Would you be okay if I record?

Sure—of course. Feel free to do more recordings or redundancy. The recording quality on this MacBook is pretty good.

Perfect. I’ll set up another recorder as a backup. Let me start by describing my project and how I hope to use the information.

Of course.

I’m working on two projects using the same set of interviews. The first is a Global Taiwan Institute (GTI) project. I’ve received a scholarship to do fieldwork here. The investigation looks at digital democracy in practice from the perspective of practitioners—civic tech volunteers, and public officials (domestic and foreign)—to see how public‑administration theories stack up against reality. This GTI project focuses specifically on Taiwan. My goal—fingers crossed—is to write a report by the end of this year, possibly with a public event about digital democracy. I’ve been here for about a month doing interviews.

You’re here for a critical month—there’s been a lot of activity.

Yes. I started by attending a hackathon early last month to see what’s being discussed now. I also plan to go to the Facing the Ocean hackathon in about a month in Kaohsiung, to see the topics discussed among Japanese, Korean, and Taiwanese civic‑tech groups and public administrators. It’s been a very active couple of months—many meetings, conferences, and great conversations. That’s the first project, the GTI project.

The second is my dissertation. I’m working on a PhD in public administration; I passed my candidacy this April.

Wow—congratulations!

Thank you. The next big step is… working!

I hope to compare Taiwan with other countries. My first comparative topic is Taiwan and South Korea—digital democracy in practice—looking at relationships between civic tech and public administrations in both countries to understand differences and similarities. I may expand the cases to include Puerto Rico, where I recently learned about civic‑tech groups working on anti‑corruption and disaster resilience—issues that also feature in Taiwan’s civic‑tech life.

That’s great to hear. Between Puerto Rico and Taiwan—or other countries too?

Possibly South Korea as well, and maybe others; case selection is where I’m at right now. Taiwan is my core case.

What’s your research question?

That’s what I’m figuring out. The beauty of fieldwork is using on‑the‑ground data to interrogate the questions you arrive with. Originally, I wanted to understand the challenges and opportunities for institutionalizing digital‑democracy practice from a public‑administration perspective.

I wondered whether public‑admin frameworks could define digital democracy—especially “network management” and collaborative‑governance models. Perhaps digital democracy is a collaborative effort between governments and citizens that can be analyzed as a network—civic tech as one essential node alongside public administration, etc.

That sounds very Manuel Castells—network theory.

Exactly. I came thinking that way, but from the conversations I’ve had, it’s a complex picture.

Right—the “network‑making power” Castells analyzed was specific to the Occupy era he directly experienced. Once practices are institutionalized, you can’t analyze them only with networking power or network‑making power. I sometimes describe a different lens: a care framework—attentiveness, responsibility, competence, responsiveness, solidarity, and so on. It’s a different angle. Network theory analyzes power ; care theory analyzes relational health . They’re dual perspectives: one puts nodes (individuals) first‑class; the other puts edges (relationships) first‑class. From a relational view, individuals are intersections of the relations that inhabit us—which is the view I advocate in the book I call Plurality .

I’ve read it—though I may have missed the latest iteration.

The Japan stories are now part of the book as well. For example, Takahiro Anno, a sci‑fi writer and AI engineer in his early 30s, read Plurality , decided to run for office without partisan backing, won the most votes anyone under 40 has received, and is now a member of Japan’s House of Councillors. His new “Future Party” (Team Mirai) won more than two percent of the national vote, explicitly running on plurality, broad listening, bridge‑building, and depolarization—essentially the book’s platform.

I’ll have to follow that story. For our chat, I’d love to hear your experiences and thoughts on the evolution of digital democracy in Taiwan—past, present, and maybe the future. To start, is there a civic‑tech project near and dear to you—one that engaged the government where you saw all sides of it? I’m sure there are countless.

A cluster of projects comes to mind rather than a single one: mask‑availability visualization, privacy‑preserving contact tracing, vaccination coordination, counter‑infodemic work, and so on. There’s no clear end to one and beginning of the next—technology evolved as the virus mutated.

During that time you were a minister without portfolio; moda started in August 2022, technically still within the counter‑pandemic period because the special act was still in effect—right?

Right. The moda began as we shifted from contact tracing and containment to a coexistence strategy focused on vaccination rates. Preparation for moda coincided with that transition. I became the founding Minister of Digital Affairs, essentially pulling into one roughly 600‑person team the government units that had interacted with me most during the pandemic.

Transitioning from minister without portfolio, were there changes in how you could carry out your work?

Certainly. As a minister without portfolio, everything was project‑based. I had no budget—not even travel funds in the early years. I remortgaged my house and paid for my own travel; there was no government pocket for digital democracy or digital diplomacy. It was intrapreneurship.

By 2022, when moda was founded, we took what we did during the pandemic and institutionalized it as public‑infrastructure budgeting. The moda is unique: as far as I know, nowhere else has a digital ministry that simultaneously includes (1) service delivery and universal service, (2) platform oversight (e.g., large platforms like Facebook and YouTube), and (3) cybersecurity. Participation, progress, and safety—in one “integrated circuit.”

In the legislature, our counterpart is the Transportation and Communications Committee—unusual globally—but we insist epistemic resilience is resilience, and all resilience is infrastructural. So we use CAPEX, not just OPEX, to institutionalize processes that outlast a single minister.

With a minister‑without‑portfolio model, projects often fade when a new minister arrives. For example, Minister Feng Yen in the Ma Ying‑jeou cabinet handed social enterprise to me; I pivoted it to social‑innovation organizations, including co‑ops and advocacy groups, and later web3 initiatives. Minister Jaclyn Tsai ran e‑rulemaking; I added more deliberative elements. That role is personal and stylistic. Infrastructure budgets are different: once parliament approves them, my successor (Minister Huang) is equally committed to carry multi‑year infrastructure to completion. One is like research; the other is development and availability.

Thank you for the administrative history. I was curious about your working relationship with Minister Tsai—around 2014–2016, during the transition.

At that time we all knew the KMT would lose the next presidential election, but they wanted to build a bipartisan bridge to the incoming cabinet. That was successful: Premier Lin Chuan, the incoming premier, was independent—non‑partisan—and had worked with both blue and green parties. The outgoing premier, Simon Chang (now Taoyuan mayor), was also independent and had worked with both DPP and KMT.

I’ve read about the “reverse mentorship” you established. From a public‑administration perspective, that relationship—public administrators seeking to build bridges with the civic‑tech community—still feels novel. Can you take me back to how it was set up and how expectations evolved?

It started even earlier when Cheng Li-chun—now our Vice Premier—served as Minister for Youth. In the early 2000s she set up the Youth National Affairs Conference (青年國是會議). Since then, the Youth Development Administration has emphasized deliberative democracy, with young people as facilitators and agenda‑setters. I’d argue this fostered intergenerational unity and helped keep young people—especially young men—from turning cynical or extremist, unlike in some countries.

Many Sunflower Movement facilitators were trained in Minister Cheng’s programs a decade earlier. We in civic tech learned deliberation from them, and they learned communication theory—Castells—from us. Unlike other Occupy movements where the “old left” didn’t invite young facilitators and leaned toward advocacy (broadcasting), Sunflower emphasized deliberation— broad listening .

The Executive Yuan then set up a cabinet‑level Youth Advisory Committee to signal that the cabinet would pivot to agendas led by young people—a tremendous gesture by Premier Jiang Yi‑hua and Vice Premier Mao Chi‑kuo. The only criterion was being under 35; I was 33. I wasn’t on the cabinet‑level committee, because Minister Jiang asked each cabinet minister to find a personal “reverse mentor” from civic tech. From the g0v (GovZero) initiative, Tony Q and I were invited by Minister Jaclyn Tsai to join her office.

What were those initial meetings with Minister Tsai like? Topics? Energy?

Frank. For instance, Kuan Chung‑ming (then Minister of National Development; later NTU president) gave a TEDxTaipei talk in 2015: “Why do citizens oppose every measure the government proposes?” He emphasized the importance of broad listening and analyzed how the government felt more distant as bandwidth among citizens increased.

He supported the vTaiwan initiative, jointly carried out by his National Development Council (NDC) and Minister Tsai. Simon Chang also participated as the cabinet’s Chief Information Officer. Initially, vTaiwan looked like part of Jaclyn’s rule‑making portfolio—responding to emerging tech—but people also wanted a general participation center: an e‑petition platform that went beyond her remit. We debated whether NDC’s Department of Information Management was the right home for the national Join platform.

So, we ran pilots outside the government (g0v, vTaiwan—Jaclyn even joined g0v hackathons) while working closely with career public servants. Reformers in government wanted citizens’ trust and were in it for the long haul. I was introduced to all level‑12‑and‑above career public servants—about 300 people—and met essentially all of them in the national HR training program across three sessions. The overall feeling: it wasn’t just us building a bridge to government; career public servants were extending a strong bridge to civil society. They know trust is easy to lose and hard to rebuild, and they were willing to trade some vertical, positional trust to learn the art of horizontal, peer‑to‑peer trust. Early conversations were overwhelmingly in good faith.

From your experience, what have been key challenges and incentives for public servants to engage civic tech or try innovative approaches?

They’re masters of institutional design and know you can’t trade time for risk . If you save time but increase risk for the minister, successors won’t sustain it. The reverse doesn’t work either: if you reduce risk by demanding lots of extra time (e.g., endless qualitative processing), it’s unsustainable.

Civic tech hunts for Pareto improvements: save time and reduce risk—or at least keep risk constant while saving time, or keep time constant while reducing risk. That zig‑zag path is how institutions change.

I’ve been thinking of civic tech as a “demo space”: citizens develop proofs of concept to show, “Here’s a problem we identified; here’s a solution that saves time.” Is that how it’s played out?

Definitely. I think of it as quantum tunneling . In classical physics, to overcome a wall (an emerging challenge), you need a lot of energy to climb it; once over, a vertical, top‑down push makes everything easy. Many social‑change theories talk about “critical mass.” Civic tech is different: instead of forcing through with overwhelming mass, it enables experiments that scale across rather than scale up . A core tenet is reproducibility.

If a team in one city runs an experiment and shares it—say at bimonthly hackathons—others can reuse parts elsewhere. An air‑quality visualization network (AirBox) solved “no one trusts the other sector’s measurements.” The same stack can visualize water quality, manage river pollution, or support local referenda. When mask availability became the problem, the same stack was repurposed overnight for mask maps and routing, with OpenStreetMap folks. The “three‑day mask map miracle” happened not because we had GPT‑5 back then, but because the infrastructure from air and water maps already existed.

So: many “advanced waves” try different hyper‑local problems. Some prove competent—saving time and reducing risk. Reformers inside the government then happily say, “Let’s make this a new institution.”

How did this approach encounter bureaucratic culture—central and local? Was there conflict?

In 2014 President Ma’s approval was 9%. Anything with legitimacy above 9% met little resistance. Urgency calls for clarity. Without that phase change, if a government enjoys, say, 70% approval—as in 2020—then “disrupting” the CECC wouldn’t work; pilots would need higher legitimacy than 70%, impossible for pandemic control. So civic tech worked with CECC. The agenda came from CECC; civic tech focused on fairness in implementation.

For example, we aimed for >75% mask access. Equalizing by time (not distance) was fairer to rural people. For vaccination, we respected preferences (e.g., some elderly hesitant about AstraZeneca), and built dashboards showing age‑by‑district vaccine preferences. Some swore by BNT (procured with help from TSMC, our most trusted institution); others preferred the locally developed Medigen; others only Moderna. We turned “anti‑vax vs pro‑vax” polarization into a healthy “my soccer team vs yours” competition that kept attention on vaccination. There was little oxygen for anti‑vax politics in Taiwan. But again, targets were set by CECC; civic tech didn’t propose alternatives to vaccination—because legitimacy was high.

I’ve been thinking in terms of three “urgency moments” over the past decade: (1) Sunflower and its aftermath, (2) COVID‑19, and (3) the current moment blending the Bluebird movement and whole-of-society resilience issues given Taiwan’s geopolitical position. Do those capture it, or are there others?

I’d add ~2018: polarization, populism, and the infodemic. It wasn’t Sunflower‑ or pandemic‑scale, but people at the mayoral election/referendums felt the information ecosystem was fragmented and balkanized. That was the only time I really felt intergenerational unity was at risk. Today it’s better. The chief populist then is now moderate and presiding over our parliament—similar to Switzerland’s experience, where participation made populists learn that civil society is more popular than populism.

A public administration innovation you proposed that excited me was the Participation Officer (PO) network—an internal “training‑wheels” network. What inspired it, and how did it evolve?

You should interview Peggy Lo. She joined the Ministry of Health and Welfare (MOHW) before I entered the cabinet and, during the KMT‑to‑DPP transition, helped foster deliberation and direct social communication in MOHW. The first Join platform citizen initiative—adapted from the vTaiwan multi‑stakeholder model—accelerated coverage for an expensive cancer drug in our national health insurance. She handled hard cases extremely well. She became our archetypal Participation Officer.

The PO network was designed to empower the “Peggys” in every ministry. Each ministry has its own culture of listening. Health relies heavily on expert opinion (atomic energy even more so); their power was translating expert knowledge into something laypeople could meaningfully engage. The Ministry of Culture, under Cheng Li‑chun, crowdsourced the Basic Culture Act—horizontally from the start. With 32 ministries and 32 cultures, a single minister without portfolio cannot direct them all. The only workable approach is fractal: each minister has trusted POs who bring their culture to a horizontal coalition, while each ministry leads its own open‑government plan. Recursively, agencies and departments have second‑level POs. Eventually, no agency fears engaging people directly on Join . That was the idea.

It also let public servants set their own agenda. Each month we deliberated two cases: one PO‑voted from citizen petitions, and one PO‑nominated by ministers. Ministers learned what topics make good cross‑sector conversations. This ran for three years before the pandemic.

During the pandemic, the PO network changed: everyone learned to deliberate online, but government legitimacy was very high, so POs took on a more communicative than deliberative role—pre‑bunking the infodemic. The MOHW PO (Peggy’s successor) became world‑famous—their Shiba Inu was the CECC’s “spokes‑dog.” Without the training wheels, we couldn’t have pivoted so quickly or effectively.

And since the pandemic?

It stayed strong. It’s all about fostering internal institutional culture within each ministry, when it comes to public engagement. No one pushes back against open data anymore—once a key PO challenge. The National Action Plan for Open Government is endorsed by all ministers and is now subject to international independent review. Participation is institutionalized within each ministry. You can debate response speed—e.g., whether the NCC should have replied to the Join petition on the Bluetooth inspection fee in one day instead of three—but that’s like saying, when submarine cables are cut, you can only stream 720p instead of 1080p. There’s room to improve, but compared to ten years ago, it’s night and day.

Were there unforeseen challenges or resistance when rolling out the PO network?

Not really. In the first couple of years of the Tsai administration, approval dipped to around the 30s, then slightly lower due to big reforms (labor law, pension reform), Constitutional Court rulings on marriage equality, etc. President Tsai governed like a “second‑term” president in her first term. Urgency and legitimacy deficits created a need for clarity—similar to the Ma‑era conditions. So there was no pushback to the idea of higher‑bandwidth communication with the public.

It sounds like urgency often motivates change. You’ve said listening produces outcomes most people can live with, leaving no one very unhappy—incremental, deliberative, and generative. But you’ve also emphasized something transformative .

Right. The primary goal is transformative instead of just deliberative or generative : people feel agency because they can help set the agenda—owning a stake in the policy outcome. You only need this when the relationship is frayed. If the government already enjoys high approval, you don’t seek to transform the relationship; you keep it resilient, not disruptive. That’s why this model works best in “warmer” times—moments of urgency that motivate everyone to invest in horizontal trust.

Looking ahead: to build and sustain digital‑democracy practices, what methods or issue areas show promise?

Citizen initiatives on Join have enjoyed a higher spike in participation in topics like the NCC Bluetooth inspection fee, child‑protection cases, and others have reached hundreds of thousands of signatures; in my time, thousands were already high.

The Presidential Hackathon empowers hyper‑local innovation with cross‑sector civic tech participation and offers winners a promise to enter public‑infrastructure as part of the fiscal budgeting cycle—a great success. Ministries rotate running thematic Presidential Hackathons. This year it’s the new Ministry of Environment; Minister Peng Chi‑ming, once a hackathon judge and a strong advocate for open data and open science, is leading it. Minister Peng is another example of a player from the civic tech and open data movement who enters into government as a cabinet minister, bringing with them a general commitment to transparency. This is extremely encouraging.

This is an example of a virtuous cycle effect of someone who first participated in civic tech at an entry stage and later took on a position of authority and was able to bring that culture into a national scale. Are there issue areas more compatible with current participation technologies?

When issues aren’t urgent or are already clear to bureaucrats, there’s no need for broad listening. If there’s urgency and/or ambiguity—e.g., “Who’s the competent authority for Uber? For the infodemic? For a legislative occupation?”—that’s where these methods shine. Lacking both urgency and ambiguity, civic engagement adds little value.

Can deliberative technology help during polarized times—say, around recall campaigns?

vTaiwan used Pol.is early in the Bluebird movement—gathering outside Qingdao East Road, from which “Bluebird” (青鳥) took its name. The constitutional roles of the Legislative Yuan vs Executive Yuan (and the Control Yuan, later the Judicial Yuan) were at issue. In rallies you saw the ROC flag beside the Taiwan‑independence flag—unprecedented in earlier civic movements. Autocratic pressure forged unity: “pressure makes diamonds.”

From late June to late July, that bridge‑building, depolarizing (transformative, deliberative, and generative) energy was muted—certain radical elements hijacked the conversation—directly leading to the epic fail of the “epic recall.” But after July 27 and heading to August 23, the civil society has been healing; I even celebrated on Threads with the hashtag “I love our majestic mountainous country”—combining “majestic” (an ROC flag anthem term) with “mountainous” (a classic pro‑independence framing). It symbolizes bridging: Taiwan is more mountainous by area, while “majestic rivers” evoke the broader ROC imagery. Combining both elements resonated.

As Taiwan’s cyber ambassador‑at‑large, looking to the next stage of technology and deliberative democracy, what are you optimistic or cautious about?

I’m optimistic about saving time and reducing risk. Previously, Pol.is‑style tools hit limits at scale because meaningful semantic aggregation was hard. With language models, summarization and topicalization are grounded enough that hallucination is negligible for these tasks—true only this year. Small pilots, like Bowling Green, Kentucky, already show good results. Jigsaw’s open‑source Sensemaker (currently relying on Gemini 2.5) works; we’re porting it to GPT‑OSS for more openness.

Now there’s a replicable open‑source stack to scale across different polities and cultures. Consensus statements translate across languages in seconds with little semantic loss. Taiwan is no longer the largest polity doing digital democracy at scale: California has engaged (we’re wrapping up “wildfire recovery” and moving to “government efficiency”); Japan has leaders like Takahiro Anno as I mentioned. Both are larger than Taiwan and similarly wealthy. Until a few years ago, outside of Taiwan’s case, practically all cases of institutionalized digital democracy were at the city-scale: Barcelona, Madrid, Reykjavik, London. All good case studies, but none were larger than Taiwan in terms of population. Now with language learning models, we are now seeing larger polities scaling digital democracy without incurring extra burden regarding time or risk on their respective public servants. With the Napolitan Institute, I believe Google Jigsaw is running broad‑listening exercises in each U.S. congressional district—the largest polity yet to try AI‑assisted sense‑making.

I’ve been reading about “AI sovereignty”—nation‑specific AI environments. How does that interact with intercultural exchange?

It’s my Oxford research topic (with the Institute for Ethics in AI) through next May. My proposal is “6pack.care”—a “six‑pack of care.” Instead of aligning AI vertically (a slave to a master—CEO, dictator, or a thin metric like engagement), we propose public AI aligned horizontally to a thick relational field.

Thin metrics—the Aladdin’s‑lamp/parasitic‑AI problem in which a “wish” of just a few words can lead to overly broad or creative interpretations by the AI—may lead to sycophancy, addiction, over‑reliance, power concentration, and out‑of‑control risks. These thin vertical alignment risks can be easily rectified by relying on thick horizontal relational fields. Horizontal alignment uses bridging algorithms like Community Notes on X/Twitter—directly inspired by Pol.is. Notes are featured not by majority vote but because they bridge polarized camps–instead of left-wing and right-wing, it identifies an up-wing , or notes that are upvoted by both camps. Recently AI can draft such bridging notes. X opened APIs so you can plug in Mistral, gpt-oss, etc., and get human votes. It’s reinforcement learning with community feedback. The AI is rewarded when previously polarized camps understand each other more—maximizing relational health within a community. When the community heals, the AI’s job is done; it retires. That addresses many failure modes of vertical alignment and existential‑risk talk. This is a promising new paradigm.

You can reinterpret Join , vTaiwan, etc., as civic‑care projects—and our AI work as aligning models to that field. In Taipei and Tainan, our 2023 “alignment assemblies” produced different constitutions (model specs) city by city. So “sovereign AI” isn’t national unison but nested stewardship : Taiwan’s 16 Indigenous nations (with 42 language varieties), Hakka, Hoklo, each city—tuning models for their cultures. That’s why we double down on open‑source AI and local, sector‑ and community‑specific fine‑tuning. We’re not training a singleton superintelligence; in Taiwan, we the people are the superintelligence.

Exactly—and the model’s responsiveness follows from that.

Right. We don’t want to be tethered to proprietary stacks (e.g., Gemini only), hence the port from Jigsaw to GPT‑OSS.

Two last questions. First: based on our conversation and your experience, how would you define digital democracy ?

In Taiwanese Mandarin, shùwèi (數位) means both “digital” and ”plural”. So I’d say: digital democracy uses digital means to provide groups with group selfies —reflections of the polity in high‑bandwidth, low‑latency ways—so they can craft measures everyone can live with.

As a Matrix fan, I can’t help thinking of “residual self‑image”—our mental projection of a digital self—and how that interfaces with the body politic.

Exactly. I’m a sci‑fi geek too—I have a tattoo on the back of my neck where the plug would go.

Second: from your definition and perspective, has Taiwan achieved pluralism?

“Achieve” is a progressive framing. I identify as a Taoist, or conservative anarchist. Conservative means we conserve the interplay between cultures—the pluralistic now—resisting homogenization or polarization into binaries. Anarchist means voluntary association, non‑coercion, shifting from vertical control to horizontal coordination and trust. I wouldn’t say I achieved anything; I would say we conserved a transcultural republic of citizens—my translation of Zhonghua Minguo (中華民國).

Always inspiring to chat. I interviewed you for a podcast five years ago, pre‑pandemic, and that conversation pushed me toward research on digital democracy. It’s meaningful to return five years later and see how much has changed—if all in the spirit of conserving pluralism.

I appreciate your thoughtful questions.

My hope with research is to add something to conversations brilliant people have already started. I take cues from John Postill’s work and want to extend that to how social movements translate into public policy and administration—your work, the PO network, from vTaiwan to Join . I’m grateful we found time to chat.

I’m happy you’re physically in Taiwan as our most recent urgency plays out—from bridging to radicalization to anti‑fragility and now healing. Intergenerational unity held this time; you don’t see elders and young people polarized. There was some risk in the four weeks leading up to the first recall vote, but we’re largely past that.

I was here during the larger Bluebird movement protests at the Legislative Yuan—there taking it all in. I was inspired to write a piece on “the two Taiwans”: Taiwan as a polity with its domestic policy debates, versus the DC‑based conception of Taiwan—war games, military scenarios. The persistent view of Taiwan as a high‑risk geopolitical factor in DC versus what people talk about here—matters that hit daily life—are two very different conversations. That inspired me to understand Taiwan from the ground and share that perspective.

No democracy is an island—not even Taiwan. Or many islands—connected by microwave.

True. Thank you again for your time.

Thank you. I look forward to your research. About the transcript: do you want me to embark on it somehow, or can we just clean it up together and publish?

Sure—that would be great.

Let’s do that. I’ll send you a link.