Audrey, what was the most important measure you put in for Taiwan’s digital democracy during your service?

Shall we record this?

Okay, here we go.

Even this — radical transparency — was very important early on. When I joined the Cabinet it was right after the Sunflower Movement , in which President Ma’s approval rate was at 9 percent, mostly because people had lost trust in vertical institutions. For the past 12 years — starting around the Sunflower period — Taiwan has been the top target (according to V‑Dem ) for polarization attacks and foreign information interference. So, to design defense and resilience, you have to shift from a vertical mode of trust (people trust an institution because it holds the “high ground” in elite conversations) to a horizontal mode (people trust those who sound like them and learn like them — everyday people).

This co‑creation — this “ uncommon ground*” that discovers common ground across ideological camps — needs to be *more viral than polarization. We have to be more popular than the populists. Radical transparency helped investigative journalists, civic teachers, and others build civic care into the basic education curriculum.

If there is only one thing I would point to, it is the 2019 education curriculum reform, where we shifted from media literacy (consuming information) to media competency (co‑producing information). Primary schoolers learned to measure air quality and create shared knowledge about pollution sources. Later that citizen map morphed into a mask availability map , which helped us overcome mask Vs. anti‑mask polarization and later vaccination Vs. anti‑vaccination polarization.

The idea of humor over rumor eventually dominated public conversation in Taiwan, so polarization attacks became anti‑fragile : we grew more resilient the more we received those external attacks.

There is so much to talk about. This radical transparency— you , personally, as someone holding office—made it a norm: “If we’re going to talk, it’s live.” That’s a radical idea. In the U.S., many people are disenchanted because the system feels opaque despite being democratic. You can imagine members of Congress adopting this practice. Within Taiwan, did that practice spread throughout the political system?

Yes — quite a lot. Initially it was just me and my office, because it was costly to transcribe in real time. Someone would record video; a stenographer — a court reporter — would sit next to us, typing everything. We adopted that in Cabinet meetings and other important meetings, but because it required very skilled court reporters, it was not close to 100 percent.

Nowadays, with language models and AI, this is no longer a problem. They are now more accurate than humans — not just at transcription , but also summarization and translation . In digital deliberations , people expect real‑time captions and transcripts — and not just that, but real‑time summarization, topical clustering and analysis — almost like a group selfie , so people can see where they are across ideological camps.

Civil society now expects this kind of sense‑making . The U.S. is adopting it too: Bowling Green, Kentucky’s What Could Bowling Green Be? initiative uses the same code (Google Jigsaw sense‑making + Polis) for bridging conversations. Work is underway to roll out similar bridge‑making in all 435 U.S. congressional districts. The goal is to show it’s faster than regular polling, as rigorous , and far more generative — because, while standard polls force you into a few bits (YIMBY vs. NIMBY), deliberative methods are generative, transformative and deliberative .

As a political scientist, I see public‑opinion data as thin—you ask people to rate agreement on a Likert scale, reducing complex thought to a number. Using focus groups or deliberative groups at scale is promising. Many people are pessimistic about AI’s effects on democracy—deepfakes, the loss of reading—but you seem to believe these tools can improve representation . Can we talk about group/online deliberation? In another interview you mentioned deepfakes— a lot of them coming out of Shenzhen and beyond —and the use of videotaped focus‑group deliberations that bring together people not living solely within their own political spectrum bubbles. Tell me more, especially because Republicans and Democrats at the mass level are increasingly segregated; even in purple places, people don’t interact, breeding dehumanization—even at family gatherings. I’m a big believer in getting people back together— even if digital .

The software we use is from James Fishkin’s lab at Stanford’s Deliberative Democracy Lab. It is a deliberative polling system. Functionally, it is like Zoom with breakouts of 10 people each (say, 45 rooms at once), but the software facilitates — no human moderator. It reminds quiet people to speak up, limits disruptive interjections to a few seconds, provides real‑time transcripts and runs quick quorum checks to see if groups are ready to move on — like an advanced chess clock that carries conversation guides to their logical extreme.

Those 45 rooms produce transcripts concurrently; we run sense‑making in the background with language models. By the time they are done with a topic, we can calculate the proportional veto core or surface the uncommon ground — packages that make everybody slightly happy and nobody very unhappy. That is a very different way to aggregate qualitative data. What used to take days of undergrads, LLMs now do instantly .

So, even mid‑conversation, we prepare a slate that is mutually comprehensible and reinforcing — think Overton window . We read it back to participants and experts: e.g., “Most of you want all political ads to carry a digital signature and advertiser KYC; most of you also want platform liability for damages from unsigned ads (say, a US$4 million scam). You want ByteDance to set up a Taiwan office, and if they refuse to accept liability, you want throttling of connections to their servers until business goes elsewhere,” and so on. People can see this is not content moderation or censorship ; Taiwan still ranks at the top in Asia for Internet freedom .

By the end of the day — although at that time no party had more than 50 percent of seats — we were able to show a comprehensive slate to the Legislature. They passed it without a cooldown period: the Digital Signature Act in May (after a March deliberation) and the Anti‑Fraud Act in July. This year there is essentially no so‑called “deepfake ads” activity operating in Taiwan.

Turning deliberation into policy — politicians acting reliably — might be an obstacle in the U.S. Another question: Who are the people in these deliberations? Do you get a representative cross‑section of Taiwan?

We send text invitations to about 200,000 random numbers. People volunteer; then we use stratified random sampling — exactly how a rigorous pollster would.

In polling we worry about silence —people refuse to participate, or skip items. “Silent voices” aren’t heard; they often don’t vote. The politically active skew older, wealthier, more extreme. Did you face this in Taiwan—populations hard to bring in?

Not at all. It is no coincidence we have among the lowest affective polarization scores in OECD‑equivalent societies — across urban–rural, ethnicity, gender. We literally leave no one behind. In rural communities, outlying islands, high mountains — broadband is flat‑rate, all‑you‑can‑eat for about 14 euros per month. For my 92‑year‑old grandma it is easier to participate online than to physically attend a town hall.

We also close the loop quickly. Citizen initiatives are guaranteed a response within 60 days; during the pandemic we sometimes responded in 24 hours. The faster you close the loop, the more trust you build — a flywheel. Once your petition is realized in 60 days, you are more likely to mobilize friends and neighbors to propose more.

There’s a rich literature on deliberative democracy: it improves trust and reduces polarization, but it’s historically expensive and hard to implement. U.S. democracy has thin participation—vote, donate, protest—intermittent and not very deliberative. Your practices could help. You mentioned Bowling Green; what about California?

Engaged California ( engaged.ca.gov ) has already pre‑committed to accepting deliberative results from that platform: first on wildfire mitigation and prevention, now on government efficiency — think DOGE, without the chainsaw — inviting public servants to propose how to make government more efficient, including using AI — empowering people closest to the pain .

When I joined the Cabinet in 2016 we were like the OG “DOGE” — radical transparency, publishing scoreboards, changing procurement rules to cut unnecessary spending. But we worked with the bureaucracy, not against it. We announced every move and waited 60 days before regulatory changes; most ideas were crowdsourced from public servants.

One interesting feature: you worked in government while proclaiming to be an anarchist. Did that create issues as a manager with a very hands‑off style?

I still am an anarchist. Anarchism is counter‑coercion, not anti‑governance. If you have a good multi‑stakeholder institution — the way the Internet is governed — Kropotkin would be happy. I am against top‑down, coercive takedowns, shutdowns, lockdowns.

In Taiwan the people understand I am at a Lagrange point between social movements and government — working with government, with the people, not for them. I resist capture by either orbit. My focus is on civic‑care capacity. I do not have a platform other than the platform. I do not need top‑down coercion because I do no thave a pre‑concluded policy outcome — only a procedural outcome: people’s ability to be attentive, responsible, competent, responsive, in solidarity . That shields me from partisan capture and polarization.

It’s delightful—and rare—to hear someone say the goal is a process ( civic care ), not a specific policy outcome. In political science we see the rise of anti‑system parties that operate in democracies but don’t really believe in democracy, trying to gut it from the inside. We need people committed to democracy above all. People with extreme policy positions are more prone to authoritarian sympathies; they’ll undermine rules to get outcomes they want.

Social media’s engagement‑through‑enragement machine prioritizes those voices.

Calm, nuanced conversations don’t sell—but there’s demand when people find them.

Segue to humor over rumor . How do we navigate algorithms designed to pull us apart with misinformation? Taiwanese media has division too. Using tech and AI, tell me about humor over rumor—and your success.

Finding de‑polarizing content and uncommon ground is step one. Step two is making it more viral than polarization — more popular than the populists . Humor helps.

Early in the pandemic, we saw polarization: some believed only N95 masks work; others said it is about ventilation and masks harm . Science was not conclusive, and the virus kept mutating. Taking content down can backfire when the science shifts.

Using uncommon ground , we found both sides were okay with the message: wear a mask to remind yourself to wash your hands — because a mask keeps your hands away from your mouth. Nobody disputed that. So we pushed a cute Shiba Inu meme: “Keep your mouth safe — wash your hands; wear a mask to remind yourself.” We measured tap‑water usage — it rose. It was cute; people shared it.

In 2021, we published a scoreboard of preferences for four vaccine brands by age and district. Some elderly people did not like AstraZeneca (news of blood clots in Europe) and preferred Moderna. Some preferred BNT (Pfizer‑BioNTech, secured via TSMC); younger people preferred Medigen (homebrew), others distrusted it. The scoreboard turned what could have been vax Vs. anti‑vax energy into “my team Vs. your team” dynamics. People competed vigorously; at the end, everybody got vaccinated. There was no oxygen for an anti‑vax political faction. The principle is pre‑bunk, not de‑bunk.

We did not require much from platforms beyond no paid amplification of political/social content by non‑Taiwanese — treat it like lobbying. Taiwan was the first jurisdiction, I believe, where Facebook’s Civic Integrity team implemented strong KYC in election seasons against foreign‑sponsored content.

Openness brings vulnerability: authoritarians exploit it to seed polarization. If we react the wrong way, we become more authoritarian—the “anti‑terrorism move.” I still believe openness wins ; U.S. universities thrive by attracting talent from everywhere, including China. Even if 1% had bad intent, 99% contribute. But Taiwan faces a real existential threat and more disinformation than anywhere. How do you preserve openness while acknowledging the threat?

It is not a trade‑off if the system is anti‑fragile : the more “virus” you get, the stronger your immune system — people exercise civic care and grow a six‑pack of care .

It only becomes a problem when an open society allows a monopoly that becomes impossible to exit — then the monopolist sets the rules. TikTok is a prime example — the only large‑scale social network in Taiwan that is also a cyberarm product. Research showed that if you ask an average U.S. undergrad to move off TikTok alone, you’d have to pay about US$60 per month to offset FOMO. But if there’s a magic button that moves everyone around them at once, they would pay US$30 per month to go somewhere more sane. That is a product‑market trap.

The State of Utah is paving new ground as a family‑loving state: it passed a Digital Choices Act . Starting next July, if you are a Utah resident and you move off TikTok to, say, Bluesky, TikTok will be legally required by Utah to keep forwarding new likes, reactions, friends, and content to your new network — like number portability in telecom. Without portability, there is nowhere for new competitors to gain traction because FOMO is too large. With social portability, users can move individually without getting re‑captured.

Europe’s Digital Markets Act already mandates interoperability for messaging; they are consulting on extending it to social networks. In Taiwan our AI Basic Act just passed first reading with similar provisions, and I am proposing a Data Innovation Act patterned on Utah.

Taiwan is rare: we do not ban TikTok outright. We ban it in the public sector (harmful cyberproducts) and in basic education environments. But if you are 18 and using your own device and network, you are an adult. Interestingly, TikTok installs in Taiwan have been gradually declining, and we are now one of the largest national segments on the fediverse, mostly through Threads federating with interoperable networks. People prefer a world where they can post on Threads but consume elsewhere (Bluesky, Mastodon, etc.).

In the U.S., big platforms feel partisan and polarized—X under Musk leans far‑right; Bluesky is “over the road”; Truth Social is its own thing.

But everyone is on LinkedIn.

I’ve found YouTube reasonable for longer, nuanced conversations.

I have many questions, but I want to be sensitive to your time.

No worries — we have this place until five.

I didn’t come here to talk about cross‑Strait politics, but I’d love your perspective on the state of things right now, particularly given uncertainty in the U.S. Trump is transactional. There’s bipartisan support for Taiwan, but Trump can shift the Overton window. How are you feeling about the U.S.–Taiwan relationship at this moment? Is there uncertainty creeping into Taiwanese society?

Most of my work — the things we just discussed — was during Trump 1.0. Not my first rodeo. Even then, we worked closely with Keith Krach, then under secretary of state, to popularize the Clean Network — now often framed as trusted technology — so supply chains (TSMC and others) source from vendors that do not cede control in exchange for state subsidies from CCP‑aligned entities.

That was our “uncommon ground” in 2014 during the Sunflower Movement, when half a million people deliberated — many more online — on whether to allow Huawei and ZTE into our then‑new 4G infrastructure. The consensus: keep reassessing whether so‑called “private” PRC firms are de facto state‑run; amortized over time the risk makes them more expensive than alternatives.

We now cooperate closely with the U.S. on drones and unmanned systems. Palmer Luckey of Anduril was just in Taiwan discussing “hellscape” scenarios and asymmetric defense. In modern warfare, if you have strong production capacity and edge swarm AI, it is less about conscripting bodies and more about adapting AI systems. Today offense has an edge; in a few years — like in cyber — defense may dominate.

Much of our military modernization follows this asymmetric logic — we cannot afford attrition with the PRC. This fits military‑industrial priorities in D.C.; they want us to buy drones and co‑design algorithms. Whether the Republicans or Democrats , the technological underpinning does not change — that is why I am optimistic about deterrence and de‑escalation. If the U.S. does not have to send soldiers to defend Taiwan, then of course the U.S. will support Taiwan — it is good business.

Public‑opinion data in the U.S. suggests a majority would defend Taiwan, even if they don’t fully grasp the implications.

Is there anything you want to ask me?

Which other democracies are you studying or learning from?

Not yet—this is new for me. We have a wish list of reforms for the U.S.: larger, multi‑member districts so we aren’t stuck with two parties; our system (two parties, bicameral Congress, presidentialism) was designed centuries ago. We’ve learned a lot since. If Pennsylvania were one multi‑member district, we might get 3–4 parties instead of two mega‑tents where AOC and Joe Biden share a party, or MAGA and moderates share the GOP.

Duverger’s law says single‑member majoritarian systems converge to two parties—that’s where we’re stuck. There’s serious conversation about multi‑member districts in Congress; members are tired of being gerrymandered out every decade. I don’t think it will change soon; things may get worse before they get better. Trump exposed core issues: money in politics (our elections cost ~40× Germany/UK), constant fundraising, corruption. We should require tax returns, ban stock trading by officials—basic steps.

In Taiwan we shortened tax filing from 3 hours to 3 minutes — real‑time filing and feedback.

Exactly—our IRS capacity was gutted; audits feel arbitrary. We’re in a crisis—but crisis creates opportunity. We need a bigger tent. I’m seeing deeper civic engagement than ever—some of the largest protests in U.S. history this year; ~3× more protests than in 2017. People are active, though still too partisan.

How do you feel about state‑level advances? Maine moving against super PACs, Utah’s radical interoperability we discussed, Engage California, Kentucky’s “broad listening,” Oregon’s citizens’ assemblies — policy juries moving from the judicial to the executive space. Do you prefer federal‑level change first, or states as labs?

We like to say states are laboratories of democracy—and autocracy. Some innovate inclusion; others innovate exclusion. Political scientist Jake Grumbach calls them laboratories of backsliding. Examples: felon disenfranchisement—e.g., in Florida a very large share of Black men can’t vote due to past felonies; the 2000 election was decided by fewer than a thousand votes there.

I don’t want to make it all Dem vs. GOP, but it’s hard to see the GOP innovating inclusion right now—too much coercion, too little civic care. At the federal level, debates fixate on redistribution—taxes, health care—rarely political reform. People think we can’t amend the Constitution, so they don’t try. That leaves states with more space to innovate.

I was reading Patrick Deneen on the drawbacks of liberalism and restoring civic institutions, including spiritual associations. DEI in universities and tech rarely considers spiritual diversity. Removing those practices creates a vacuum filled by conspiracy and rabbit holes. Maybe restoring faith and family is easier than electoral reform.

We need people to self‑organize. American civil society has declined for 40–50 years. Robert Putnam’s Bowling Alone was right: human connection matters. Churches, unions—everything’s eroding, especially for the young. I’m hopeful this moment forces reconnection in person. As a digital thinker: what do you make of the phrase, “Online community is not community” ?

Try Partifu l — an app that makes it trivial to host house parties; it amplifies in‑person connection. In Taiwan we are among the least lonely societies by various measures: community colleges, vibrant spiritual practices (I am a Taoist), the second‑most religiously diverse society. Highly educated people often practice spirituality publicly. That strengthens the third sector.

I was also called the minister of Plurality. I believe the digital space can connect otherwise unrelated people — and those connections translate offline. The two go together.

The density of civil society here is striking—more temples than 7‑Elevens (and there’s a 7‑Eleven on every block). People often think Taiwan’s vibrant democracy is about letter‑of‑law institutions, but it’s also about civic participation. Martial law ended not that long ago—how did democracy take hold so deeply so quickly?

Two reasons.

1. Civil society legitimacy even during martial law: disaster relief for earthquakes, typhoons. On an island with daily tremors, you cannot afford to splinter; you exercise civic muscle in a constant civic gym.
2. The existential threat from an authoritarian, atheist CCP regime. Many fled to Taiwan during the civil war — bringing incredible religious and intellectual diversity (and cuisine). That lineage is everywhere: cooperatives, consumer unions, credit unions. The ROC tolerated and showcased vibrant civil society — “Free China” — especially against the Cultural Revolution.

Both civil society and the state had reasons to promote plurality. And yes — taking out the trash together every night helps build everyday civic habits.

Certain crises necessitate cooperation; they seed civil society. In the U.S. we’re in a political crisis; for the first time, 30–40% think democracy itself is dangerous. Our democratic muscle was unexercised; it’s finally being reactivated. But the news cycle is psychologically destructive— doomscrolling breeds demobilizing anxiety. We need to channel optimism.

Once we take the right action, hyper‑arousal fades.

How do you look at the mainland these days? I haven’t gone since before the pandemic. The security environment changed; exit bans; it doesn’t feel welcoming. In the meantime, our expertise atrophies. What’s your sense of the political climate in China? Rumors about Xi—legacy, health—elite churn. What’s your hot take?

That is more my dad’s expertise. He was a journalist in Tiananmen from May ’89, left on June 1 — good for our family; a colleague was shot on June 4. He later studied the student‑group dynamics for his Ph.D. and attended a Berlin reunion a year after the Wall fell. We were close to reformers; my living room in Germany had very young people thinking about democracy and technology. That is my starting point.

More people inside the PRC now feel that the once ironclad control of speech pre‑COVID has weakened. The lockdowns went on too long, testing mental health and endurance, leading to the White Paper/A4 movement. The swiftness of policy reversal after direct action bolstered confidence in mass demonstrations. You now see more exercise of speech — and even some association — compared to the height of Xi’s control.

Exhibit A: deepfakes. I made BBC news by showing that DeepSeek and other PRC models lacked guardrails against deepfakes and info‑security harms. You would expect them to purge Tiananmen from training data, but a single prompt can surface it. Simple narrative control is failing.

Today LLMs run locally on laptops at near GPT‑4 quality. You cannot deny access — it is just a file of a few hundred gigabytes containing much of Wikipedia and more; one can surgically ablate alignment layers, yielding base models that say the quiet parts out loud. People in the PRC are making such models; they’re not perfectly rewriting history.

There are tamper‑resistant, censorship‑resistant tools — zero‑knowledge payments, privacy primitives. These used to be nerdy, expensive, GPU‑hungry; now they are democratizing. You no longer need to be a master coder — you can speak to a model and it assembles tools. My competitive coding edge is diminishing. That democratizes mobilization even inside the Great Firewall, offline .

This echoes older debates. In the 2000s, scholars like Guobin Yang argued the party couldn’t withstand the Internet—optimism about collective‑action potential. Then came the backlash: the CCP mastered control—censorship, the 50‑cent army. Now your point is that new tools may obsolete old controls. It’s cat‑and‑mouse. The party is leading the world in containing generative AI.

The battle is still very much on.

Within political science, China studies leads broader authoritarianism work because the CCP is the market leader in control. Techniques diffuse to the Middle East and elsewhere. In U.S. discourse, you now see people openly embracing authoritarianism. Curtis Yarvin debating Glen Weyl about monarchy—are we really doing this?

Yarvin’s CEO‑king idea got a real‑world experiment with Elon and DOGE — to very mixed success. The good thing about Americans: you are not afraid to experiment and share results.

Let me pivot back to Taiwan’s democracy. There’s a narrative—e.g., Lev Nachman in Foreign Affairs —about potential backsliding. How resilient is Taiwan’s democracy?

Very. We had four weeks of heightened polarization before July 26 — before a mass recall some dubbed the “Great Recall.” It was the first time since 2018 that something resembling McCarthyism gained traction: talk of impurities, treason laws. But it quickly fell out of favor after the vote. Taiwan voters rejected populist McCarthyite rhetoric.

You might expect Eric Chu of the KMT to be jubilant; instead he was humble and did not escalate. Society healed rapidly. We are now heading into a nuclear‑power referendum, historically polarizing — yet this is the least polarized referendum I can recall.

Democracy requires restraint. Parties must avoid pushing power to the max or treating each other as enemies. In the U.S., a classic breach was Mitch McConnell blocking Obama’s Supreme Court nominee—legal, but norm‑shattering. From the outside, it wasn’t a good sign when a party tried to remove legislators en masse via irregular methods.

They argued self‑defense: the Legislature was trying to deny the Cabinet’s budgeting power, deadlock the Constitutional Court, and even zero out parts of the Control Yuan (we have five branches; the Examination Yuan was unaffected). Civil society launched the recall as democratic defense against cross‑branch overreach. The Legislature dialed down those attempts; by the recall vote, most assaults had stopped. After the Great Recall fail, a recalled member cannot be recalled again that term — so we shall see if survivors go hawkish or de-escalate. Signs point to de‑escalation.

Tell me more about cross‑Strait salience. In the U.S., observers reduce everything to DPP = away from China, KMT = closer to China, ignoring domestic issues. Did recent months mean the cross‑Strait issue is less salient, and voters just wanted change after DPP’s long run?

Neither the KMT nor the DPP holds a majority in parliament; the people do not want dominance by either. In the last presidential race, none of the three candidates even entertained “One Country, Two Systems,” especially after Hong Kong’s National Security Law. What can Xi say now? The ROC national flag (“free China”) and the Taiwan independence flag (“free from China”) flew together at this year’s anti‑authoritarian interference rally — first time in 40 years. Call it ROC Taiwan or Taiwan ROC — same polity now.

Also, independents — people like me who do not attend party rallies — are the largest segment: about 40 percent. More people believe in direct action: e‑petitions, participatory budgeting, deliberation — democratic expression outside party sorting.

Is TPP the outlet for younger frustration?

Initially, yes — mobilized younger “none‑of‑the‑above” voters. But recently the TPP has maneuvered to be a better‑branded KMT — aligning with older KMT constituencies rather than just broadening the youth base. This pivot followed the speaker/deputy speaker vote, where KMT secured both. If TPP had gotten a deputy speaker seat while DPP took the speaker, TPP might have pivoted the other way. It is a three‑body problem.

On models for cross‑Strait relations: “One Country, Two Systems” is dead. Is there any alternative—an EU‑style arrangement, some supranational body? Being subject to Beijing’s laws is a non‑starter. Perhaps we’re just kicking the can, hoping the mainland changes.

If asymmetric deterrence is fully in place — so allies do not lose soldiers defending Taiwan — then a hostile takeover will not be an option. At most you get moves on Kinmen and Matsu or blockade scenarios — but our war‑gaming shows blockades can escalate quickly. Politically, Xi gets more juice squeezing Taiwan than risking thermonuclear war.

Trump said Xi told him that as long as Trump was president, he would not invade — I decline to comment further. The point is: once deterrence is credible and the kinetic domain becomes defense‑dominant, Taiwan’s people will be more willing to talk. We are fine with interdependence where others depend on us (our chips power PRC science too). We are just not okay with one‑way dependence — us depending on Beijing.

Beijing often says it won’t talk to the DPP. “Only Nixon can go to China”—would meaningful progress have to come from the DPP?

Again: deterrence first — kinetic, cyber, information. Only then do talks become politically viable. By then, KMT might rebrand itself as a culturally conservative party rather than “pro‑Beijing.” The partisan split may be on other issues.

In the U.S., support for Taiwan is strong—but there’s a performative aspect and creeping McCarthyism about China. I worry we’ll hug Taiwan to death—e.g., legislation for formal recognition could trigger a rupture or war. Beyond weapons packages, what does authentic support look like—citizens and government?

The only way Beijing can take Taiwan without firing a shot is by convincing the world that democracy never delivers and only leads to chaos — the authoritarian narrative. In 2014 some in Taiwan flirted with that; not anymore. But in the U.S., both extremes risk buying into that narrative. If the U.S. fragments — culturally or kinetically — into civil conflict, the PRC will point to that and move without kinetic escalation.

So, the most useful thing Americans can do for Taiwan is to foster civic care and civic muscle at home. No democracy is an island — not even Taiwan. Without a vibrant U.S. example, backsliding accelerates globally. I have talked about this with large U.S. livestream audiences; people across the spectrum agree.

How does the U.S. stand up to China? By making our democracy work—health care, roads, trains—the lived quality of democracy.

Exactly — the stable attractor of the first Cold War wasn’t just missiles; it was rock ‘n’ roll and blue jeans . That made it possible for Taiwan to democratize peacefully — both the ruling KMT and nascent DPP looked at American democracy and said, “We want that,” so we did not kill each other on the way.

Today you see backsliding because there’s no obvious model like the U.S. of the 1950s.

Techniques that undermine democracy diffuse—Turkey, Hungary, Brazil, El Salvador—sharing notes.

Yes, that is why the U.S. example matters.

As for Yarvin’s CEO‑king idea: the DOGE experiment shows that unchecked CEO power yields mixed results. Public experiments are valuable; we learn by trying.

Do you travel to the U.S. regularly? Would you visit Princeton? We have CITP—the Center for Information Technology Policy —bridging CS, statistics, and public policy.

Sure; I am in the U.S. every other month. I am also an AFP Fellow in Philosophy at Oxford along with Alondra Nelson, working on machine ethics — training multi‑agent systems to practice civic care. That sits at the intersection of CS, politics/mobilization, and philosophy — natural allies for CITP.

Wonderful. I’d love to host you. About our recording—what’s your protocol?

We can negotiate, but usually I make a textual transcript, lightly paraphrasing, send it to you to co‑edit, and once we are both happy, we publish it to the public domain.

Perfect—send it to me.

Of course. Thank you for your time.

Thank you.