Hello and welcome. While the video projection is still being set up, let me first say welcome to the Executive Yuan. This is my boring office. Normally I am not here. I am here every Thursday only for the cabinet meeting.

Normally, I am in the Social Innovation Lab, which is a much more likely place that I will project in a second. Still, welcome to Taiwan and I look forward to have exchange with you. I will be on the record, but under certain house rule for you. For me, I will attach my name.

We will make a transcript so that every one of you will receive the transcript and we can all co-edit for 10 days before we publish to the Internet. If you would like to add your name to your questions, you can add it afterwards. Otherwise, we will just say audience member for all of your questions. If everyone is OK with that, then we can begin. How shall we proceed?

Tell us about your role?

I’m the Digital Minister in charge of social innovation, open government, and youth engagement.

You look very unlike a minister – the long hair, the robe.

You have something against long hair? [laughs]

No, I’m fine with long hair. Yeah, it is good to see a minister like you.

OK. Thank you. As I said, I’m a horizontal minister. In the Cabinet, we have 32 vertical ministries with 32 vertical ministers. Above the 32 in the Cabinet, we have 9 horizontal ministers. Each of us are in charge of cross-ministerial issues.

I’m one of the nine with the portfolio of, as I said, social innovation, and open government, and also youth engagement. I promise to show you my real office. This is my real office.

It’s much more interesting than this one. Aside from Thursday, I am usually there. Every Wednesday, everybody can come and talk to me for 40 minutes at a time. We recently tore down the walls, so people can directly walk in from the street.

This is in the heart of Taipei, what we call the Social Innovation Lab. It’s next to the Daan Central Park and the Jian Guo Flower Market.

I read that you took initiative to the direct democracy where ordinary people can influence the decisions between the elections. Could you tell how this practically works?

Ways of direct participation, let’s correct a little bit. I didn’t quite advocate for direct democracy. We usually say it’s participatory democracy. When many people participate and listen to one another, we say deliberative democracy.

Direct democracy summons the idea of a referendum. Because our work is primarily in the discover and define stage, and not at the decisional stage, while it is also direct – it’s indirect, of course – we don’t want to summon the words referendum here.

We are in the participatory or deliberative phase of democracy, but I’m happy to give you some examples.

How do you do it?

Very carefully.

I will, in very broad brushes, say for an ordinary citizen how they may engage with this, what we call, co-governance or co-gov system. One way they can do it is simply by showing up on a Wednesday and bringing with themselves any idea that currently don’t have a policy to allow.

For example, people may bring self-driving tricycles and want to test it out for a year or so. This is called a sandbox application. That’s one way. The second way is that they can get 5,000 people on a petition system.

With 5,000 people, any minister must respond within 60 days. When it concerns two or more ministries, then I can show up. I will be summoned by the ministries to the place where they did the petition.

It could be on the rural places, remote islands, or any of the indigenous nations, for example, to solve the issues there. That’s the second way. The third way is that they can participate in their regular town halls.

I attend those town halls by myself with a small staff every other Tuesday or so. When I attend, they just show up in where they are already meeting, but I bring with me the 12 ministries across video conference in five municipalities to listen to people carefully in a distributed way, this is called the Social Innovation Tour.

You can find all the records in si.taiwan.gov.tw, so that’s the third way. The fourth way, is that they can propose an idea with the public service on the presidential hackathon, to propose for example, using machine learning to solve water pipe leaks. They prototype for three months, and maybe they win a trophy from the president.

Once they won, the President promised to do whatever they prototype in the 3 months into public policy on national scale within 12 months, so that’s the fourth way. All the previous four ways may in any stage also reach one of our reverse mentors.

In those 12 ministries, each one has maybe two reverse mentors, and so all the ministers relating to social innovation, for example the Minister of Labor, has two reverse mentor and they are all under 35 years old.

A reverse mentor, for example 黃偉翔, who was just 20-something when he joined as reverse mentor for the Minister of Labor, proposed we have the world’s skilled, like the most skilled people in Taiwan who won collectively the third place in the annual World Skill competition to not only go on a national day parade like the athletes, but also integrate them into the basic education system to revamp to the schools together with the children so people learn to aspire to the skill level, and the creativity, and the design of these people.

This is not direct participation, this is more about reaching out to a reverse mentor who take the various idea in those four stages directly to a minister. The minister agreed to be a kind of champion of that idea, then ever four months, the Prime Minister says OK to those reverse mentor ideas, and that becomes national policy without having to wait three months.

That’s kind of a short cut, but is determined by the idea that young people lead the direction and the older people implement those directions, so that’s the five main venues that I’m in charge of. Yes?

Hi, my name’s Daniel from New York. I’m curious, one of the things that I think has driven innovation in Silicon Valley, for example, is lack of bureaucracy, a lack of government interference. I found even though Taiwan has all these great tech companies, I think it’s limited, from I’ve read, less innovative technologies, more kind of commoditized technologies.

I’m wondering first whether you think that Taiwan is an innovative society and innovative economy, and if not or maybe it’s not as innovative as it could be, what kind of steps are you doing in your capacity as Digital Minister to spur innovation?

The World Economic Forum seemed to consider Taiwan as one of the four super innovators for two years running now. We just got a report from the Civicus Monitor, again for the second year running, that we’re alone in Asia in being the only society that is completely free and open for the civil society.

What this means is that the bureaucracy is not in the way of freedom of expression, of assembly, of trying out new ideas. That has something else, a different connotation, because you’re primarily saying that the private sector should not be interfered by the bureaucracy while they’re pursuing private-sector innovation.

The problem is that a lot of those private-sector innovations has negative social externalities. When that happens, for example surveillance capitalism, and things like that, we treat them as important issues to tackle, and we don’t think surveillance capitalism is somewhat preferable to a surveillance state-ism.

In Taiwan when we say innovation, it is always trisectorial, so our unicorns don’t pollute the social or environment spheres, they have to pursue inclusive worlds. For example, Gogoro, which is a unicorn in Taiwan, is at once solving the motorcycle ridesharing issue, but also is a investor in renewable energy, battery management, electricity, and things like that.

They have to prove in all the 33 bottom lines to be given free reign by the sandbox system, Presidential Hackathon system, whatever other system. If they pursue economic liberty, but actually taking away from the society or from the sustainable environment, then, of course, our co-governance mechanism is designed to keep them in check. Yes?

I’m from Italy. Since I was thinking about this kind of direct democracy, since in Italy, we have the major, the five stars to promote this kind of direct democracy, who runs the platform of the vote? Is there a way to verify themselves of the vote? How can people be sure that the results is real?

Really good question. Before we talk about open innovation and about cross-sectoral partnership, the most important thing is that how reliable is the data that each sector is handing to one another. This is, of course, a crucial issue.

In our social innovation platforms, most of the governance is done by the social sector, meaning that, for example, the AirBox, which is the example of people measuring air quality in their balcony and in their schools, all those 2,000 points are governed by the people themselves.

Meaning that they use distributed ledger technology, also known as blockchain, and other ways of computation to make sure that nobody here can take control of the other people’s measurements, modify them after the fact, and so on.

Using secure technologies, people can feel easier when new people participate in this network. Collectively, they form what we call a data coalition, meaning that they may serve different interests, like the schools may only want to teach about environmental sustainability or about data stewardship.

Of course, people who are living closer to the industrial areas and so on may have also the campaign will to basically fine the people who over-pollute and so on. While people have different interests, they can rest assured that if they join this kind of distributed ledger network, they can reinforce each other’s bargaining position.

They form a coalition called a data coalition. They collectively bargain with our environmental minister. Says that, “Look at us. People are trusting our numbers, and they are not trusting our numbers, because you only have 80 stations, but we have 2,000 stations.

“We want to share with you our legitimacy, but we want something in return. We want you to run our equipment, joining our network, but in these areas that they want to measure, but cannot.” What are these areas?

These are industrial parks. These are private property that they suspect may or may not be contributing to pollution. They cannot break and enter. It turns out the public sector owns the lamp. We can go in and install 10,000 lamps.

In 10,000, their AirBox, in their protocol, using the same, what we call sensor things, protocol. What it does is that it gains the government legitimacy, because the people would say, “Our numbers agree with the EPA’s.”

EPA also give people legitimacy by participating and supporting, but do not take control of those data coalitions. That is the usual way in Taiwan that we’re forming those data coalitions. Once people trust each other enough, they may evolve to become a data cooperative to govern this in a democratic fashion.

The trust must first be gained with the social sector who trust each other, and then the private sector and the public sector join. At least, that’s what we have found to be sustainable in Taiwan.

Do you receive many hack attacks from China?

Yes.

You block them?

Yes. That is why the cybersecurity industry in Taiwan is booming, because we don’t have to hold training lessons. They are just in the field. [laughs] That doubles as a battle-hardened testing for our equipments and so on. If they can survive in public sector for an extended period, it’s probably pretty safe.

The 1987 in the corner there, this isn’t from 1987?

No, that’s the number of stations, and it’s an old picture, so it’s way beyond that now. Around that time – it was two years ago, I think, when I took the snapshot – there’s an additional 1,000 or so stations all over the world.

They don’t have to pay patent or copyright license to Taiwan, because all of this is on GitHub. Everybody can just download the code, find some Raspberry Pi or some open hardware, and start participating in this network.

It’s called AirBox?

It’s called AirBox, yes. The same team participate in our Presidential Hackathon and is now working with New Zealand on a successor project called the WaterBox, which does the same thing, but measures the waterways in the arable lands. Yes?

First, it is a little bit unusual that a minister, someone at your level, would be so engaged with a social innovation lab at this level. I have two questions. One, what are you building this model off of, and two, what do you ultimately hope to achieve?

Build off, as in resources that it takes to maintain this work?

No, the origins of this approach.

Ah, the origin. How did I learn this, right? OK.

I dropped out of junior high school when I was 15 years old. I told my principal that I learn of this new thing called the World Wide Web, just invented then.

I wanted to participate in the building of the World Wide Web, and I find some really good websites, like arxiv.org, that host the pre-print. People, before they publish to journals, post their drafts for the academic community to review.

Even now, as a digital minister, I still publish on social archive to give back to our tribe, so to speak. In any case, this way of open access and open source – at that time, still called the Free Software Movement – eventually leads me to a political system called the Internet Society, the Internet engineering task force.

By participating in this Internet governance system, I learned a new way of political work, which is people without coercive power. The Internet doesn’t really have any, what they are calling it now, any army, navy, or whatever other way of force.

What the Internet has, however, is a set of protocol that we show people, if you adopt this voluntarily only, and allow for what we call permissionless innovation, then it’s good for everybody involved. Using this way, the Internet is at once owned by everyone, but also no one.

Eventually, the Internet Society become not reporting to any sovereign entity. This whole journey taught me about the way of what we call rough consensus and running code way of governance. I worked in that community for six years before even I get the first vote in Taiwan because I was very young.

Because of that, that has always been my native political system. It was only used to govern the Internet itself. Now, we are using the Internet to govern with people as well. The key difference with the Internet way and the traditional way of governance is that traditional governance builds legitimacy by showing that they’re working for people.

The Internet society builds legitimacy by showing people that they’re working with people. That is the crucial difference. That is, I think, ultimately what democracy need to move toward to.

Thank you.

Yes?

What to do with maligned-formed regimes who seek to use the Internet in order to attack democracies. You being here in Taiwan. We’re seeing these sorts of attacks worldwide.

My understanding is that you’re also looking at solutions as to how to address that challenge as well. What is Taiwan doing right now in practical terms of defending its society against its government, against attacks, disinformation that happening online, as well as cyber?

Sure. The cyber part is rather easier to explain because we allocated, for all new government initiatives, five percent to seven percent of budget just for cybersecurity. That’s aside from the ICT budget.

This is very important because it ensures that in Taiwan, if you are a white hat hacker, someone who exploits system flaws and let the people know about it and how to mitigate it, you get paid pretty well. You get a lot of social recognition. We say the cyber defense is national defense. You meet with the minister and the president quite frequently and so on.

Basically, by giving white hat hackers the due social recognition and payments really, and a really good career wherever which initiative to want to go to, for being an ethical hacker, we ensure that they don’t fall to the dark side which always has more cookies. [laughs] That is the easy answer about the cyber part.

The harder part is about disinformation because it changes literally every second. They exploit all the wedges in the society, try to not really interfere with any particular result but just to get people disappointed about democracy. Just to get people disenfranchised, disempowered from the political process. That is the ultimate goal of sowing discord and disinformation.

The way that we found of working toward disarming disinformation does not involve infringing of journalistic freedom. Many jurisdictions took the shortcut of infringing of journalistic freedom, and I would not comment on those jurisdictions. [laughs] In Taiwan, CIVICUS Monitor says that we’re free and open.

Our three reactions to disinformation, the first one is timely response. We have this triple two principle. Within two hours, all the relevant ministries need to roll out clarification of 200 characters or less and at least two pictures. They are now good enough that most of them can deliver within 60 minutes, the clarifications.

The clarifications are the work of memetic engineering. It means that the clarifications must go viral by the moment without people seeing the disinformation, because it doesn’t go viral by its own. It actually reinforces the disinformation message. It has to be a memetic work by itself.

This is one example. It says, “Popular rumor, perming your hair will be subject to $1 million fine starting next week.” That is not true. The younger version of our premier says, “I may be bald now, but I would not punish people with hair.” A fine print that says, “Labeling requirement for hair products start to take effect on 2021.”

The premier here says, “But if you keep perming your hair many times a week, while it will not damage your pocket, it would damage your hair. When serious, you may look like me.” It is actually funny because he makes fun of himself. It’s humor. It’s not really sarcasm. It goes viral.

If you search for perming hair fine, all that shows up is this memetic engineer product. We wrote this out quick enough within one hour, then it actually goes very viral.

When people see this kind of pictures, when they see the next disinformation payload that raises anger, they will not turn the anger, which is helplessness, into outrage which is fake action, really. They will actually turn this into fun, which is humor.

Once you laugh about it, you really cannot be mobilized by the disinformation package anymore. We become inoculated. That’s the first response, is through humor. The second one is through voluntary participation.

20 years ago, when we worked on the issue of email spam, Bill Gates was saying at that time that we must start charging postal stamp for each email. Otherwise, email is broken because it costs nothing to send an email.

It turns out that we didn’t collect stamps, but rather, we collected spam. People voluntarily set up honeypots to attract spam. People voluntarily flagged their inbox with spam in it into international social sector or organization called Spamhaus. What they did is they used machine learning to analyze the pattern, behavior, not the payload but the behavior of the spammers.

The next time they send an email, they go into the junk mailbox. They do not go to the inbox. We’re setting something very similar as Spamhaus in Taiwan with participation of the social sector. For example, the LINE, which is like WhatsApp, is end-to-end encrypted. We really cannot peek into its content. LINE corporations say it cannot peek into the content.

LINE relies on people to report whatever they feel as abusing their personal, friendly relationship of messenger so that whenever they see something that has the potential to go viral and they suspect that it might be disinformation, what we try to do here is that we invite people to flag it.

LINE agreed to partner with four social sector organizations that are MyGoPen, Cofacts, Rumor & Truth, and the Taiwan Fact-Checking Center so that every time, actually every day, this is today’s number, there’s 147k people flag of rumors on the end-to-end encrypted system, which is like WhatsApp. We have to rely on people to flag. Otherwise, we don’t really know what’s going on.

They have reported uniquely 32k different disinformation packages. 1.5k of which gets trending, so the clarifying fact-checker spend their energy on it, and they goes back and circles back to the people who reported and also everybody else who have flagged it. These are the trending disinformation of the day.

If you look at the Taiwan FactCheck Center, which is TFCC, they are professional journalists that does the same fact-checking as you do, but they are very public about it. When they publish something that is false, for example, there’s many interesting disinformation recently.

For example, there was this one which was clarified in November 27 that says, “Yesterday, there was a protest in front of the presidential office, but the news says nothing about it. The government must be controlling all the news agencies or something.” That was the payload. It very clearly says what the video showed was from 2016. It’s a very old protest.

Of course, that’s malinformation actually, it’s not just disinformation. When it’s already shared for 160 times, the TFCC goes public with its fact-check. Once it does that, Facebook tunes the algorithm so that these no longer reach the top of people’s newsfeed. You have to scroll for two hours to see that. It’s like moving this the spam folder.

Once they do that, it stops spreading. If you check, it’s still there. It’s not a take down. It’s just it prefers not to show it on the newsfeed. It relies entirely on this crowdfunded, nonpartisan Taiwan FactCheck Center to control the algorithmic input of this kind of behavior.

Basically, by combining this downgrading of virality by fact-check social sector and operating of the clarification messages by memetic engineering, we make sure that people can see the government not as a fact-checker or as a journalist, but as an eager participant in the puzzle-making process of newsmaking.

Everybody can potentially become a fellow journalist. That is what we ultimately want to do. I hope that answers some of your questions.

Yes, thank you. That’s absolutely brilliant.

Thank you.

Who’s the liaison with Facebook for example? Is it the government or is it the…

At the moment, I am the liaison.

Only for disinformation litigation. Yeah?

Facebook has agreed to actually dial down…

Yes, they had, because they are reasonably sure that the TFCC is not controlled by any political party.

This is Facebook Taiwan, not Facebook in many…

In HQ. The HQ has to approve of it, of course. You make a really good point. Like LINE, the LINE dashboard that I just showed you, is worked by the LINE Taiwan CSR department. LINE is, of course, operating in other jurisdictions as well.

Unless they can feel that their social sector partners are actually social sector and not captured by a private interest or the ministerial or party interest, they would not do something like that. This way of solution is only possible in the CIVICUS Monitor, like completely open and free society, because then you have a higher legitimacy in the social sector to not be captured by the private or the public sectors.

Yes, your point is a very valid one. Also, for example, our Control Yuan, which is a separate branch of the government that does national auditing, they publish all the campaign finance including donation and spending. They do so completely transparently.

They do the audit, but they publish the raw data of each incoming campaign donation and spending so that people can also do independent analysis. We have data scientists and also investigative journalists to look at from the previous election how the campaign donation flows.

They found, for example, for social media advertisement, they just don’t go into campaign finance and donation because people just pay directly the social media for precision targeting.

We say to Facebook and to every other large platform like Google, Yahoo, PTT, LINE, and so on, that says, “OK, this is the norm here in Taiwan that we have a separate branch of government publishing completely transparently all the campaign donation which you can only donate when you’re a citizen to begin with, not a foreign citizen.”

Because of that, we want the same norm in your platform. We want you to bend for unresponsive propaganda. We want you to review it to the same degree as we do in campaign donation. Otherwise, we’re not passing any law. There may be social sanctions.

Facebook said, “OK.” They published a report just last week actually of all the social issues and political advertisements, the complete breakdown of exactly as our campaign donation. For this election, Google and Twitter says, “OK, maybe we’re just not running political ads altogether.” There’s, again, a norm first way of liaison with their multinationals.

We heard so much the other day from the Institute for National Defense Research in regards to their concerns with the upcoming elections, and certainly the disinformation they’re expecting from China.

Is there a coordination going on? I heard you mention obviously the other ministries involved with the innovational lab for something like this.

As part of my working condition joining the cabinet, because I’m at a Lagrange point between the social movement and the government, there’s three conditions. The first one is radical transparency. Including this meeting, all the meetings that I chair is published, completely transparently, online.

As I mentioned, this is a good idea because people who argue for private sector interests would have to argue from a Sustainable Development Goal perspective instead of from a purely selfish lobbying perspective. This is the first condition.

Aside from radical transparency, there’s also the condition of voluntary association, meaning that my office is one delegate from each ministry, and they must join voluntarily. The National Defense, which is the Ministry of Defense, never send anyone to my office so I really don’t know about what they are doing.

This is a good firewall. Otherwise, all the proceedings, all the meetings that I meet will be public, and that will be learned by adversaries. I only work on things that we’re not afraid for the adversaries to know about. This is a delineation of responsibility.

Finally, the third condition is that anywhere I’m working, I’m working. It’s teleworking. That’s why I get to work in the Social Innovation Lab or, indeed, tour around the world, but still is working. I hope that answer your question. I know what’s going on in the abstract, but I cannot learn about any specific case.

There’s no coordination that’s happening at this point?

There is coordination, but it’s one-way. What we’re doing is coordinated into the National Security Council, but what the National Security Council is doing on a case-by-case basis, I know of the, for example, Rapid Response System and mechanisms they’re setting up, but I’m not party of any specific case that they’re handling.

Thank you.

Yes?

It’s all really fascinating, and I’m following it, I think. I’m very tech dumb, so I find it all really fascinating. To your knowledge, this project where you’re marrying a social movement with the workings of government, how unique is this among governments around the world? Is it something that maybe Taiwan is a leader in?

We’re definitely a leader, but we’re not unique. We learned a lot of our ways, we co-developed a lot of those thoughts with, for example, NESTA in the UK with Geoff Mulgan and friends. They have had some success in, for example, introducing in the UK the Policy Lab, which is a cabinet office, very similar to us, that works on scalable listening.

Our system, for example, we don’t develop ourselves. The AI-based listening device was invented in Seattle by a bunch of Occupiers. What we’re doing here is – as I said, whenever there’s a social innovation, people react differently – we use AI for listening.

For example, this was the UberX conversation. Everybody can see where they are among their friends and families in the opinion clusters. It doesn’t pay to troll or to mobilize people. If you mobilize a thousand people voting exactly the same, it will just be a dot here, and the area which measures plurality will not expand.

By sharing the data, starting from social sector data, and asking how people feel, this AI system resonates with people’s feelings so that by saying agree or disagree, you can move among your friends and families who are not nameless enemies or adversaries. They’re your friends and families.

Once you agree or disagree a few times, you can share your sentiment for other people to resonate with. This invented in Seattle and now deployed, for example, to Bowling Green, Kentucky, USA. This is a real town hall. We see the same shape every time we run pol.is, but this is from Bowling Green, Kentucky.

You can see there are five very traditional ideological, divisive statements that instantly tell whether somebody is Democrat or Republican. These are the consensus statement that most people agree with most of each other on most of the things, most of the time.

For example, they all agree that the arts are important. STEM need to become STEAM. Nobody is against that. It costs practically nothing. Any mayor that adopts this gets reelected chance maybe improve by 0.5 percent. By getting those rough consensus out, we show people the polity is not what people think from the ideological divisive messages.

This, again, is not invented in Taiwan. We just deploy it here. The petition system we learn from Better Reykjavik, which is the Icelandic city that builds this from a offshoot of Pirate Party International. I think it’s called the Best Party of Iceland at the time. We learned from their platform as well.

We’re also working with the CONSUL system, which is the 15-M/Podemos movement that, when they took Madrid City a few years ago, started working on a Spanish-speaking participatory budgeting platform for all the Spanish- and English-speaking countries. They got a UN Public Service Award for that.

We took a lot of these elements in our participatory budgeting element as well. As I said, it’s a global movement. Taiwan is a pretty good lab, but it’s been done in many places, but usually on a municipal scale.

That’s what I mean, on a national…

Taiwan is just a large municipality anyway. From Taipei to Kaohsiung is just 95 minutes by high-speed rails, so people feel municipal even though it’s 23 million people.

A digital republic. This look fantastic, but we was, before today, by the Sun and Moon Lake. Here, we was in the four-stars hotels. These four-stars hotels was so horrible Internet that send email was problem. My question, is 100 percent of island…

Covered?

…covered Internet? Thinking about all people, but about the old generation when sit not by the computer and don’t have new mobile, don’t have Internet. What we look here, this is, I think, for the young, middle-aged generation. Sometimes, for old generation, this is a little bit other-world.

I totally agree.

This is a very big part of voters, very big part of the society.

I totally agree. Was it the hotel WiFi that was bad, or was the 4G signal?

Maybe WiFi. I don’t know. [laughs]

If you have roaming, then Chungwa Telecom should work. If it doesn’t work, it’s actually my fault so I would love to learn about it.

I heard that all the 4G stations, combined, covers 98 percent of Taiwan’s areas and population. It should be very affordable, like 15 euros per month to get unlimited data access. If you don’t have 4G reception on Sun Moon Lake, that’s probably my fault.

I really need to learn about it.

They just had to reset it. It didn’t happen to me…

It wasn’t cellular. It was WiFi.

OK.

I can’t vouch for any hotel’s WiFi. That’s really beyond me.

My point here is that your point is correct. The elderly people often prefer face-to-face meetings. We found that when we’re discovering the common issues, defining them together, face-to-face meeting is a must.

In Taiwan, we say, “Face-to-face meeting builds 30 percent of trust,” and over the Internet now, using a projector, maybe 20 percent of trust. This means that not only high bandwidth live streaming is important, but it’s important for me to physically travel to the elders, both tribal, like indigenous elders, and also elderly, and listen to them where they meet.

Sometime, I spend two nights or at least a night with them on a ethnographic hangout and just hang out with them for a couple days. When we hold these public town hall meetings, for them, it’s just showing up to where they’re already meeting, talking about the issues they’re already talking about.

We scaled this part of listening by showing how to do it to the district offices. The district office, when it talk about regional revitalization, which is our flagship national participatory budgeting project, they gather the elderly people face to face, but we use technology to amplify their voices. We bring tech to people. We’re not asking people to come to our platforms.

The e-petition and so on just sets the agenda of what’s to be discovered. When we’re defining the problem, we always end up in a face-to-face meeting in the place where it happens. That is the way to bridge the younger and the elderly generation together.

We don’t think this problem solves itself with time. We think broadband as human right need to be constantly deployed to make sure that we take care of the people who are most vulnerable. The remaining two percent are around mountainous areas, above 3,000 meters, and so on.

Even for these places, the Minister of Interior agreed to use helicopters for training sessions to help set up telecommunication towers. There’s really fanatical [laughs] commitment, and we say that because otherwise we cannot with a straight face say it’s a digital democracy. It would be excluding people, systemically, and that must not happen.

Are you going to use also the 5G technology?

Yes, of course. The 5G auction starts real soon now. We’ll finish next January on the first band release. We’re still on the leading group in that.

We also say each social innovation application, if they need 5G to test out, for example, self-driving boats in areas that really need this kind of transportation, medical uses, or things like that, they can apply for a specific 5G testing area, a sandbox that uses a different band than the commercial 5G.

They can partner with telecoms, or they can partner with the equipment companies. They can do a local test on whether 5G actually solves this social issue without relying on the commercial telecoms to deploy it to those lower-resource areas. Both sandbox and the general availability of spectrum starts this month. Yes?

I’m wondering how you think about the scalability of these things. I think we’re all impressed and encouraged by some of the distributed governance stuff that you’re doing. When you compare it to across the Strait, they do facial recognition and social credit scores that are done at such a massive scale with so much data coming in.

It’s hard to see this catching on in a way that some of the more coercive and violent technologies do. I wonder, even aside from your role as minister, maybe just as a technologist, how you think about the challenge of scaling distributed technology versus more authoritarian technology.

As a expert in distributed computing, I assure you distributed scales better than top-down models. We just saw that in Hong Kong. We saw that authoritarian technologies versus distributed technologies in a place where you still have Internet connection – that’s, of course, a given – distributed methodologies won pretty handily.

What I’m trying to get at is that the economy of scale is one thing, but scaling in terms of social legitimacy is another altogether. No matter how quickly the PRC is running into the opposite direction, their direction doesn’t naturally scale to people with expectations from liberal democracies.

They only get to convince people if these people are already disappointed enough with liberal democracy. They think that one vote every four years do not reflect truly people’s voices. In these regimes, sometimes the authoritarian model is convincing.

If the general public feels that they have the chance to participate not only every four years, but every quarter, every two months, or in the weekly tours every week, then that’s a natural inoculation layer against this authoritarianism scales better.

This scaling up narrative, scaling out, and scaling deeply is inherent in any social configuration that the people feel that they still have a stake in where the society is going. That goes really deeply, and I’m still cautiously optimistic about our way of scaling.

You had a question?

From your point of view, do you think the China issue is exaggerated?

Sorry, which China issue?

The China threat.

The PRC’s threat to Taiwan’s democracy, let’s frame it like that. It is true that when we look at the CIVICUS Monitor, you almost feel that we’re defining ourself as what the PRC is not doing. What PRC is doing gets questioned very heavily by the Taiwan civil society as something not to follow.

I wouldn’t say it’s exaggerated. I would say it’s constitutive, meaning that whatever the PRC is doing on the opposite direction serves as a reminder that we must not, for example, cite disinformation has harming “harmony” and infringe on journalists’ rights. We must instead roll out funny cat pictures. [laughs]

I don’t think it’s a exaggeration. The threat is real in both legitimacy and also day to day areas, but I think the threat is challenging us to find more solutions for liberal democracies.

Building off of that point, it seems, at least to me in a very short amount of time here, four days, that the whole country self-identifies through the prism of China.

Of being not PRC, yeah.

Yeah. I come from a country where, as well, our neighbors are definitely not allies.

We don’t, internally in Israel, self-identify through the prism of not being them. We take a very internal self-identifying approach. Do you think that it’s helpful in Taiwan, or maybe I’m just wrong in my perspective, that so much of society, from politics to social to tech to commercial, everything is seen through the prism of not being China.

If you look at this picture, there really is no other easy contrast.

Value-wise, we’re quite similar to Japan and to Korea. We don’t have to define ourself as somewhat like Japan and Korea because they don’t define themself as that as well. [laughs]

With Japan in particular, in the Global Cooperation and Training Framework, Japan just proposed that they join this – originally bilateral between Taiwan and US – training framework. They came up with the idea of “Indo-Pacific.”

I wouldn’t say it’s just Taiwan that is defining as not PRC. The entire participants in both CPTPP and the Indo-Pacific Initiative is saying, “We’re trying to build a world order that is still based on the ideas of liberal democracy and the rule of law.” That is something that we have in common.

You’re correct in saying that Taiwan is, additionally, defining ourselves in a kind of nation-building way, with the PRC constantly trying to define in a way that says “breakaway territory” or something like that. I think it’s a natural reaction to the “breakaway territory” narrative.

As I usually say, the breakaway was true, but it was during the Neolithic era. The Neolithic era was 8,500 years ago when the land bridge in the Taiwan Strait gets submerged by water again. After that, it’s territorially, geographically breaking away already.

The very fact that I have to come up with these counter-narratives is that the international community keeps asking me, “Hey, they keep calling you a ‘breakaway territory.’ What do you say about that?” It is, again, a mutually defining thing. I do agree.

With time, maybe we will identify, instead, more about being beautiful islands in the whirling oceans. That is definitely the way to go. We are still in the process of that.

You have said that you’re going to use the 5G technology. However, after the Huawei scandal, 5G has been accused for spying, in other words, more cyber attacks. That’s my first question.

My second is, after the Sunflower Movement, you were invited to join the government. Will you keep your position after the elections in case that KMD take the presidency?

The first question is easily answered in Taiwan context. The Occupy was done by 20 or so NGOs at the time, each deliberating about one specific aspect of the Cross-Strait Service and Trade Agreement.

The legitimacy theory was that the MPs refused to deliberate on it, so people took people’s representatives, Parliament, and start deliberating on it. It’s very legitimate, as you can see. Those 20 NGOs, one of them is deliberating on the telecommunications services.

At that time, there was pre-4G. When we occupy, we were still using the YMAX technology, which serve its really good use for the Occupy. [laughs] At that time, we’re migrating to 4G. The entire core and periphery network was not yet built back in 2014.

The consensus on the street with half a million people and many more online is that it’s not about any specific company. We didn’t mention any specific company. We say that the non-market force, that’s the CCP, has control of any so-called private sector, so-called companies in their economy so strong that we can anticipate, at any given time, they become de facto state-owned when the situation calls for it.

That’s the first one is the control argument. The second one is the path dependency argument. Exactly as you’ve said, even if we do a thorough penetration testing and this line security perimeters, that is only good for that particular version of the firmware.

The next time when there is a disaster happened at backdoor discovered, accident or not, then we will have to be dependent on the vendor to upgrade the equipment and not to say that 3G, 4G, to 5G. 5G is still evolving. You have to continuously dependent on it.

It makes each operate more expensive, but migrating away, even more expensive. That’s the path dependency argument. These two arguments were the rough consensus of people at the time. The National Communication Commission and National Security Council at the time, basically look at what people had consensus on the street, and basically said we agree with it. That systemic analysis of risk is already done five years ago.

That’s why we don’t have much debate here about the 5G telecommunication network, because no PRC component are allowed in the 4G network anyway, we’re free of the path dependency. Because of that, that is not as hot a top in Taiwan.

Your second question, when I joined in late 2014 there was still Ma Ying-jeou’s presidency, and I worked as a reverse mentor to a minister at the time, Minister Jacklyn Tsai. I wouldn’t say that it is particularly partisan about it, this the public service collectively seeing the occupy as a demonstration, but not as a protest. As a demo, as a demo of this new way of governance actually gets consensus more effectively.

I personally trained maybe a thousand public servants in the 2014 and ‘15 era before the understudy minister become a real minister. I think this way of working, which is pretty-well institutionalized now, using a lot of regulations on, we’ll carry on no matter what.

This position of taking all the sites and getting new values out of existing positions I think will remain, and no matter who become the Digital Minister, they can only add to this system without subtracting to it.

One more question from a little bit the other side. You are very enthusiastic for democracy and freedom, but do you sit in this moment down at the portrait to Chiang Kai-shek? How get this together, and what mean you and the young general about the time of Chiang Kai-shek dictatorship here?

First of all, I believe this is Dr. Sun Yat-sen, if I am not mistaken. Dr. Sun Yat-sen is, as you know, a radical revolutionary, all for common ownership and all sorts of economic policies that are heir to a great social innovator, Henry George, as pictured here.

There really is nothing at odds between the work we’re doing and Dr. Sun Yat-sen’s idea of a transcultural republic of citizens, which describes the name of the country. Basically, a transcultural republic of citizens means that we have many different cultures here. We’re trying to be across cultures, and also taking the positions of all the cultures, to form a new culture.

That is the original explanation of the word republic of citizens, or Mínguó, as explained by Dr. Sun Yat-sen in 1916. There’s nothing at odds with Dr. Sun Yat-sen’s vision.

Thank you.

How is Taiwan creating the data privacy?

Very carefully.

Good.

Yes, we are getting GDPR adequacy real soon now. That should say a lot about our way of looking. We think that data is the beginning of a relationship. It is always relational. There is not only privacy, as in only pertaining to one individual.

It is also about how people together can form data coalitions, and eventually, data collaboratives, that can bargain together with the surveillance capitalism. That is the basic idea of what we call data dignity here.

We take pains to institutionalize – like the Presidential Hackathon and so on – creative ways to make the social sector collectively bargain on the data that they produce by the amount, instead of individually bargaining with the multinational corporations, which is like the laborers before labor unions, always at a losing proposition.

For privacy, the GDPR adequacy is just the bare minimum. We’re also looking at innovative ways of privacy-enhancing designs, like open algorithm, like split learning, and all of those new trends that enhances privacy by getting people govern more in the data-oriented future.

Once you participate and become a co-governor, people will then be empowered to produce more data that are prosocial, instead of the data that are purely being exploited outside of their originally-intended use.

That’s the broad brushes. I’m happy to provide finer details, but I think data relations, instead of data as extractive oil, which is the worst metaphor ever…10 years ago, people used that metaphor only because, like oil, data was difficult to extract.

Otherwise, they have nothing in common. Then come deep learning, and insight extraction is suddenly very cheap. Data mining can be now done automatically with unsupervised learning and other deep learning tools.

Data and oil now has nothing in common. Our language now reflects that reality of data as a relational thing, instead of an asset that could be exploited.

China is using big data in a big way for surveillance. Are you also using big data for all these…?

Yes. When we say transparency here, we’re always saying that the people altogether forming data coalitions are making the government transparent to people. Of course, the PRC is now building ways to make people transparent to the state, which is the same word, but opposite directions.

This AirBox example is a case where the data’s obviously pretty easily…Or not easily, but it’s retainable on an individual basis. What’s your thought about how…Or maybe you’ve actually worked on specific ways to actually retain data so that these coalitions can be built in contexts like online types of data and things like that.

Yes, a lot of it, actually. The data collaboratives in Taiwan encompass all the different stakeholders, and the environmental data, which is easily explained, is just one of it. We also have, for example, a collaborative approach of using data – again, on the environment side – using computer vision to detect marine debris and predict their flow.

It’s inherently multinational. You have to participate with all the Indo-Pacific partners to solve marine debris not by cleaning our own sand beach, but rather by predicting the flow. For things that are more social, this is an interesting case.

All of this is, by the way, from our Presidential Hackathon. The idea of combating illicit financial flows by analyzing for all the publicly-listed companies their transactions in public. They have to declare, for example, their selling and buying of land and things like that.

When combined with land price data, with environmental data, with GDP growth data, and all sort of data, people can train a machine learning algorithm together with all the data controllers to predict how likely some company’s engaging in shell company behavior for the next quarter.

They really got this trophy, and the trophy is a presidential promise, saying that whatever you’ve done, we will do it on a national scale within 12 months. Now, we have to work with all the data controllers, because this is very private data, like transactional data.

They must never leave the data controller, but we can share this training algorithm so that they build the machine learning network by themselves. It’s a little bit like, when we’re talking, we’re sharing the topmost of our awareness, our mental workspace.

Actually, nobody remembers the first day when they learned how to speak or the first step that they took walking by themselves. It’s because our neural system also discards the raw data and only retain the wisdom or the abstractions.

By co-training, using split training, only those abstractions, we can compromise no privacy, no personally-protected data, but then together, learn something about potentially illicit behavior. This is maybe one of the best cases for this kind of thing, because when done wrong, it could justify very easily surveillance statism.

If people participate voluntarily in this approach and form a data collaborative to collectively predict illicit financial flows, then obviously, it’s a net gain for the society. We’re treading very carefully here to find a way that is democratically governed, and people, when they don’t feel like it, they can opt out of it and take control of their not only private data, but also, social data as well.

Can I just ask one very quick follow-up?

Yes.

How did you convince the government to…I apologize if you touched on it in the beginning, but these are pretty radical initiatives. I just wonder how you are able to get them through the bureaucracy.

My theory of change, very simply put, is three different axes. For the higher level of public service, what my way of saying is that it increases social participation, so we can mutually accountable. They understand that this means that it’s less risk for everybody involved. It’s risk-reducing.

For the mid-level, what we’re saying is that, previously, when things go wrong, the minister always blame you, and when things goes right, the minister get all the credit. That’s not fair. You should get some credit.

The mid-level public servant who actually propose this in the Presidential Hackathon gets the trophy. The trophy is actually a micro projector that, when turned on, projects the image of the president handing the trophy to them.

It’s really good credit for any public servant, at any time their director-general say, “Oh, but we don’t have budget for it.” They just summon the president, and there is budget. [laughs] For the middle layer, it’s about the credit, and also the gratefulness from people, when people see face-to-face what they’re solving on.

For the front-line career public service, what we’re saying is that, if you automate away these chores that you don’t like doing, anyway, everybody become more creative, and you get home sooner every day. It saves work.

Reducing risk, improving credibility, and reducing work. We cannot improve on all three in each of our endeavors, but we take pains to not sacrifice any one for the other two. It’s always Pareto improvements, and that is the theory of change.

I was just going to ask, really briefly, following up on your previous comment, you mentioned about essentially how constituents feel when they have access to make choices in ways that are important to them. How do they feel? This initiative has been going on for how long now?

For three years now.

For three years, and so I assume that you are able to generate some kind of survey or paint some data on how folks are perceiving the initiative. How are they feeling?

It differs by age group. The more people participate, the more curious about public service, about sustainable, and so on, they feel. The participation peaks, there’s two peak, around 65 years and 15 years old.

These two are the age groups that engage the most and feels that they got most out of it. I don’t have any working theory, but anecdotally, my hypothesis is that they have more time on their hands. That’s the first one.

The second is that they care more about public welfare, like Climate Change and so on, and less about private benefits, because they both care more about the next generation. These two, and also for the 15-year-old, because they cannot vote. This is the only way they can participate.

Whatever the 15-years-olds are doing, the 65-years-olds are very eager in helping to implement that into reality. We get the most positive feedback from these two age groups. The middle-aged groups, the age group between them, are somewhat more cynical.

They feel that this may be a way to solve the lower-hanging fruits, where we sacrifice no people’s interest, but sometime, they feel that structural changes cannot be effected by gradual, just looking systematically at, as I was showing, systematically looking at the lower-hanging fruits.

They still feel that these issues need to be tackled in a structural way. We merely say, we surface that these are the issues that divide the society, but we do not possess the capability, in this co-governance approach, to make a unilateral decision that sacrifices one part of the population for the benefit of the other.

This is not what it’s designed to do, and so they accept that limitation. Of course, they’re not as eager or participatory than the other two age groups. There’s no real difference between, say, the municipal, the rural places, and so on, which means that our inclusivity is done pretty well.

Thank you.

Before you could read the phrase, “education for global citizenry,” and I just wonder, what is the content, the meaning with it, and how is it done? It’s unique.

This is one of the sustainable targets. There’s 169 of them, and is target 4.7. In our basic education system, starting this year, we are shifting from a standardized, answer-based, very East Asian education system that somewhat encourages core competitiveness to a new curriculum that emphasizes core competency, which is intrinsic motivation, rather than extrinsic motivation.

The students are encouraged to design their own capstone projects and design their own curriculums. The role of teachers switches from being the bearer of standardized answers to a co-learner. This is a real sea-change in basic education, taking into account 10 years of experimental educations learned with them into the basic education system.

This is the first curriculum that is co-created, not only by parents, but also by students themselves. It’s very democratic. The sustainable development is highlighted among all age groups as maybe one of the most important thing in education.

The other being inclusion and innovation. Sustainability, inclusion, and innovation are the rough consensus across generations of what a lifelong learner should focus on. That is how we are designing our curriculum, so that any school can really choose to use any way, but preferably open source hardware and software, to teach about, say, media competency.

When we say media competency, we don’t mean media literacy only, which is about being a reader and a viewer. The broadband as human right means everybody is a broadcaster. It means that they have to also participate in, say, fact-checking, in, say, source gathering, and developing their points of view.

Basically, democratizing journalism, so that people can become responsible media partners. That is, again, also part of lifelong education.

Sorry, I’m just following up. How effective is that? As a former teacher, and in the UK especially, the curriculum that has been designed by parents has actually been failed.

We’re not saying designed by parents. We’re saying designed with parents. [laughs] Important difference. The alternative education system, the experiments for the past 10 years, some of them are curriculum designed by parents. Some worked, some didn’t.

We look at what worked and took what worked into the basic education curriculum. We look at what didn’t work and put cautious stories about it [laughs] in the basic education curriculum. The relationship between the Asia’s most liberal experimental education system, all the way now to the university-level, legally, 10 percent of students can participate in this designed-by-parent system.

The other 90 percent is still designed with parent system, but it’s the research and development relationship. We took what’s worked here into the basic education. I think we mitigated against some part of that, but thank you for raising that point.

That’s exactly why we need to include all age groups and experts into the curriculum design committee. Other questions? Well. OK.

All right. Well…

Sure, yeah. It was just some remarks from the group. On behalf of the international press group, we want to thank you not only for your time, but some really fascinating work that you’re doing. I think we were all blown away by not only what you’re doing, but certainly the boldness in which you’re doing it. We’re very moved by it, so thank you for your time.

Thank you.

Thank you for sharing with us and willing to answer our questions so effectively. We really appreciate it.

Thank you, and thank you.

If I may have two more minutes, I would like to conclude this meeting with a prayer. [laughs] This prayer is literally my job description. If you saw my Twitter, you already saw it, but anyway, this is my job description three years ago, when the HR asked me, “There was no Digital Minister in Taiwan. What would you do?”

I’m like, “Oh, just target 17.18, 17.17, and 17.6,” and they were like, “Nobody really memorizes Sustainable Goal targets, Minister. You have to say something in plain language.” I transliterated that into plain language as a prayer three years ago as my job description, which I will share with you. It 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. And whenever we hear that a singularity is near, let us always remember the plurality is here.” Thank you.

Thank you.

Thank you.