Good afternoon. Thank you, Minister Tang. We are very honored to be here at the Executive Yuan to have this conversation with you. It’s a highlight of our 10-day visit to Taiwan and Hong Kong.

The theme of the 2018-2019 Luce Scholars wrap-up really is about democracy and the challenges that both Taiwan and Hong Kong facing in trying to build a more democratic system.

Democracy is not just about “one person, one vote.” Governance and how government function matters enormously how the democratic system can succeed.

We are very honored that you take the time to meet with us. We hope this is a very substantive, wide-ranging conversation. I give the floor to you.

Welcome all. Here you can see my actual office. [laughs] It’s more fun. Anyway, we are meeting here, in the Executive Yuan building.

This picture is the Social Innovation Lab. This is where I usually do my public work. Every Wednesday from 10:00 AM to 10:00 PM, anyone can visit this space and book 40 minutes of my time to talk with me on anything. The only requirement is coediting for 10 days and we publish the transcript online.

Because of this principle, everybody can see very easily how is it like to be a digital minister in day-to-day. Since I’ve become digital minister two and a half years ago, I’ve talked to 4,100 people, almost 1,000 meetings. There’s 200,000 or so utterances.

This is not just lobbyists and interviews but also internal meetings. Anything that I am a chair that results in a decision, we also capture the entire context, except it’s 10 working days, but otherwise the same.

The greatest thing about this is that it’s quotable in a sense that it’s not just one fragment or the other after the meeting for people to interview. Rather, people who work on investigative journalism, for example, can understand the full context of the conversation, but also when and where this kind of conversation appears.

I would encourage you to maybe...our website is SayIt, S-A-Y-I-T. You can find it in sayit.pdis.nat.gov.tw, but if you just google for SayIt, and with any keyword, then you can have immediate access to the context of me talking about that particular thing.

Thank you for your contribution to the comments, because this conversation will be part of that radical transparency.

As a beginning I would like to just take, maybe, three minutes or so to briefly sketch what the digital minister’s role is about, and then we’ll just open for each of you to ask questions and until two hours has passed, if that’s OK with you.

In Taiwan, we have 32 ministries. Each one has a minister. That’s called a vertical minister. Above the 32, there are 9 horizontal ministers, also known as minister without portfolio, or MWOPs, interesting acronym, but anyway, horizontal ministers.

We are kind of pseudo-ministries, because each of us have an office, but office isn’t the same as a ministry as dictated by law. Each of the horizontal ministers can shape our office however we like.

My office is literally one person poached from each vertical ministry. Theoretically, that gives me 32 colleagues, but at the moment there’s only 22 colleagues, meaning that some ministries never send anyone to my office.

Like the Ministry of Defense, I wonder why? [laughs] The people who join my office still report to their minister, but I ask them to do two things. First, that everybody must work out loud.

Meaning that we must be always prepared to show the Wekan board, that’s our Kanban board, that each and every ministry’s delegate is working on and this is entirely public. Everybody can see who among those delegates are [laughs] online, and who are not online, and then what projects they are working on.

If you work in agile teams, this is a standard, so we have a shared communication facility. That’s the first thing. The second thing, aside from working out aloud, is that they are voluntary association members.

Meaning I don’t give them scores, I don’t rank them. I don’t determine their salary. They have to score and rank themselves. They have to be responsible of what projects to do.

When they want resources, this is just a platform to which they can ask other ministries’ delegates here in order to work out something to their mutual interest.

For example, a recent case is the Ministry of Transportation and communication, the Ministry of Economy, and, to some extent, Ministry of Culture all working on these self-driving tricycles. The Social Innovation Lab is like a sandbox where people can experiment how is it like to have your vicinity a lot of those self-driving vehicles rolling around.

They’re very slow so they won’t hurt people. They are open-source, open-code, and open-data so that everybody can tinker it.

Through a year or so, people just tinker for it to have two eyes, not one eye, to have an express number of expressions, to figure out the norm of having these self-driving tricycles integrating with the society, a process that we call co-domestication.

Rather than having the code decide the law, decides the market, decides people’s norms, we’re entirely the other way around. We ask people to figure out the norms that informs the market policies, that informs the regulations. Finally, we turn them into law. Norms-first approach is what we’re saying.

The governance, nowadays in the UN, they call it co-gov or co-governance. I think President Tsai put it best. She said, "Before when we think of democracy, we think of a position between two values."

It could be economic development on one side and environmental sustainability on the other. It could be disruptive innovations on one side and social justice on the other.

Each knot is like a vertical minister, and the rope here is the anonymous career public servants that absorb all the tension and gets none of the credit. [laughs] That is the bad old days.

As we realize that we cannot invent an agency whenever there is a trending hashtag, so people’s organization and mobilization has really changed.

For each emergent issue, there’s no way to have a emergent agency to look at it, so we should ask a different question. The old way is about asking, "Who are the representatives, and how do I make a fair arbitration between them?"

Now we ask a different set of question, ask if people have different positions, and we don’t know what to do. Despite different positions, are there common values? Given the common values that people discover, is there innovations that can deliver on those values without leaving anybody worse off?

That is the main thing about co-gov. We have lot of ways. We have sandboxes where people with a disruptive idea can ask for a year to break existing regulations, other than money-laundering and funding terrorism. Those two, we know what would happen, no need to experiment.

Everything else is fair game. [laughs] You’re given one year to show the people that your idea of a self-driving hybrid vehicle or whatever is a good idea.

After that year, it may be scaled out or scaled up. At any time, the MPs can say, "OK, we want a new law about it, and we are going to deliberate for three or four years."

Meanwhile, the experiment, including the business model, is still legal. This is limited-time monopoly in exchange of open innovation by sharing the data and all the things that we learned if it fails. If it works, of course, it become part of the regulation.

The other example is that I personally travel every other Tuesday or so to the rural places who are least likely to participate in Taipei policymaking. I go to where people are with technology. Instead of asking people to come to technology, we bring technology to the people.

The local co-ops, social entrepreneurs, charities, and activists just gather in this kind of facilitated conversation. I’m the main the facilitator. People in this kind of setting they have to speak about public benefit.

If it’s one-to-one, it’s often about private benefit. Because it’s radically transparent, people talk about things in public benefit ways.

The different municipalities, Taipei, Taichung, Kaohsiung, Taoyuan, and now Hualien and Taitung have all joined through telepresence. Whenever anybody’s talking about a local issue, all the municipalities learns about it in real time.

In Mandarin, we say 見面三分情, you get 30 percent of trust just by meeting face-to-face. Through high-bandwidth connectivity, we get maybe 20 percent of trust versus face-to-face.

A great thing is that whenever it’s cross-ministry, all the career public service don’t have to copy each other on A4 papers. They can be just sitting next to each other.

Instead of saying, "We’ll have to copy the Ministry of Interior on that," minister of interior is literally sitting next to them. They have to really brainstorm something out in order to respond to the people’s emergent real social needs.

The greatest thing about radical transparency is that they get all the credit if things get resolved. If they have names on the public transcript, when the journalist interviews me, I make sure that they are fairly represented, that I actually let the journalist know who is the actual person that came out with the social innovation so they get all the credit.

If things go wrong, if people start flipping tables -- it’s not actually very easy, because you’re across screens -- but if there are controversies, I absorb all the risk. In this kind of environment, the public service become very innovative.

We introduce such new ideas in exhibition cities like that Shalun Smart Energy Science City just outside of Hainan High-Speed Rail Station, which is also part of the Taiwan autonomous car lab.

Finally, just one last example, whenever there’s thousands of people caring very much about an emergent social issue, we use AI-powered conversation to review the fact that they actually have much more in common about this than we previously thought.

We publish all the open data, which in Taiwan means open government, but also social, private-sector data, and we allocate three or more weeks for people to reflect, to share their feelings. That’s what often went missing in previous generations of democratic social designs, is that there’s no sufficient time allocated for feelings around the same facts.

I may feel happy. She may feel angry. It’s all OK. It’s only after the feelings resonate with each other that we ask about brainstorming ideas. The best ideas are the ones that take care of most people’s feelings that we can turn into discussion.

In practice, it looks like this. You see an avatar that represents you. You see fellow citizens’ statements. You can agree or you can disagree. As you agree or disagree, your avatar moves toward the cluster of people sharing the same feeling with you.

There is no reply button, because we discovered that if there is a reply button, people with the most time wins the argument automatically, because they have more time to troll [laughs] people. If you take away the reply button, there’s no way for a troll to grow, and instead of just trolling each other, people start sharing their authentic feelings for other people to resonate.

I will end on this slide, because this is the most important slide. Every time that we run such an open conversation, people discover that what the mainstream media and the social media shows is just 5 divisive statements out of 200 that we would have thought that it is what drives us apart. Actually, most people agree with most of their neighbors on most of the issues most of the time.

That is often forgotten by the popular discourse, and each conversation then drives those rough consensuses that we agree would bind ourselves to hold as agenda, meaning that we talk face-to-face on these and only these consensus statements.

That’s how we work to regulate issues that requires a people’s common understanding of the norms, and we just trend those norms into implementable ideas. We have run this with many partnering organizations, such as Columbia University in Bowling Greens USA, Alternativet in Denmark, in the UK, in Madrid, in many other municipalities and cities.

They invariably all report the same, that people thought they were divided, but they were actually not. After each process such as this, people’s trust with each other, as well, as the public service trust to the people, increases.

That is the main thing that we’re designing toward. It is a way for the government to trust the people. People sometimes trust back. Sometimes they don’t, but people grow to trust each other more.

That concludes my seven-minute opening. Feel free to ask any questions.

Just from the background that I read up on you is that you had a very rocket-ship career in tech. You were in Silicon Valley, and I assume that you could have stayed in the private sector and amassed a good fortune, and worked out your career and your change that way.

My question is what really compelled you to leave that private sector to move into government? What motivated you? Why do you care enough to work in the government?

I’m not really working in or for the government. I’m working with the government. I’m kind of at a Lagrange point between the movement and government.

That’s still a very good question. I actually wouldn’t consider my career particularly fast. It’s just I started early. My first startup was when I was 15 years old, in 1996. Like everybody else, it took me 20 years or something like that to reach retirement.

That is just by the virtue of our junior high school teachers at that time. I just showed my teachers my conversations in email with this, at that point, very new website called Arxiv.org. It’s still there. Now there is also SocArxiv. There is all sorts of archives.

It turns out that the researchers there who published the preprint doesn’t care whether I’m 14 years or I’m not. As long as I can write an email, people just considered me researching partner.

I just convinced my teachers and the principal, saying, "You’re telling me that I need to work 10 years just to get to that lab, but look at this. I’m already exchanging with the director of that lab and we are figuring out research proposals together."

They were very reasonable people and they said, "OK, from tomorrow on, you don’t have to go to school anymore." That instills in me an optimism about bureaucratic flexibility in a very tender age.

Then I discovered Internet governance. There was a way, five years before I even had the right to vote, for everybody to participate in the shaping of the Internet, the Internet Society and so on. For me, democracy is always that.

Five years later when I was 20 years old, I started learning an outdated retrogression of democracy that is representative democracy. I always thought that in Taiwan we got the World Web and the presidential election more or less on the same year.

Democracy here in Taiwan is just like any social technology, and as a technologist, I’m happy to just adapt and tinker the technology in a way that makes the fit better with the society.

In a sense, I didn’t leave the startup scene. This is maybe my seventh startup as an entrepreneur. I’m just leading this theory of change within the government, but having this kind of lab space and organized pretty much exactly like a startup.

The retirement "from the private sector" only means that I would like to maximize the social impact of my work, but I’m still working very much with the Silicon Valley and with the other side, for example, New York City. I just joined the board of directors of a new NPL called RadicalxChange with Neal, Ellen, and Glen Weyl, and Vitalik Buterin of Ethereum, and so on.

We’re spreading these new democratic technologies everywhere in the world. Meanwhile, being a minister. This is...I think the word is slushy. The idea is that I’m still working as entrepreneur, it’s just not for personal profit but, rather, social impact.

I wanted to follow up on your radical transparency, the way that people respond to it. It’s clear in the way that you’ve set up your office that you’ve worked really hard to be transparent. You’ve put information online.

I wonder how much do people go out of their way to access that information? Do you see a lot of Taiwanese people following up on these platforms?

Yeah, totally. We do so because it’s not just the text to a transcript. People have different modes of communication. Some people prefer text, and that’s fine, but we also publish small snippets of video. We also publish mind maps. We also publish all the different modalities to make sure that people understand the context of our everyday policymaking.

I also write columns biweekly on "Apple Daily" here, and also monthly on "Business Weekly." All this is just a way to translate this raw material, but opening up to any other journalist as well. You can also see investigative journalists taking data as raw material and then package their own narratives.

There’s even people drawing manga comics based on these materials. This is the raw or the source material, but people who usually access it out of interest are the people who are actually cultural translators, that are making sense for their communities using the same raw material.

This works much better than the previous norm, where after a meeting, three people get interviewed by three different media in these three different meetings. It makes sure that people don’t quote each other out of context. Indeed, the 10-day editing period is very fruitful for the people who do attend as well.

Sometimes, they run to me saying, "Minister, I finally understood the other side’s argument." When people are on this kind of meeting, sometime they just speak instead of listen. This kind of coediting also makes sure that people go through each other’s argument, and then makes it much better as a listening device, rather than a speaking device. This has multiple uses.

Do you mind going back to that slide you had last on your presentation? I was super fascinated by this product. Quick question behind how this works. Is this an online website?

It is.

If you’re a citizen of Taiwan, you go online, you can pose one of these statements?

That’s right.

The first part, you define a divisive statement versus a consensus statement. How do you define which is which?

This is by k-means clustering and principal component analysis. No human do that. There’s an open algorithm. Right now, if you Google for digital dialogue AIT, you will see one conversation going on about how to promote US/Taiwan economic and commercial ties. That is proposed by AIT, the de facto embassy of America in Taiwan. You can see how it looks like.

People currently are grouped into two groups. Group B believe that US and Taiwan should start talking about FTA right now, and Group A, only half of them actually have opinion on that. Half of them disagree. Despite the different emphasis on priority, actually, people do agree on widely a few things.

This is the user interface. This is not the analysis. This is what people see after answering yes or no to resonate or not on each other’s statements. They can see their avatar going toward the cluster of people that share their views.

We have an actual conversation that was just before this one. This one is still ongoing. We had one that asked people how to promote Taiwan’s role in a global community. At the end of the two-month conversation, if you click group B, you will find the most divisive statement.

For example, that every time the PRC closes an international door, the US should try open one for Taiwan someplace else. Half of people really like the idea. The other half really hates the idea.

If it is a mainstream conversation, it maybe just ends there, and people will just amplify the controversy, because there is no reply button, we actually let people see there’s only one very divisive statement, and that’ the one I just showed you.

There’s a few that has less than 50 percent consensus, but otherwise, everybody agrees on pretty much everything. Then we use that as agenda to ask the AIT and Ministry of Foreign Affairs and so on, to respond to those 10 consensus, and only those 10.

Awesome. It’s such a cool mechanism. Do you start with a prompt and start a conversation, and then from that readout, people start to make these statements, and then people vote on those statements to understand what they align with?

Yeah, exactly. It’s a way for the crowd to collectively reflect not only on the opinions, but also have a reflection of their crowd’s shape in real time.

Who is the...It goes out publicly? Are you...

I write a column to introduce it.

You write a column to introduce it. How big was your correspondence, conversation?

It depends on the subject. The one that we had the first time, that was in 2015, around Uber and their use of people without professional license to drive for profit, that was the prompt. That’s the actual shape, I think, two weeks in. At the end of it, I think it had...just a second.

At the moment we’re having one, actually right now, about electric personal assistive mobility devices, such as Segways.

The consultation will actually be covered by BBC and it’s this evening, and it’s again driven by Polis. We can look at some prior conversations. Airbnb, UberX, cyberbullying, and things like that.

The UberX one I think had -- that was back in 2015 -- easily thousands of people who have voted nine or more times, to be visualized. They do have a lot more in common than people previously thought.

Everybody cares about insurance, cares about the insurance for passengers as well, that we need to keep the conversation going instead of shutting down Uber.

That safety is paramount, that fairness, if they run an e-rental, we should tax them like a rental, but if they run an e-taxi, they should be taxed or incentivized like a taxi. Everything like this, basically.

The government should then have a jurisdiction about ensuring economy communication platform, and said that the government should have some say after all.

This actually was very splintered. At the very beginning, you can see the four groups literally in the corner, and then they slowly converge into the consensus items.

I think the most participation is maybe in the thousands of people, the less interesting ones like the Segway, maybe that’s hundreds or, at most, 1,000 participants.

I have a question related to sample size.

It’s not a sample... Everybody speaks for themselves.

What I mean by sample size is that if you’re using this to help inform policy dialogue...

Just the agenda; none of this is decisional.

In theory, policymakers can still consume this information as you choose to present it to them. Even if it’s helping set the agenda, it could be the case that, particularly if I’m a policymaker with not a lot of time, and I don’t bother spending that much energy reading up or understanding the nuances.

I could look at the data you present me and think even if the sample size is only 500, I won’t pay a lot of attention to that. A large majority of those 500 agree or disagree with something, that I might be inclined to pursue a certain policy agenda that might not be representative.

That’s why all we do is that we present this as people’s feelings instead of anything concrete for ideation. This distinction is very important, because if you look at this picture, group A is just 40 people, but yet they have an area, a size that is almost the same as group A.

That is to say, it doesn’t matter how many people vote exactly the same. It’s just going to be a dot here. Usually, when we present, we just take off the numbers, because this is measuring the diversity in the feelings landscape.

This is not measuring the statistical representativeness. This is nothing like a referendum, basically. What we are measuring in design thinking terms is just how divergent and a little bit convergent, maybe to this degree.

What kind of insights and themes can we sense from this collective sense-making exercise? Nothing less and nothing more, so we always make sure to say that the face-to-face conversation where we help each other to find a common value among different positions is the only result.

The iteration, evaluation, ideation, everything in the second diamond is off the table for this process. This augments to the prior, rather than replaces to the middle of the existing representative democracy, theoretically speaking.

Hello. Thank you so much for your time. It seems like a lot of your work focuses on this idea of ultra-transparency to the public amongst your coworkers, amongst your team.

Do you feel that you’re putting yourself in a position of risk by having such transparency to the public and others when you’re in a position to be a decision maker, having to make tough or at times divisive or controversial decisions, or has it been relatively risk-free? If there is a risk, do you find that risk worthy to have this ultra-transparency?

The risk is entirely perceptual. Sometimes, the public service doesn’t want to make the drafting stage -- as we call it -- transparent, out of habit, not out of any careful analysis. Indeed, in most of the freedom of information acts worldwide, the administrative drafting stage is one case is specifically closed to publication.

If you write an FOI request, anything that’s already decided, then, of course, it says you’re entitled to have a copy, but if you ask, "What about the other five alternative that was considered and then ultimately not implemented," no matter how open your jurisdiction, it’s likely that you only get a sketch, if any, of those discarded options.

That is out of habit than any careful consideration. My work mostly is existential proof that if you open up the why of policymaking not just the what, it leads to less risk for everybody involved.

The social sector will advocate for the ideas that ultimately not implemented by the public sector, can understand what we thought about them, and that we agree on most of the common values. Maybe we lack the resource politically or whatever to deliver those.

Then the social sector might take those idea and implement by themselves. If we don’t publish the ideas that were discarded then they will think it’s out of malice, or ignorance, or whatever while it’s nothing of that sort.

After two and a half years, I can say with confidence that publishing the why of policymaking lead to less political risk.

One follow-up. You also mentioned earlier on that you pull from each of the vertical ministries.

Yes. That’s right.

Do you find that the ones that are not sending over representatives are the ones that do not believe in this model and find it to be a risky model? How legitimate do you think their concerns are, such as you invited the Ministry of Defense?

The Ministry of Defense. That’s right. The Ministry of Defense in our constitutional design reports to the president and also the National Security Council. It’s understandable that we, in the premier’s office doesn’t have a jurisdiction over the Ministry of Defense, the Council of PRC affairs -- perhaps I should translate that as Continental China Affairs -- as well as the Ministry of Foreign Affairs.

These three are considered president’s team. On the other hand, the MOFA, the Foreign Affairs, after witnessing our work for a year, started sending people over. There are at least 10 people lining up to join my office.

We don’t agree for them to join at the same time, otherwise, it will become a MOFA branch. [laughs] The reason why MOFA is suddenly interested in that is that there’s a lot to say about public diplomacy as the one conversation about AIT already had.

This gives a different kind of legitimacy in international relationships. It doesn’t have to be bilateral or multilateral. It could be multistakeholder.

That’s something that they’re very much willing to align. That is why after demonstrating this conversation and more, there’s no ministry that says, "OK, we will never send people over again." Rather, they slowly, culturally...What’s the word? Osmosis. they are immersing themself into the culture.

The people who are not yet sending people over other than those three that reports to the president are seriously considering it. I would say that it’s still a forward way. My three working condition, the first one is voluntary association. I won’t give or take orders.

It takes time for the culture to disperse. The other two is, location independence, wherever I’m working I’m working and the idea of radical transparency as already explained.

At a time where you see that a lot of technological development has impacted a lot of social change across the globe, here in Taiwan where do you think are some of the clashes between that technological development and some of the more traditional mores and customs here?

What are some examples of, perhaps, areas from a culturally moral standpoint where you feel like technological development has had a spread in that wrongfully so? What are some areas in that same area of cultural and moral tradition, where technology and its development has helped clarify or helped the people in that way, if that question makes sense?

Sure, great question. In Taiwan, we don’t have a pre-Internet democracy. There is no traditional democracy to speak of. The first time we had a presidential election is already post-World Web. In Taiwan, democracy and Internet, they’re not two things. They’re always sold together.

That is like Estonia. It is a blank canvas for people to paint because we didn’t have hundreds of years of republican tradition or whatever to uphold. That’s the first answer.

The two examples, I would say for sure that the current, some people say disinformation crisis, is a great illustration of people feeling much closer to each other, but may live in different filter bubbles so that they have very different takes on reality.

That is something that’s amplified, at least, by technology and especially mobile computing that could threaten democracies. The ironic thing is that the more liberal your society is the more in danger you are. The people in highly authoritarian societies already have a way to take away people’s freedom of expression. It’s less likely for those things to go viral on the Internet.

In Taiwan, we’re the only one in this region that doesn’t have anything to take away people’s freedom expression. The ministers like me certainly doesn’t enjoy more freedom of expression as compared to any civic media. That’s very rare in our region of the world. Because of this, we have a legal definition of disinformation.

Our entire counter-disinformation design rests on this definition. We say disinformation is information that is intentionally harmful and untrue.

All three need to check. Harmful means that it harms the public and the democratic institution, not any particular minster’s image -- that’ just good journalism.

What we found is that the most important way to counter that is using exactly the same technology. That is to say if it’s a rumor that spread online, or on Facebook, or on PTT, or whatever, the clarifications are posted to the same channels and always within an hour.

It’s always within 20 characters the title, the payload always less than 200 characters means that it fits one single screen, even the mobile phone screen and always contains two pictures. This is a classic example.

A couple months ago, there was a popular disinformation that says if you perm your hair twice a week, you will be subject to a MT one-million-dollar fine. 60 minutes after this circulates, you see this picture, this memetic engineering product being posted by our premier’s office.

The photo you see here is the premier when he was young. It says, "This rumor is not true. I may be bald now but I will not punish people with hair. Premier Su." The fine print reads, "What we’ve introduced is a labeling requirement for hair products and it only take effect on 2021."

I didn’t translate this part. This is the now-bald premier saying, "However, if you keep perming your hair multiple times a week, it will damage your hair and you may end up looking like me."

This is good humor. He makes fun of himself. It went viral, much more viral than the disinformation. It serves as an inoculation.

People still see the disinformation but they see it after this spreads. People become inoculated. Instead of taking down anything, we’re making sure that people get inoculated by making our clarification messages fun and that they can spread.

This is one of the three ways. The other two being, we ask people to flag, like flagging email as spam, to flag messages as disinformation.

The social sector, the fact-check center, can fact check and classify something as false. Once they do that, it’s not disappearing but rather Facebook agreed to make it not preferred to show on people’s newsfeed.

Unless you only have one friend then they have to show it. If they have two friends, then it’s the other friend’s things on your posts wall. You can also go back to that rumor and you’ll still see it.

It’s just that Facebook will say, "A related link was to the fact-checking center." Not unlike how people deal with spam and junk mails.

The final thing is that during election we ban precision targeting sponsored by nonvoting noncitizens. That is treating them as campaign donations, basically.

The three lines of defense is our countermeasure against one of the most destabilizing technology, that is to say, social media, on democracy. The same thing can also be used in a collaborative way.

Usually, people who care about air pollution in other jurisdictions have to protest or campaign for the government to set up more measurement stations closer to their homes. In Taiwan, all the 2,000 or so airboxes that you see here are entirely from the social sector.

The citizens set up those airboxes. Each one is the same $100 US and sometimes even in schools. People can see, at one glance, the digital divide of Taiwan and also the air quality in Taiwan, the trend.

What the government does is that instead of beating them we always join them by saying, "OK, are there places where you think you should have measurement devices and you would love to but you somehow cannot get into?"

They say, "Oh, in industrial parks." They are of the private sector’s lot, they cannot go in even if they suspect the industries there is polluting the air. We’re like, "Oh, we own the land in industrial parks so we can hook air boxes there." People say, "We want a airbox here to tell domestic versus across-the-street air pollution."

It’s very hard. If you fly drones, they run out of battery really quickly. It turns out we’re setting up renewable wind plants, wind turbines, there.

We can change the contract and say, "Whomever wins the contract must carry airboxes." We’re in complementary relationship with the civil society. It’s all open source and open hardware. People across the world download it.

We make sure that all of this is featured in the data collaboratives relationship so that even people who are very young, like K-12 students, can all have broadband access to one of the top 20 supercomputer in the world that already has these crowdsource data and that already use distributed ledgers, otherwise known as blockchain, to make sure nobody can change previously shared data.

K-to-12 people can always start programming AI and machine-learning models to make sense of the air quality and the airflows nearby. This is the way to truly teach the children data stewardship that is something that is impossible to teach unless you are, yourself, a data operator.

That configuration was the same technology as the one that caused social division is then repurposed to use to have solidarity and collaborative relationship across different sectors and academia.

I I’m going to follow up on that.

Sure.

Did you have any issues with building the credibility of your fact-checking site? How did you deal with that?

The international fact-checking network, which is entirely social sector and not government-sponsored, have a way of saying, first, that they only accept micro-donations like crowdfunding. It doesn’t accept any money from any party or any political entity. The funding source is transparent.

Then they always include the entire fact-finding process in their so-called fact-checking report. This is much like, for example, the latest one, if you see a child carrying a purse that is orange, it means that they live with deafness and, please, slow down. It turns out it’s not true. [laughs]

This is something that a NPO in continental China have promoted as a campaign. It’s not part of Taiwan. Instead of saying, "It’s not true," they interview all the relevant stakeholders and also publish the entire log of their fact-checking with those stakeholders so that everybody can draw your own conclusion after looking at the due diligence and fact-checking.

Sometimes it’s quite ridiculous -- like if you look at this picture and see which four digits then you have nearsightedness, or farsightedness, or something like that. It went viral anyway. Anything went viral is subject for fact-checking. They would say something about spatial frequency about contrast sensitivity function.

They interviewed a professor from NTU department of psychology, as well as eye doctors. Everybody builds context based on their publication. It’s not a blind faith or blind trust into the capability of them as journalists, but rather the source and the public transcript of their conversation with the source that understand the legitimacy.

This is all supported by the social sector. The public sector, the only commitment we bring to the table is that if the fact checker asks us about anything, we’re committed to give out a layperson-understandable narrative within 60 minutes.

That’s the only thing we do. Anybody else can do that as well. We don’t monopolize this way of legitimacy.

Before you became digital minister and set up this format and this way of operating in your office, what existed within the Taiwan’s government to address any of these topics?

I was one of the people who supported the Sunflower Occupy back in 2014. In March 2014, the main contention was that the Cross-Strait Service Trade Agreement or CSSTA was considered by the administration as a regulation level, not a law-level conversation so the MPs cannot have a substantial deliberation about that.

The ministry should have the entire authority in making it happen. Theoretically, the ministry need to pre-announce their regulations. At that time, it’s usually for 7 days or for 14 days.

All it contains is a telephone number or, at most, an email that you can write to during those one or two weeks. Everybody who called have no idea, already 500 people or 5 people have called before them.

They have no idea that what are the major arguments or consensus around them. It’s not very effective. That’s an understatement. It’s not effective at all to reassure the people that after those one or two weeks of so-called public commentary, that the administration is taking any of the ideas into account. That creates what we call a legitimacy crisis.

Because of that and of the Occupy, after the Occupy demonstrated that, even with half a million people on the street and 20 or so different NGOs each deliberating on one particular aspect, we can, nevertheless, agree on a set of five consensus items after three weeks of Occupy.

The ability for people to form consensus is demonstrated to the entire country. Then after that, there was a national forum.

Many people said, "Let this become the new norm. All regulations need to be up for 60 days of deliberation. All the commentary need to be made public and with no exceptions and that applies even for projects of a national any ministry-level projects that runs for one year or more. The KPIs, the budgets, and everything also need to be publicly discussed."

This is the actual interface. You can see all the different 32 vertical ministries, what their projects are, the 2,000ish projects, and how many people care about.

For example, the most cared about is long-term healthcare, and then sanitation, and then social housing. Then, for healthcare, you can see that it’s 10-year project and they were on the third.

Then, how is the KPI procurement and the budget going? Every quarter, the minister of health and Welfare go back and say, "Thank you for your commentaries and feedback. This is what we have changed this quarter thanks to your input."

This is radically, not just transparent but accountable. This is what people already demanded in 2014. My work as well as my previous work as understudy to another minister, Jaclyn Tsai, is to realize this common understanding and vision of the 2014 Occupy.

You had mentioned earlier about the government trust in people and how this has facilitated that and improved that. It reminded me about, at least some of the situation that you see in mainland China where there is a question about whether or not the government does trust its people, its much larger population of people.

How do you think governmental trust in the people exists in a purer form in Taiwan in comparison to China? Do you think that democracy has anything to do with that difference?

Yes, we are on the frontline of global confrontation with authoritarianism. That’s one of the consensus points in the digital dialogue with the AIT. One can contrast very easily that with our freedom of speech and expression, not even democracy.

Our commitment to freedom as core value, that takes the same technology but we apply in very different ways. Whereas in Taiwan, you just saw the joint join.gov.tw e-participation platform that makes the state radically transparent to the people, we do see that on the Chinese continent, they are introducing social credit that makes the citizen radically transparent to the state.

It’s more or less very similar visualization, actually. We use the same color palette of green and red and things like that, but it turns out that the direction is completely opposing.

In social credit, if somebody loses such a credit, there is no due process at all. There is no way to appeal. It’s entirely managed by a close algorithm, whereas here, it’s entirely open. Just open versus closeness, I would argue, is more fundamental than democratic/authoritarianism. Taiwan also had authoritarian past.

The vertical ministries are sometimes quite authoritarian in their ministry or management protocols. As far as they agree to become more and more open, in the sense of being more accountable to the public, then they will inevitably adopt technology in a way that foster this kind of change, and also with the private sector.

The previous slide, I talk with the social sector. With the private sector, I talk about that, how people with new ideas, like self-driving vehicles, that our law doesn’t address, is given one year to experiment in the public, and for the social innovation to drive regulatory innovation.

On the Chinese continent, the Communist Party have set up a party branch within every enterprises above a certain size, so they can ensure that the party’s agenda become the private sector’s agenda, instead of being set by the private sector innovations.

The "party innovations" become the private sector’s "agenda." It’s the same idea about working in opposite directions. I would say that a lot of those technologies are implemented in full force and very quickly, but it’s on opposing directions, is my observation.

This is a bit of a pivot from Taiwan, but looking more at your expertise in implementing new technologies in governance. In a project that I’ve had the opportunity to begin to work in my place, my country of Indonesia, focuses on trying to bring these kind of innovations to ASEAN.

Where the conversation, I think policymakers want it to be elevated to this level. The reality is that digitization of electronic records is the goal, let alone getting people to go online and take a poll, and having AI then be implemented to come to these conclusions and outcomes to inform policymakers.

How would you suggest or propose a path for these countries to take a lead from Taiwan in order to get to this stage from their current point of development?

I would say that all the efficiency or effectiveness arguments that we basically just talk about it doesn’t really make sense if you have a governance apparatus that cares more about reducing risk.

Usually, when talking with our partners in this region -- I visited Bangkok very recently -- I talk with the people there. I always emphasize not efficacy, not sharing credit was the two main points that I made. I always say this is a way for the mechanism to absorb risk for you. I bring out a very different set of slides. [laughs]

Instead of saying, "This is a public commentary," which is way beyond their current situation, we usually bring up the case of the National Audit Office.

The National Audit Office, before we introduced this e-participation forum, is in the rather unenviable position of having to say yes or no for each governance innovation, because they’ll have to establish new auditing principles for them. For example, I would use one concrete example.

There’s somebody from the Taipei City that has the great idea of setting up their hall for social entrepreneurship in such a way that instead of employing X number of disadvantaged people with handicap to work on specific products with specific KPIs, they’re renting the space to any qualifying social enterprise to run their own projects and employ as many disadvantaged people as possible.

They may fail, but if they fail then of course they invite somebody else in. They had waves of people participating from doing business model, like building modeling beam or AI tagging or cooking or whatever that they think is good for disadvantaged people to do.

The National Audit office, if we step ourselves into their shoes, is in a very unenviable position because the disadvantaged people, the campaigners, the activists have been clamoring this for decades. If you say no, you’re the villain, basically, that stops progress from happening.

If you say yes and you audit them with the old auditing mechanisms, you may not look good after a year. Then the members of parliament will lambaste the National Audit Office for squandering taxpayers’ money on something that is uncontrollable and, indeed, may be unaccountable. No matter what they do, they’re doomed.

Our position when talking with the National Audit Office is that instead of asking for people’s feelings, you can assume people feel fear, uncertainty, and doubt. If they’re very positive, then you’re preaching to the choir anyway.

They frame their conversation as, "What’s your fear, uncertainty, and doubt concerning renting at zero dollar of social entrepreneurs for that SE building in Taipei?" Sometimes when discussing budget, you can see that most have maybe 20 or 30 conversations every quarter, and we consider a success.

When they ask for fear, uncertainty, and doubt, they have at least hundreds [laughs] in conversations. It’s always easier to express distrust [laughs] than expressing concrete suggestions. To express concrete suggestion, even feelings, you have to grow through the actual plan. If you’re just saying, "Oh, I really doubt that this will happen," that can be written in no time. [laughs]

They always get much more input than we in the administration do. Once they do that, they sort them into nine categories, nine clusters of fear, uncertainty, and doubt.

They ask the minister in charge of social innovation, that’s me, to say, "Here are people’s doubts and distrusts. What do you say about them that can reassure them?"

I’ll have to ask my Scottish counterpart, my British counterpart, my Italian counterpart who have implement something like this before of how they managed to work with the people who doubt those ideas, because it’s the first time in Taiwan.

I took a couple week to write a detailed response to them. Then they explained this in laypeople’s terms and responded to everybody and say, "Hey, Minister Tang have outlined the new auditing criteria," that, "We are going to audit exactly the way that Minister Tang have suggested."

They win no matter what. They address people’s fears. People trust them more. If this new auditing strategy doesn’t work, it’s me to blame for the MPs. They are in a no-lose position just by virtue of having a public conversation about fear, uncertainty, and doubt.

For many people in this region, this is the actual selling point. The career public service are all reformist at heart. It’s just they don’t really want to pay the political risk of introducing reforms. If they’re just shuffling those fears, doubts, and distrusts around, that totally they can do. It’s a way to lower the risk for everybody in the career public service.

I would suggest that you introduce this concept first, and then, without sacrificing effectiveness and due credit, basically make it about Pareto improvement. You make the risk management better without sacrificing the other two.

Then you will have a political moment to increase the other two without sacrificing any of the other two. Basically, small increasing returns will soon kick in. Sooner than later, it will become a norm that people feel, "Oh, it’s just natural to do this."

My question is also about feelings. I thought the example you showed us, correcting the misinformation about hair perming, was really interesting.

It seems to me, in that case, although that misinformation was widely disseminated, and then you were able to correct it and inoculate people against it, I would imagine that ad labeling hair products is not a particularly ideologically divisive issue.

I’m curious with respect to issues where misinformation is being shared in connection with some very deeply held emotional or ideological belief, how you deal with the fact that some psychological research suggests that presenting people with contradictory information can actually cause them to double down on emotionally held beliefs.

How do you deal with that when you’re trying to correct misinformation?

That’s why we always say we’re clarifying instead of correcting. If you present the information as if you cause the viewer to react with a bad feeling of being corrected, it doesn’t work, exactly as the study you cited shows.

Other memetic engineering that our ministers do with those hour is about appealing to people’s sense of humor, appealing to people’s sense of...Here we call it kūsō (くうそう), which is hard to translate but somewhere between a parody and a satire.

We always phrase or frame our clarification message as not a confrontation to the rumor. Rather, it’s a funny take, a funny interpretation, almost onion-like, except the content is true. We learned a lot about the packaging of disinformation, but send the clarifications in that way.

We are not shying away from causing social controversies when making fun and when making humor, but it’s always about ourselves. It’s not at some other people’s expense. This is the main difference between a traditional fact-checking as in true, not true, or whatever.

Here, we always make sure that there’s two photos, at least one of which are usually a cat or a dog. That [laughs] just softens people’s feelings into more willing to consider our piece of the puzzle, so this attitude of presenting our piece of the puzzle, plus the strategic use of cats and cats’ picture, together means that people are much more receptive, even if it’s on politically charged misinformation and disinformation campaigns.

All the ministries’ social media experts have different personalities, but nowadays, if you go to our ministries’ outreach like Facebook, Line, or Twitter, they all tell you that they share one common hobby, which is 吸貓 -- how do I translate that? -- literally inhaling on cats. [laughs] Not marijuana, cats. [laughs]

Everybody is very adept in adapting animated gifs of cats in all their conversations. There’s another, I think it’s a MIT Media Lab research, that it’s even more viral than baby pictures. You can’t really beat cats [laughs] when it comes to virality of the message. That framing is the main difference versus traditional fact-checking.

Two questions, and I’m going to ask the first hopefully easy to answer and the second one a little bit more elaborate.

Fill in the blank. You have mentioned two times now "Continental China," therefore Taiwan is...what China?

Taiwan is just Taiwan. Well, I guess you could also say “Free China”...

I was just wondering what’s the corollary because I hadn’t heard it.

This is from our President’s recent speech in the US where she referred to Taiwan as "an island off the Chinese continent."

Constitutionally, the Amendment of the ROC Constitution calls the PRC-governed territory 大陸地區. Many people translate that as "Mainland China," but that kind of implies that they’re the mainstream and we’re kind of on the fringe. [laughs] That gives the wrong picture.

Now we could also say "the Taiwan islands" and "the Chinese continent."

Other question, you talk about transparency and accountability. In principle, hard to argue with that philosophy, in practice, how are the 32 ministers dealing with it?

They don’t have to join if they don’t want.

They don’t?

I would say a vast majority have accepted that this is the new norm. The others at least consented to have a team of what we call participation officers within that ministry, all reporting to the minister, in charge of transparency. It’s just they’re not yet sending people to my office in a radically horizontal way.

I would say that nobody is against that. In the DPP primary, as was in the KMT primary, all the presidential hopefuls just doubles down on this kind of open government. If they have anything to complain about this approach, they’re saying that it’s not doing enough.

We need to be even more radically transparent and participatory. We’re in a very fortunate political landscape where this is considered a norm of people since 2014.

My question is related to the point you made about the direction of transparency from the government to the people, from the people to the government. You put the Social Credit Score system as an opposite of that same spectrum.

My placement for my year on the Luce program was in India, where Aadhaar and digital identity number’s been a really big conversation. I was curious what your take on that was. On one hand, it seems to be a solution that when you reach a scale such as India, the government does need some information about the people to serve them in any kind of capacity.

It seems like some degree of transparency from the people to government’s necessary. It’s just a question of how much. I was curious what your take on Aadhaar was, if you’re familiar or you have some commentary on that. More broadly, as you scale up to a billion-plus-person country, how do you start to think about information a government uses to serve the people adequately?

Aadhaar, Estonian eID, as well as other designs that we’re seeing -- maybe not Libra because Facebook have not yet declared independence [laughs] -- all the nation-states, we call it identity management platforms, are the same expression of the fiduciary duty that the government has on people’s data entrusted to the ministries or to the state.

Traditionally, what we would argue is that the more anyone feels that a fiduciary body is acting in their best interest, the more likely that someone would voluntarily disclose more personal information. You see it in our relationships with our accountants, with our doctors, with our psychiatrist, with our lawyers.

These are the professions that really need your private data. They are paid to act on your best interests, sometimes paid by the state, but still acting on your best interest in the case that they’re defending lawyer.

The more that people view data as the beginning of a relationship and the less that people view data as an asset, like oil...It’s the worst analogy ever.

The analogy was invented when both were hard to extract. It used to be very hard to extract insight from data, but that was before machine learning. Otherwise, oil and data has nothing in common at all. That’s a really bad analogy.

The more people see data as beginning of a relationship -- GDPR is going to that direction -- the more likely the citizens will voluntarily give data to a data trust. The other trust may be a social enterprise, may be a co-op. It may be a nationally mandated independent unit. It may be a DPA, anything.

The point is that they have to continuously earn the trust. Once you have that healthy relationship, then it can scale to more people as well as to more regions.

If people don’t take this long route to earn trust, and take the shortcut of forcing the people to trust "blindly," then it actually creates a situation where any breach in trust -- it could be server security, it could be anything, it could be a social engineer, it could be a domain hijacker, whatever -- then the people by default doesn’t trust the government who was acting in people’s best interest.

Even if they were, they wouldn’t be trusted any more. Then it becomes progressively harder to introduce any new identity system.

In Taiwan, we are by and large taking the long route of having each ministry being their own data protection authority, of having no central database of anything, of having the National Development Council to interpret and introduce the latest technologies without appointing any CIOs in any ministry or agency.

Of having the municipalities not having to syndicate their database to the central government, so everybody must prove that they are acting in your best interest in order to entrust people’s data trust into that particular governance system.

This is sometimes criticized as slow, inefficient, too much decentralization, or whatever, but in Taiwan, we believe that this is the only way forward for the government to show trust to people, that we are acting in people’s best interest, and be called out if we are not.

I’m really curious about if we can give up the government as an organization, how your ministry’s work fits into that and from a managing perspective almost.

I’m first curious is the end goal of your ministry to work yourself into irrelevance, and otherwise to build the capacity to do this kind of work across the vertical ministries, rather than have this long-lasting horizontal function.

I’m also curious more in individual level, how your staff have...how they respond when they first join your team, how capable and prepared are they with the skillsets and mindsets necessary to do this pretty unique kind of work, and how much training and handholding do you have to do early on.

Very good questions. Digital transformation is a phase. It’s just a decade or two. By year 2030, maybe there will be no digital ministers. Exactly as you said, people would have finished digital transformation. Maybe we have one analog minister and everybody else is digital.

When digital is no longer an emerging issue, then, of course, we don’t need a horizontal minister for that. In that sense, I’m designing myself out. You’re completely correct.

On the other hand, horizontal ministers are in charge of new subjects across ministries, there’s constantly new subjects in social innovation. It may not be digital at that time. It could be, I don’t know, quantum. [laughs]

As long as there are things that are emergent that is outside original designated function of any ministry, there’s a reason for horizontal ministers to continuously exist. That’s from a theoretical point of view.

Whereas nowadays it’s open government, youth engagement, and social innovation, we’ll actually see that my mandate has already changed in the past year. It’s no longer just social enterprise. It’s social innovation. It’s no longer youth council. It’s about youth engagement and so on.

All this is subject to the different ministries’ capacity building, that once they are done capacity building, then they don’t really need me anymore, and that’s for the best.

The second thing is the delegates sending to my office. I only have two HR criteria. First is that the person dispatched here needs to have at least one competency or skill that is above the existing members.

It could be anything. It could be making coffee. I’m not joking. [laughs] It could making visualizations, illustrations, real-time interpretation. It could be anything. They have something to teach.

The second is that they must be at least as much a giver as they are a taker. It’s OK to build your own personal career, but one must stand ready to help each other in their careers as well. That’s really the only two things that I’m looking for.

Everything else is a learning organization. We don’t really have a curriculum, so to speak. Instead, every month, we have at least two collaboration meetings among people’s e-petitions. For example, whenever people start a new petition about how -- there was two years ago -- the tax filing experience is explosively "hostile," we need to vote on this matter.

I promised to give due credit, so 楊金亨 here, at the time the Ministry of Finance participation officer, just look at this e-petition.

Even before it reaches 5,000 people of the mandatory response threshold, he just posted publicly saying, "Thank you for all your criticism. Everybody who complained about Ministry of Finance is now cordially invited to the Ministry of Finance for a collaboration meeting to design next year’s tax filing experience."

It turns everybody who complain into chefs that participate in the kitchen, so to speak. We meet the presenter of the petition, and it turns out he is a professional interactive designer. He cares the most and so suffers the most.

We livestreamed the conversation, and he really helped mapping what we call the user journey of everybody receiving the tax filing advertisements to enter the tax filing system to finish filing their tax. What their actions, needs, problems, emotions, feelings, and solutions to make them feel better.

The greatest thing is that even thought at that time, there’s easily thousands of commentaries online, if you have this kind of structure, what we call service design structure, actually, people who post the same thing 500 times, it’s just one Post-it note on this graph. It lets people see the signal out of the noise.

That become on-the-job learning for everybody involved. The only thing that I need to remind my colleagues is that whenever there is people online saying that it’s explosively wordy, public service, sometimes just instinctively, when they’re writing the Post-it note, even without any awareness, writes this that maybe it’s a little bit verbose than we would have liked, to tone down.

The continent China calls it the harmonious transformation of online comments to make more harmony, but then you lose solidarity if you don’t report people’s comments as is. We need to reign in our instinct to make harmonious words.

If the tax filing interface is so baroque so that it confuses the hell out of people, we can take out all the exclamation marks, but the words need to stay here.

After that, then people who complain the most, that’s these people, and the people who they are most toxic toward, that’s these people, meet face-to-face in full workshops. It’s very difficult to spew the same toxic words once you’re co-designing.

After that, then they deliver collectively a new tax filing experience that has 96 percent approval rating last year. I heard that it’s 99 or so this year.

It’s really good, and the net budget for this is negative, because in one of the workshops, somebody -- an IT engineer -- observed that we’re paying the same CPU and bandwidth cost throughout the entire month of filing our personal tax, but actually, people file mostly on the first two days and the last two days.

All the bandwidth and CPU cost is squandered when we do the old rental. Now you can charge by hour or at least by days, quite elastic computing. Once we switch to that kind of payment, it saves a tremendous amount of budget, and we use a fraction of that to run the workshops.

This is a concrete response to you in the sense that people don’t really need an abstract case. Whenever they come here, every two weeks, there’s an actual case like this coming in, and people just take up whatever position needed to address that.

I was thinking about what we read about you earlier, how one of the first things that you designed or invented was something that would help your younger sibling with fractions.

Yeah, that’s right.

That came from the standpoint of providing a service that would help someone, that would help to improve a specific difficulty or problem that you saw.

It looked like this, then you see some balloons, and you can enter one half. Then you see something like this, and then you have to keep guessing, and then just build up the fractions. It’s on Scratch, actually. Nowadays, there are still primary school people using that for their fractions. Go ahead.

That’s amazing. As you’ve been talking, I was envisioning you like 20, 30 years from now. You’re older and you’ve probably had people that you’ve impacted and that you’ve trained, and like you said, maybe your position is even phased out, and you go on to another service.

My two questions that I really want is what is the importance of empathy for you in life? What does a successful career look like to you?

Empathy, for me, is the capability of taking all the sides. In a 20-sided conversation, if a facilitator can argue from each of the 20 sides authentically, then they have a much better chance of getting people to see past their differences and positions and build a common value together.

If one can only empathize with half of the sides, of the stakeholders, and just naturally consider the other stakeholders as enemies or as others, then they’re not a very good facilitator, and it’s almost guaranteed that you will get nothing very productive out of this meeting.

Being able to take all the sides is why I think empathy is very important in this kind of facilitative governance. A successful career is just something that you follow your own sense of intrinsic motivation.

When I joined the cabinet and worked with, not for, the government, the premier at the time, Lin Chuan asked, "Do you join the cabinet out of a sense of duty, a sense of social impact, of maximizing whatever purpose?" so on and so forth.

I’m like, “No, I joined the cabinet for fun. I’m just enjoying this experience of listening to people and taking all the sides.”

This keeps me motivated every day because, unlike passion which leads to burnout, fun is something that you can have every day.

Optimizing for fun is the main thing that let me continue to have very difficult conversations with people to feel strongly about one another’s positions. I drive this conversation because I derive enjoyment from getting to know people better. That is success.

It’s not something that requires anything external to certify or to assure. As long as you’re fulfilling your own intrinsic motivation and sharing that joy -- fun, when shared, is joy -- with people, then I think it’s success.

Throughout your presentation, I’ve been thinking a lot about American society, the political landscape. America is seen as this pillar of freedom of speech, and OpenNet, and everything, however, despite that, as an American citizen I’d say, at least for me personally, I’ve often been uncertain how to engage politically in these discussions.

I’ve seen conversations going on in the comment section and I feel disheartened by it. That’s not to say America shouldn’t be esteemed for certain things. I’m grateful that we have OpenNet and freedom of speech.

At the same time, I see the ways in which it’s become not quite efficient or ways in which it’s been exploited. We have all these issues.

I wonder for you personally, as you face another election, when you reflect on freedom of speech and the way social commentary and dialogue is being conducted in America, do you think about ways in which it can be improved? Do you look at what’s not working in the system or have any thoughts on that?

In the US as well as in the UK after the referendum...I just had a long conversation with people from BBC who is here for four days to cover my work and the democratic innovations from Taiwan.

People usually attribute this kind of disempowerment to the rise of Internet-enabled populism. Then people’s ideas are being reduced to soundbites or clip bites. [laughs]

It’s a much-reduced form of expression even if you have very nuanced idea, very resonating idea, those ideas are not given the space to surface as we’ve seen in Polis, in e-petition, in sandboxes here. I don’t think populism is the root of the problem. I think it’s populist tribalism.

If it’s populism that appeals to all the people, to everybody, populism that doesn’t exclude but rather include, then I don’t have a problem with populism. It’s like classical populism. People who feel excluded from political process can be encouraged to express their authentic selves and being listened to, enjoy the experience of being listened to.

The main problem is that people would imagine that there is half a population that’s not going to listen to my point no matter what the point is. That is the problem. I would encourage everybody in that kind of setting, it’s especially difficult in the months leading up to election, because then people become automatically polarized because of gerrymandering other things.

If there’s still hope, if you can talk about things of a longer planning horizon...The election rhetoric is usually about the next four years or five, but if you start talking about things that are in the next 10 years like, I don’t know, sustainable long goals but many other things. We invited, for example, two very different sides just before the marriage equality referendum here.

The 護家盟, Family Guardian Coalition, as well as 伴侶盟, Taiwan Alliance to Promote Civil Partnership Rights, they are both famous here in Taiwan.

We talked about the e-petition case. The question was about whether to allow reproductive rights, artificial insemination, for women who are single. That could be a very controversial topic to talk about.

We framed the discussion to say, how might we change in our society so that the newborns to such families can enjoy equal inclusivity in a society 10 years down the line? Once you reframe the conversation like that, these different so-called camps discovered they have much in common.

The stakeholders are literally not born yet. [laughs] We’re planning for the next generation. When preplanning horizon is 10 years, 20 years, people have a lot more in common than they would have thought. If we are talking about next quarter, then it’s a lost cause.

In addition to the traditional theory of overlapping consensuses of being very specific in your ideas and use cases, I would say also increase the planning horizon and then keep talking about these things. You would find that across party lines you will have more alliances than you previously thought.

This is going off of the previous question on the topic of populism and what are the driving factors there.

It also fits into the debunking of hoaxes. Internet-based political engagement has led to the rise of iPod politics, consuming the information that you clearly want to consume in your own ideological bubble, which is problematic when you’re trying to show folks that there is more in common than these divisive ideologies.

How have you seen your role as involved in this technological space ensuring that the population is being able to break out of their own ideological bubbles and not get purely trapped inside of them?

The point is not to focus on technology. If we’re talking in technologist terms like Internet of Things, or machine learning, virtual reality, or whatever, these are instruments where people project their own fears, uncertainty, and doubt in psychological projections on these tools.

I call myself a poetician because I don’t give or take orders. What I do is I write poetry. That’s my main work.

The first one as digital minister is when the HR here asked me for a job description. Taiwan didn’t have a digital minister before. They asked for a job description to put somewhere on the website to explain what the post is really about.

That’s the first time that I sat down and started seriously thinking how to not talk about technology in technologist terms. I’ll read you my job description. [laughs]

The job description goes like this, "When we see the Internet of Things, let’s make it an Internet of beings. When we see virtual reality, let’s make it a shared reality.

"When we see machine learning, let’s make it collaborative learning. When we see user experience, let’s make it about human experience. Whenever we hear that a singularity is near, let us always remember that plurality is here."

It’s poetic in a sense that it brings people’s false projections about technology into something that’s obviously shared no matter which technological cap you’re on. For people who are more graphically minded, we also have the SDG icons that explains the same idea.

In that, we use technology to enhance availability of reliable data, that we build partnerships out of those reliable data, and that we offer such open innovations always in a co-creative way instead of a colonizing way for our international partners.

This images, this very beautiful icons, thanks the UNDP for that, is very useful to, not only categorize the large picture which is by the triple bottom line, but also about each of the specific projects that we’re promoting. People can see instinctively which project correspond to which sustainable development goal.

It’s very rare that anyone would find that their goals are missing from the palette. Back in 2015, one criticism of the SDGs was there are no goals left behind. [laughs] All the goals are here. You will be bound to find something that is according to your view of the world that is somewhat represented by one of the 169 goals.

Just by promoting the idea of sustainable development in poetry or in graphics, that brings people to a dimension where people are much more ready to listen to one another than if we focused on specific technologies.

When you were discussing co-creating with opposing sides, you talked about the importance of face-to-face interactions. I noticed that some of the other techniques you use reorient communication away from face-to-face interactions. I was wondering if you could talk a bit about where you see face-to-face interactions being crucial.

We never replace face-to-face interactions, we augment them. That is the crucial difference. Instead of asking people to forget about town halls and come to this website, we make sure people still help town halls.

Their argument are mapped and archived and presented to the public on the website, that it become the agenda of the next town hall, that people can track each other’s promises, not just governmental promises but private and social sector stakeholders on the table who promise to do something to make sure that people have mutual accountability.

All this doesn’t take away anything from the face-to-face conversation. This is a democratic backbone for each meeting to maximize its social impact for it to lead subsequent conversations instead of being lost in translation. All the additional tools including Polis is never designed to take away people’s time facing face-to-face.

We call them...There’s many words like ambient technology, ambient computing, COM technology, and so on. All of these techniques are to ensure that we can spend more time listening to one another knowing that we have similar agenda, similar values already before entering the door to this conversation, and that people build long-term relations rather than one-shot transactions.

Too often in e-petitions, and participatory budgeting, or whatever, people would feel that it’s a transactional experience that leave people more empty than before. If you can continue and follow up this conversation all along its policy-making lifecycle, then people feel empowered and it becomes a relational thing. That’s my design criteria.

I wanted to revisit the conversation you were having with Maddy earlier about clarifying disinformation. You had this example about the hair product. I can’t assume that’s a one-off, because I don’t know what the discourse on hair products is in Taiwan.

In the US, there are a couple of things that are constantly inspiring disinformation like rumors about vaccines and their side effects, incidents of voter fraud, things like that. I was wondering in Taiwan if there are any topics like that you find you’re constantly needing to address.

It’s a fair question. First of all, I would say that we’re partnering with the global platforms on what we call a public notice system, especially on these things. You probably already are aware that whenever you see a flat Earth video on YouTube it will contain a link to Wikipedia that explains the spherical nature of the Earth. The same for vaccination and autism, I guess.

There is what they call infoboxes. That whenever people are spending time watching the conspiracy videos that, no doubt, earns YouTube some advertisement money, that they are using some of that gains to reduce the negative externalities that their product have on the society.

They’re being socially responsible to a degree, because these videos are the one that sells the most advertisement because it’s most addictive. It reaffirms people’s very conspiration theory that people already have some tendency to. I would say that people treat this as an addiction like liquor or things like that.

A little bit, it enables people to think more creatively but a lot, it damages critical thinking and creative thinking. We view Facebook to more a degree but also YouTube and so on with very similar eyes as large liquor companies.

I personally have a filter for that that I would recommend to people working on those rumors to protect ourself against those memetic hazards. It’s called Facebook Feed Eradicator.

I installed this in every computer and every browser that I use. It does what it says on the tin. It eradicates the Facebook feed. It takes the feed away and replace it with a quote, this time from Adler. You can customize that.

What it does is it takes away any parasitic artificial intelligence’s chance of learning what kind of things keeps you engaging in a non-externally useful way with those conspiracies. Rather, it makes every action on Facebook intentional. You have to search for something to see that. You have to visit a friend’s profile to see that friend’s profile.

You have to enter a livestream to participate in that livestream, not unlike a browser. It takes away the dopamine-addiction-psycho part, the FOMO part, the fear of missing out part from Facebook. This is in our K-12 media literacy curriculum to show the children how to shield themselves against the negative parts of the externality. Of course, it’s all opt-in.

Sometimes people enjoy the feed in any case. What I’m getting at is that if we with a calm mind see those recurring conspiracies as a symptom rather something that causes social rift, it is the symptom of that social rift.

By addressing those social rifts, not the symptom, in a way that are proactive like making sure that anybody recurring on one of those conspiracy theories, one of which concerns me personally...There’s a recurring conspiracy theory that I’m able to detect the brainwaves of any Facebook user and control their account, X-Men style. [laughs]

People who are locked out from their accounts or see their post disappear, or flagged, or something, they will say that Audrey is exercising her power of mind control. Sometimes they say this as satire. There are people who really believes that I can read the GPS signals of their mobiles phones and they have to turn their GPS off in order not to be tracked by Audrey’s mind-controlling machine.

I would say that this is a symptom of a social rift, that this shows our common value of caring a lot about freedom of expression, that people don’t want government surveillance of their everyday move and things like that.

Whenever people mention me by name in a public social-media post in Facebook, or Twitter, or PPT, or anything, I go and reply by myself in very short, to the point replies only on the part that are authentic to their own experience instead of the copy-paste style part of it or personal attack.

I was able to ignore all the personal attack. Whenever I see any word that makes me upset, I have a new nonverbal stimuli meaning that I put on some new music. Recently I’ve been listening to negative harmonies, that is to say, remixes of favorite songs but in a mirror image. It’s a new stimuli, or I make some new tea, or whatever.

Then I associate that particular upsetting personal attack with this new stimuli that feels pleasant. As long as I sleep for a full eight hours that night, it will form a new long-term association in my brain. Next time I view this personal attack I will feel very calm and indeed joyful by triggering those words in a different part of the brain.

It shows the trolls that they cannot make me upset, that if they propose such conspiracies, I only respond to the words that are authentic to their personal experience. You can Google it. It’s called troll hugging. It’s my hobby.

If I hug the trolls enough then sometimes they show up on Wednesdays to my office and we have a good chat, so by asking them to enter into a relational attention-seeking behavior rather than a transactional one. When they troll either they’re automated, which there’s different ways to detect and handle.

If they’re not automated, they usually crave something that they cannot receive in ordinary social study, that the stable relational attention. They troll someone, they get attention. It’s transactional. They wake up still feeling very empty, troll somebody else.

By responding only to the constructive parts, people reveal more and more of themselves. Finally, they can enter into a fruitful relational conversation. This kind of interaction I can do even without thinking as part of my instinct now.

People generally think that it’s an easy way to summon Audrey [laughs] just by mentioning my name and asking something they want me to clarify.

If people say, "Oh, it must be a robot. Audrey must have been using the transformer GPT-2 model to compose her replies," I usually just record a 10-second clip of me just authentically expressing my views. All this is just to make that perceived distance between the people and me much shorter than they had with traditional ministers.

Now, pretty much all the ministers and the premier have been learning this art. People find a much more closer relationship with the ministries.

That is the thing that heals the social rift, is that we don’t make the long silence the fruitful place for the negative projections and conspiracies to grow.

If you have a friend that you’re meeting every other day, at least every week, for dinner or for a movie, if you hear a conspiracy theory about them, of course your instinct will be, "Oh, I’ll check the next time we meet."

If that friend meets you every half a year, and speaks only in bureaucratic language, and you don’t even understand where they’re getting at, then of course conspiracy theories have their room to grow. I consider them symptoms, not a threat in itself.

You mentioned how you use satire and humor to address some of these things. Can you speak a little bit to if you use storytelling at all and how you do that?

Sure. Usually, we make the storytelling by way of videos, because we’ve found that they’re by far the most useful way across different modalities. I’ll just see if I can show some videos here. Supposedly there’s an album that says favorites. Here.

For example, this is the first film that I made as digital minister, but I actually have one with English captions. I’ll just translate it along. This is our first attempt at storytelling. We got better. This one feels special, because it was literally the first one.

We’ve since learned to make the captions larger so that, even on mute, people can see it very clearly. We’ve introduced better, more interesting music in the first 15 seconds to capture people’s attention and so on. We improve on the technique.

Mostly, the storytelling element is exactly the same. We’re just sharing our common values and making the values in a way that’s relatable to people, in this case just people cooking together in a kitchen, and making sure that we use the lay people’s words and, indeed, use people’s already imaginations.

For example, when we talk about accountability, there will be violent groups. There will be people that resorts to violence when they’re threatened, as shown in the movie. There will be people who just quibble among the slightest implementation details, as you showed here.

It’s not a harmonization of the actual policymaking process. This is the actual policymaking process with all its challenges somewhat comically represented, but all of them real. We can expand on that metaphor more in the interactions with citizens. They will not say that we’re utopian. Rather, we’re just working out solutions to those common challenges in governance. Yes?

I’m curious to get your take on the increasing critique of the philanthropic and private sectors replacing the role of the government in providing any number of services. What’s your take on that happening in Taiwan? Maybe if you can talk a little bit about any role that the private sector might play and some of the work that you all are doing.

A lot of my work is not just on reforming the state, but also to re-form the social and private sector so that they can work together on what we call social innovation.

Social innovation is defined as new ideas or techniques that everybody can adopt and participate, and is aimed at public benefit rather than private benefit. Participation from all to the benefit of all, and it’s new, so it’s social innovation.

In Taiwan, we’re very fortunate in that the lifting of the martial law, which was in the late ’80s, precedes the first presidential election by a decade or so. People who are taking co-ops or foundations or even companies for a social purpose have 10 years head start to build legitimacy as compared to the government.

When we say when the environmental activists set up their own air measurements, we can’t beat them, we must join them, I’m not saying something that is philosophically controversial. I’m saying a fact in Taiwan. If there’s a disaster and the local government publish a number and Tzu-Chi publish a number, people are going to trust Tzu-Chi number.

It’s just a matter of fact that the social sector has better legitimacy than the government. We accept that as public service. The payoff for a public servant to engage in social innovation in Taiwan is quite different from other places where there is not as strong a social sector.

Public service see the social sector here as partners. They’re not just community people that advocate but never implements. They’re people who implement sometimes better than the government.

All we need to do is just to keep each other honest, is to make sure that everybody declare exactly what they are working on and whether they’re following through with that promise.

We have a new website, SI.taiwan.gov.tw. We use this new domain, taiwan.gov.tw, for all the thing that transcends the central municipal boundaries, that transcends ministerial boundaries, so that people don’t have to go through the different policies.

They can just Google for "AI Taiwan." I think that’s the first one that shows up if the Internet is working. If you -- just a second -- Google for "AI Taiwan," you will probably see ai.taiwan.gov.tw. That basically shows the entire AI policy of Taiwan’s different ministries and different municipal aspects.

If you Google for AI Taiwan, you’re bound to find ai.taiwan.gov.tw because the domain is its own SEO. [laughs] It’s very hard to beat that SEO, [laughs] because the address is ai.taiwan. Then you can see all the different things that are ongoing in Taiwan as regards to AI.

The same goes for social innovation. It’s always at least bilingual. Anyone who speaks English is able to follow our government strategies. Here, you see the social innovation database where people can see registered organizations working on social innovation.

There’s two characteristics. First, it’s not limited to organization type. You can see companies limited by share, inter-firms, co-ops, foundations, associations, and every organization, including university centers for social responsibility. It is cross-organizational list.

You get qualified on this list. Every different organization need to follow different things. If you’re a company, you must disclose your company charter, and the company charter must at least list which sustainable goals you are working toward, how you’re allocating your governance procedures. That is like B Corps.

If you already have a B Corps certificate, you can just reuse that as an entry here. If you are an association, then you have to publish your annual financial reports and show how well the social impact corresponds to the declared association’s public benefit and so on.

We’re not making people conform to the same organization type. Rather, we’re making sure that the social sector, no matter which organizational form it takes, can reflect on the goals that they’re reaching.

The second characteristic is that every municipal, county, and city government can also choose the focus of their work in terms as SDGs. For example, Taoyuan City, which has its own voluntary local review of SDG KDIs, can both list the associations, companies, and co-ops registered Taoyuan that are helping the city reach its own SDG goals.

Also, those SDG goals, across Taiwan, what kind of other associations not based on Taoyuan is nevertheless willing to work with Taoyuan on these matters? This is a very systemic, holistic picture of people’s common goals.

In this, the private, the social, and the public sector basically keep each other accountable. The government is not saying this one is private sector and this one is social sector. We have a lot of co-ops that are bridge-making between those two, so we let them declare by themselves.

Once they agree to this accountability, if they enter the supply chain, that is to say, if they have a product or service to offer, and other entity purchase more than five million NT$ per year from the registered entities, then I personally go out and give an award to those purchasing entities.

We make sure that people integrate each other’s social, environmental, and business cause into their supply chain. The supply chain is then made with purpose rather than with just GDP, GNI, or other purely quantitative things.

We’re not saying that GDPs are going away. We’re saying that we’re complementing the quantitative measurements of economy with also the social, environmental part. All of them will become quantitative at the end.

Anyone who use this mark of #TaiwanCanHelp basically agrees to be led more or less by their partners of the same SDG values in a network governance or co-governance way. This is a much larger picture than digital transformation or anything like that.

This is fundamentally shifting the relationship of the private and the social sectors. We’re still early on in doing this, but we’re already very well encouraged by the partner of 400 or so organizations joining this program in past couple years.

I’m very aware of the time. It’s five o’clock. We’ve maybe one more question. We don’t want to impose on you.

Mine is quick. It’s a fast question. I want to ask if you are an avid reader before I ask the next question. Do you have books that you read a lot or not really?

I prefer digital rather than paper books. [laughs] I would say that if you look at my Kindle, I’m an avid reader.

What are five books that you would recommend? What are your five books you like most? It doesn’t have to be recent. It could be that you’ve read over the time you’re...

More recently, I would recommend "The Age of Surveillance Capitalism." It’s a pretty good take. It’s a little bit mainstream in the sense that it’s still using mainstream, US-based economic analysis. It’s not Marxist, but it make really good points. That’s one.

I just, a couple months ago, met the professor of offensive realism in international diplomacy, John Mearsheimer. That’s his latest book on "The Great Delusion." I learned a lot from that. Although not very mainstream, I still think that it offers a structural way of understanding international relationship.

The BBC’s anchor interviewing me this week is Carl Miller, the person that wrote "The Death of the Gods -- The New Global Power Grab" that talks about network-making power rather than hierarchical power. This is, again, very contemporary. That’s the three contemporary books.

Otto Scharmer, with the "Theory U," is sometimes seen, of course, as kind of a little bit too idealist about the potential of organization change, but I do think that it’s now a good time to reevaluate the Theory U.

Previously, it requires a lot of adeptness of the facilitating leader. Nowadays, you can actually imbue a lot of that into space design, interaction design. If you view it through an interaction design lens, then the essential of Theory U is actually still very relevant, as well.

Finally, I would recommend "Radical Markets," but full disclosure, I’m on the board of the director of RadicalxChange. This is kind of our basic 101. Radical Markets, I like it. Not because of any of its concrete proposals, but rather its firm rejection of traditional left-right angles.

It’s basically using mechanism design to deliver a more long-term planning and, indeed, more fair society by using exactly the same language as traditional economists. You may find it very strange, but actually, I think, it paves the way to think outside the boxes of left versus the right.

Are there any Chinese language titles?

If you want a Chinese title, I would recommend Ursula K. Le Guin’s translation of the Tao Te Ching. It is still a classic.

This is actually not the Ursula K. Le Guin translation, it’s the more scholarly one by James Legge. I still drive a lot of my main attitudes from the thoughts of Lao Tzu, and the thoughts of Lao Tzu is why I call myself a conservative anarchist, because they both conserve the tradition, as they are, the various traditions.

Also, they insist on anyway goes, anything goes. This is like a core against method, a way of anarchism without objectives. Just a Taoist thing, but Ursula K. Le Guin’s rendition, you can find on YouTube, her reading it out loud and with musical accompaniment, and I think that makes it crystal clear in a poetic fashion what Lao Tzu’s thought really is about.

If you want to learn about Taoism, I think that’s easily the first and the most clear source of wisdom from that particular tradition.

I hate to break your habit. Something very heavy is coming your way. It’s in paper.

I’m delighted to see, in your quick scrolling, I did see T.S. Eliot and selected poems, so I see you do read some traditional...You are a poet yourself, so thank you for two hours of narrating poetry, storytelling. You love digital, this is analog.

That’s OK.

This is a thing, but it’s about people. It marries both the transitions you’re trying to make...It’s a book on a traveling show, "The Empresses of China," which the Luce Foundation funded, and it’s heading off on its road trip. It was birthed in my home state of Massachusetts, PB the Essex museum, but it’s about to go on the road.

It just shows that history repeats itself, because it’s about very strong women, so the #MeToo movement comes a lot, but it goes back centuries.

For your coffee table, I hope it doesn’t collect dust. I think you’ll enjoy it. It might be refreshing to hold something. On behalf of all the Luce scholars, and the Luce Foundation board, and the Asian Foundation, Ling and me, from the Luce Foundation, I thank you very much.

We all thank you. It’s been just a breathtaking two hours. I’ve learned a lot, but I’m old school, so this is very helpful to bring me up to the 21st century. Thank you so much.

Thank you.