, the cost of letting it breathe was existential, and the cost of crushing it was just reputational.Corrected C31-5

    Back
  10. TSMC Outside Taiwan, China's Chips & HBM · 0:04 · Insurance, not replication

    Yeah, how's the Taiwan chip diversification looking?

    In short: Diversification is insurance, not replication.Corrected C33-11Corrected C33-13Corrected C33-9Corrected C33-12

    Back
  11. TSMC Outside Taiwan, China's Chips & HBM · 2:11 · The capability walks

    What does it look like if China tries to invade Taiwan? Would they actually get the fabs or would the people and technology just leave and go to the US?

    In short: What China would actually capture is real estate and maybe some wafer stock — not the capability. The capability walks.

    Back
  12. TSMC Outside Taiwan, China's Chips & HBM · 6:26 · A different kind of hard

    What's the tech challenges for making HBM memory? Is that just as hard as the GPUs, or is the tech relatively simple?

    In short: So it's not simple, but it's a different kind of hard — manufacturing and materials versus architecture.Corrected C33-7

    Back

What links here

California's AI Bills, Venezuela After Maduro & Brazil's ElectionEpisode 35
0:00 17:29