We have 40 minutes at least.

Awesome. That’s perfect. Cool. Should we start right away or?

Yeah, please.

Where are you sitting now? What’s the background?

I’m at the moment in the Social Innovation Lab. The lab is the place where people come to me in my office hour because this is a Wednesday. From morning onward, people can just walk in from this door and just have a chat with me for 40 minutes. The only requirement is that we publish the transcript just as ours now to the Internet after 10 days of co-editing.

Were there, at any time, any bad experiences with this openness? Anybody tried to harass you or any of these things, or was it always easy?

There was someone who was here talking about his invention very passionately, a perpetual motion machine. There was someone who was here to talk about building an embassy for the aliens.

They were pretty reasonable by the end of the conversation because I sympathized and also offered constructive ideas of how to turn their ideas more public. I guess, it may be tense at the beginning, but it’s always good at the end.

It’s just surprising to see that a minister actually can be so open and can have an office without closed doors and guards outside or anything.

We’re just taking the walls out. In a week or so, this will be like a park. Everybody just walks in from the street.

I sent you some of the questions beforehand. I think I will start with them, and then we see where we go from there.

You don’t have to stick to them. Feel free to start with any question, really.

You went from hacker to Apple consultant to minister. What surprised you the most on that journey?

I’m most surprised by the innovation from the public sector, that the bureaucratic staff is so innovative. We wouldn’t know from the outside. Actually, within the public sector, there’s many people who are as innovative as social sector and private sector innovators. It’s just because in Taiwan, we have this idea of public service anonymity.

They cannot actually talk about their innovations publicly without the OK from their minister when the ideas are still on the drafting stage.

That’s the main perceptual gap, is that the public thought the public sector is always very risk-averse. Once I start working with the public sector around the end of 2014, I met a lot of very innovative people.

How to free up the innovation in public sector?

Two things. The first is through radical transparency. In our Freedom of Information Act, I’m sure in US as well, the government is obliged only to publish after they make a decision that affects people. During their drafting stage, where people are still brainstorming within the ministries, they’re not required to publish anything at all.

All the meetings that I am a chair, even the very early-stage brainstorm meetings, we publish every transcript after 10 working days of co-editing as well. It’s the same as my office hour, except it’s 10 working days, not 10 days. [laughs] Other than that, it’s the same.

The people can actually see the innovations when they’re still being brainstormed. They get a credibility. They get a credit because they can see also sometimes through video conference. Every other Tuesday, I go to places where it’s hard to reach Taipei by high-speed rails, because if I only sit here, it privileges the side with public transport and the high-speed rails.

I go to the sides of Taiwan, the indigenous areas, the rural places, the offshore islands, and have a real meeting with the local stakeholders also, but with the public service here in Social Innovation Lab. We connect with Zoom so that they can see eye to eye and have a conversation.

My point is that radical transparency, whether through video or through text, builds credibility and shows that the public service trusts people to have a connection, and so they get a credibility. I absorb the risk because I am the chair.

If I publish and there’s a backlash, I, of course, will be someone to blame, because I’m the only one traveling to the local vicinity, not the 12 ministers here in Social Innovation Lab. If the people there are very upset, you cannot hurt people over the Internet, so I’m at risk again. By absorbing the risk and spreading the credit, that is the way to make the innovation more widely known.

That sounds pretty interesting. What was the outcome of this process that you are so far most proud of?

One of the very interesting result is that in the offshore island there was a nurse, a local clinic, that was not trusted enough by the local people. Always, they require flying helicopters to send their loved ones or their families that are sick to the main Taiwan island, even though it may be dark or it may be raining.

Through this approach, we discovered the true reason is that the telemedicine at that time did not allow a local nurse to act with consultation with a telepresent doctor to give diagnostics or give medicine or give medical treatment. They always have to have a local doctor but a local doctor may not be in the same specialty. Because of that, it decreases the trust.

It’s the same in the south of Taiwan, in Hengchun, as it is in the offshore island of the Orchid Island. Through a Presidential Hackathon, they banded together and developed in three months a new system that ignore existing e-signature law, and telemedicine regulations, and everything, and made it work.

They got the winning trophy from the president, which is a projector that shows the president giving the trophy. We changed one law, five regulations, a lot of personnel, a lot of budget, and now more than 100 indigenous and offshore islands enjoy telemedicine before sending helicopters and the nurses can act based on collaboration with remote doctors.

How is the process in your ministry now faster – it sounds, at least, faster – than it usually would be in other ministries? Can you compare that?

The fun thing is I don’t have a ministry. My office is literally all the 32 vertical ministries can send one delegate to my office. My office is entirely horizontal. Theoretically, I can have 32 colleagues. At the moment, I have maybe 20. It means that not all ministries have sent delegates. The Ministry of Defense never sent anyone [laughs] for some reason.

There is a reason.

What other people facing was the Ministries of Education, Interior, Communication, Culture, Finance, Law, the usual suspects, they send people to my office and make sure that all the innovations that I work is co-created by the people who are sent to my office, together mobilizing their own ministry’s resources.

There’s no comparison of my ministry versus other ministry because my ministry is at least 12 ministries.

It’s not so much a competition between ministries but more of a consultancy for other ministries in a way.

There is because it’s a rotating delegate. I limit each ministry to send only one person. There is some competition. In the Foreign Service, at any time there’s a handful of people willing to join my office.

Maybe there’s some competition going on there. I heard there’s a queue. If a dozen people from Foreign Service join my office, then we become a section of the Foreign Service. The same holds for other ministries as well, which is why only one person.

You have written in, for example, “The New York Times” about the digital decision-making platform, Polis, and how it works. I won’t go into detail here because you already detailed it. What are the limits to that system?

The main limit is that you need to feel more or less in the same space. It’s based on open-space technology, co-presence. You need to feel that you’re part of a polity. That’s easier in Taiwan because broadband is human right, because from north to south by high-speed road it’s an hour and a half.

Even though 23 million people, 10 million on the participation platform, by writing the same word, we still more or less feel that we’re in this together, that we can live with it even if the outcome doesn’t go all the way to our position, and that, when we share common values as shared by the resonance chart, we feel that there is a polity instead of five different polities joining together.

In a federated polity, this will have to be designed differently. In a single polity, usually, municipal governments in any jurisdiction is very easy to rule it out. People in a municipality identify as a polity.

Let’s say if you take any bigger state that is very divided at the moment, for example, UK. You have the one part of the people who feels like they really want to leave the European Union. Then the other part really believes that should not happen, for example. Would that work there or is it too difficult?

Right. The point here is to look at large ideological binary decisions and separate through mechanism design into five or more, what we call, opportunity of overlapping consensus. I’ll use a example. We had a referendum here. One side is arguing about the equal rights and duties of people regardless of their sexual orientation to marry.

The other side is arguing that the kinship family relation is what marriage is about and it builds connection between two families. They think that different sexual orientations results in different kinship orientations as well. They don’t accept this as a civil code wedding of the families. These two sides are polar opposites.

When we design the participation to ask specific questions, for example, whether it’s OK for women who are not wed to have artificially inseminated children, this single aspect, not everything else. Then the stakeholders are children who have not been born. It’s in the future.

People can imagine the different societal configurations where the society is more supporting of such families and a society who would discriminate against such families and things like that.

When the stakeholder is yet to be born and we’re specific enough and not talking about ideologies, we see both sides coming to common understanding. Dividing it into very specific cases like a single mother that talks about her child’s everyday situation, that binds everybody together.

The idea would then be to go more into detail and leave the big binary question out for a while.

That’s exactly right. Once we understand that for these things we, at least, have something that everybody can live with, not necessarily prefer but can live with, then it builds the polity. People feel that we’re in this together a little bit more, a little bit more. At some point, you will be able to tackle the more divisive ones.

Do you think that could work in big states as well or would that be too complicated?

No. It’s not about population. Even if Taiwan’s population doubles, as long as the same feeling of polity is the same, this is still going to work. Even if, for states with population fewer than Taiwan, say five million, if they’re seeing themselves as five different countries conveniently banded together, then that’s probably not going to work.

Where did the idea come from to use this specific tool that you found in, I guess, Seattle?

The civic tech community try it out first. I think it was first used in the JavaScript conference in Taiwan or something like that. The tool is generally useful. The civic hackers started deploying it and working with popular media such as “Apple Daily,” who used the same tool for deciding, after there was a random killing, whether the random killing has anything to do with death penalty.

That was the case where the public media used the tool on. After a few cases in the social sector and in the media, we’re like, “Yeah. This is such a good tool. Maybe we should convince it to be open source.”

Now, it’s open source to be hosted by the government. We still contribute back all our changes like the automatic bilingual translation and so on, to the wider open source community. Gov tech is built on civil tech.

Let’s say this method becomes mainstream, what would that do…?

It’s already mainstream here. [laughs]

OK. In a wider circle, I mean. Doesn’t that also change the idea of politics? Now, for example, in Germany it’s a way you vote for a politician. The politician has quite a strong opinion on some case, be it let’s say, gay marriage. That’s what you vote him or her for. Then he goes out and tries to press through this opinion and make it into law.

The Polis system that you just described, it sounds a bit more different than having representatives who press through a topic and more like a politician becomes more a person that organizes different opinions to fit together.

Yeah. We’re mostly facilitators. The point here is that instead of allowing people only one bit as in referendum or three bits as in election, of upload of bandwidth, there’s far too little bandwidth for people to upload. Their collective intelligence must be compressed through this very linear lens.

This is almost like if you walk into Blockbusters. I hope you still remember there were video tapes and things like that that you go and rent to view. It’s sorted in five different categories like drama, and, I don’t know, horror, and things like that.

Of course, my taste cannot be captured in these categories but I’m forced to walk to a category and then start browsing. This is like political parties. It’s forced categorization of a multi-dimensional opinion into the limited shelf space.

Now, if you’re on Netflix, for example, and you watch a few films and give your ratings and so on, Netflix use exactly the same algorithm as Pol.is, such as K-means clustering, principal component analysis to sort you into the dynamically generated cohort of people who prefer a similar taste but there is no word for it. Do you know what I mean?

Yeah.

It’s not representative. Representative means you use one label to represent a broad swath of people. This is more re-presentational. It presents again your original choices along with everybody else’s and use your own words to determine the agenda. This crowd sourced agenda is re-presentational. It’s not representative.

That sounds pretty interesting, actually. [laughs] Wouldn’t that make traditional political parties somewhat unnecessary?

I would argue, though, that it’s two-fold. First, it makes it possible for the legislators if the parties are forgetting the seats in the parliament. The legislators, they are more informed in design thinking terms. They can focus on the development and delivery of a regulation.

The define stage and the discover stage…Discover is that people typing in their sentiments, the define is facilitated conversation. The agenda is basically provided by the crowd. They can deliberate with much more balanced options to choose from because people have settled on a common value that people can live with.

It makes their job easier and also enable them to produce more high-quality work, make their work more satisfying, because if they have to use old town halls or other ways to do the discover and define process, it’s very time-consuming and also it’s impossible for them to avoid this principal agent problem even if they want to.

If politicians are now more facilitators, they also could be just desk officers, but they wouldn’t need to have necessarily a party association, do rallies, and all this kind of things.

That’s right. That’s the campaigning part. I just talk about the legislative part where political parties may still be very useful because they represent different angles of facilitation and so on. The campaigning, which is more like game theory, you have a single simple game of voting and you try to win this game so you develop game theories, strategies, counterstrategies, and so on.

What I just proposed is reverse game theory, which is mechanism design. You start with the outcome, which is people coming together, forming a polity, getting to know each other’s true preferences and common values, and then you design the game in reverse to effect those outcomes.

In mechanism design, they could be participatory designers. They don’t have to spend a lot of energy on campaigning and largely canceling each other’s efforts out. It may make a different brand of parties that use, I don’t know, random controlled trials or whatever, new methodologies of mechanism design, and compete on the ways of mechanism designs instead of keep playing a plain and simple single game.

It actually could free up a lot of energy that is now used for battling each other from the trenches of party politics.

Half of the people fell they have lost after each referendum or election, or everyone in certain referendums.

Does a country need a crisis to implement a system like this, or would that also, in your opinion, work without a defining crisis first?

It needs a legitimacy crisis to research a system like this, but it doesn’t take a crisis to deploy.

We had similar ideas in Germany a few years back. There were the Pirate Party. They came up with a system called Liquid Democracy. It failed spectacularly. Why did it work in Taiwan? Why did it fail in Europe?

Liquid Democracy is still representative democracy. It just says that your representative can change based on issue and based on your mood. [laughs] As mechanism design, it doesn’t really incentivize me to reveal my true preferences, and still incentivize me to essentially gain the system. All the papers are there. I don’t have to quote from them.

I would say that having the energy to do mechanism design is fine, because I read those liquid democracy paper, too. When I designed the system, I know which pitfall to avoid. [laughs] It will always contribute to the next generation of mechanism designers. That’s how political science or, indeed, any science progresses.

How do you think is politics, apart from Taiwan, going to look in 10 or 15 years? Do you think those ideas that you just developed in Taiwan, they have the potential to also change politics in other places or?

Yeah, of course. Yes. We’re seeing a lot of adoptions actually. Usually, I would be honest, usually in municipalities, this is what we’re looking at. After 15M, Madrid become the heart of the council in all the different participatory design in Barcelona for example. For a while, Reykjavik, we learned a lot from the Reykjavik Pirates.

We just had a new party formed yesterday called the irresistibly happy party. I think it’s very much shaped like the best party of Iceland a few years back. It’s anarcho-surrealism. [laughs] All the different tries, I think, works very well, as I said, when the people feel commonly as a polity.

It could be a country. For example, Singapore is also trying to this out. Because, again, people feel as a polity regardless of their four languages, which of them they’re speaking and things like that, we haven’t seen significant adoption on a multi-polity confederation or a federate level.

Are you in contact with people from Estonia about this? They tend to be the most digital country.

Yes, of course. They also informed us about making the infrastructure system shared source, at least open for public to see in creative common terms that’s nonderivative or at least noncommercial licenses. It’s useful to have this kind of public-private partnership with social sector looking at a source code to ensure legitimacy.

Our deployment of distributor ledger infrastructure is also heavily informed by their use of ledgers in their cross-ministerial communication. It used to be called X-Road. Ours is called T-Road and so on. We’re in constant contact.

Let me just look at my questions. I also wanted to talk with you about China, because now when we talk about Estonia, it’s like there’s different models on how digitalization can shape a country.

Whereas Estonia digitalizes traditional political systems and you have your new approach with Polis, China seemed to go in a complete different way with also a very digital state. Do you think the more liberal versions of digitalization of politics are going to prevail naturally? What’s your take on that?

Whereas Estonia or Taiwan talks about radical transparency, we talk about that a lot. We mean making the state radically transparent to the people. When the PRC talks about transparency, they mean making their citizens radically transparent to the state. Same word, very different directions, exactly as you said.

It’s arguably the same set of technologies. Our technologies serve as the great amplifier no matter which direction you’re going. We’re currently running very fast and in opposite directions. [laughs] I don’t know if it’s circular. Maybe it wraps around, but I don’t know whether that space is circular. [laughs]

At the moment, there’s no issue about prevailing because, really, it’s, I wouldn’t say entirely opposite, it’s orthogonal and sometime opposite. We’re not really interfering on different implementation of the same idea. It’s the same implementation of very different ideas. I don’t see a competition there [laughs] because it’s, frankly speaking, just two very different systems.

I was just wondering, you wouldn’t think that any of those systems will evolutionarily, naturally prevail and be the model that more and more countries adopt? You know what I mean? That other countries would see “OK, maybe we need to digitalize a bit more,” and then they look around and see different systems and also see maybe the Chinese way and also see the Taiwanese way. Would you see…?

It depends on the fundamental relationship between the different sectors. In Taiwan, we always say a journalist’s word is worth more than a minister’s. If you start with that basic assumption, absolute freedom of speech media assembly, then, of course, your innovation is going to look like ours, which is transparency of the state to the people, that the government trusts in the people.

If you start in a jurisdiction that says the minister’s word is always better than the journalist’s, and if the journalist’s word disagree with the minister’s, the journalist is forbidden to practice journalism, then, of course, you’re going to end up a with a system to optimize with that particular value.

Across jurisdictions, when they’re shopping for governance technologies, naturally, they’re going to only use the ones that are compatible with their current social configuration. Nobody starts from a blank state.

I think what we can do is just to show the people who value their basic freedoms that it’s actually as effective and always generates more trust if you care about that. If you care only about efficiency and not at all about trust, then our approach is not going to be attractive.

Is the China topic a topic that you would deal with in terms of digital power? Is that something you’d leave to the defense ministry people who wouldn’t show up to your meetings?

Five years ago, we made a collective decision after a systematic risk analysis that our 4G network, which was just being built at that time, we are not allowing PRC components in the core 4G infrastructure anywhere into Taiwan telecommunication.

We based on that two assessments, which I helped facilitate and distribute but it’s not my point. It’s more like the common understanding of people.

The two things is, first, that there is no such thing as a pure market player from the PRC. At any time through the party branches, the state actors can force any market player to become a state-owned player, de facto player.

The second is that if we’re depending on the PRC components in critical infrastructures, in telecommunication, then every time it upgrades from 3G to 4G to 5G, it’s going to be path dependency and we basically become reliant on the upstream patches.

There’s no amount of cybersecurity test that can test for future versions of their firmware in terms of a disaster recovery and things like that. We made the collective decision saying, “I’m sorry, but this is not a matter about the market. This is more about a national security issue.”

That discussion I did at least helped distribute and facilitate, but that may not be the PRC issue you have in mind. This is more about how to do system analysis.

I’m not sure how the topic is talked about in Taiwan. In Germany, the mood is this: You see how China is getting ever more powerful in the digital space, be it with new social networks like TikTok, be it with prevailing in telecommunications industry or semiconductor industry. People getting a bit worried about where will that digital power lead to?

We don’t quite feel that. That’s because two things. First that we know as a fact, because of TSMC and friends, that in terms of semiconductors, we’re at least five years ahead of the PRC. They may catch up in a decade or two, but they said that a decade ago. We still have a very comfortable lead in this kind of research and development around the semiconductor supply chain.

That’s the first thing. The second thing is that we had a taste of what kind of authoritarianism was like in PRC. I still remember the martial law. It was not long ago. When I was a child, my parents were journalists. They also risked being censored or disappeared if they publish anything that disagreed with the ruler.

Because of the martial law and the fact that people who are of my age and people older than me all remember the martial law, we are not at all impressed with any efficiency gained by what they call Democratic Centralism. It’s a Leninist name…

Well, Taiwan was like that. Now we are not willing to trade a little bit of marginal efficiency – even that very doubtful as their GDP numbers – for this kind of authoritarian control, because we had a taste of it.

Have you ever heard the term digital Iron Curtain? Is it something that you see on the horizon coming up? What’s your take on this?

I don’t really know what that means, actually. It could be used to refer to the fragmentation and balkanization of Internet, of Internet being no longer being inter, but rather just net. That’s one use.

Another use is around the technologies stack in ICT communications, such as 5G. That term also has been used on that. It could be also used in a cybersecurity, like proactive defense way. I’ve seen this used in various ways. I don’t know which one you’re referring to.

I think it gets clearest between China and US. When the US says, OK, a US company shall not buy these components of Chinese companies, and then Chinese companies boycott American companies on the other side.

Then you have a tech divide, and then also next to the tech divide, you have the Chinese firewall. Then you have a divide going through different industries, and sectors of technologies. That’s what I meant or what I thought people mean when they talk about a new Iron Curtain or a digital Iron Curtain.

I see. We are very familiar with the Great Firewall, of course. I was an early translator and advocate of the Freenet technology, which is precursor to the torrent technology and all the technologies like GN and so on that circumvent the Great Firewall.

In fact, the Great Firewall became so great because it’s like a generated adversarial network where the people try new ways to break it, so they have to deploy a lot more resources on deep learning and everything to make those walls still hold.

The skirting or the firewall, as I mentioned it, has been building ever since the Great Firewall, or the Golden Shield project, and all those projects were being considered because they specifically wanted a different norm around the Internet that is less inter and more intranet.

I think it’s around the turn of the century, actually 2002 to be particular, that I started to work with people who had GFW circumvention needs like human rights activists and so on. For me, this is going on for 17 years. It’s not a new thing.

When people started talking about 5G and risk, we’re like, we had a discussion five years ago. It’s a different sense of time for me. When you refer to these things, for me it’s not a new thing. It’s always been there. If you call it the curtain, that’s fine. We call it the Great Firewall. It’s always been there.

Hong Kong has been a lot in the news recently, and I guess it’s also a topic in Taiwan. Is the battle for Hong Kong also a digital battle?

I think a lot of people in Hong Kong are, of course, using Telegram and other technologies designed for end-to-end encryption and real-time mobilization as we did five years ago in the Sunflower Occupy.

The main difference is, of course, at that time, the Sunflower movement is basically convened by around two dozen NGOs, each specializing on the labor, the human rights, the environment, and so on, different aspects of the CSSTA. They are still leaders. It’s just multicentered. It has like 20 or so super nodes, so to speak.

The Hong Kong one is maybe in the B. Walter spirit, there is maybe around 2,000 to 10,000 super nodes, which makes it impossible, without shutting down Internet, to look into the whole picture. There is no such thing as a whole picture anymore.

It’s a memetic protest. It’s a movement that doesn’t need the countable super nodes. It’s a purely collaborative and less cooperative way of organization.

That is very much worth studying. Back in June, I wrote on Twitter around I think it was June 16th, the first large parade, that I think this is a chance for the PRC to listen to the people and advance the system or promise that they made in 2015 on peace, on human rights, and the rule of law.

I think the Internet, as we both have discussed so far, can amplify these values, or it can amplify the counter-values of this.

My last question for you: How will the rest of your day today look like?

Its 7:00 PM now.

How was your day?

I had eleven 40-minute meetings today including yours. Today I began with a lecture and interaction in Argentina. Their, I think provincial minister of education and I had a really good talk and with interaction with the local entrepreneurs.

Then on 9:00 AM Taipei time, a student of Dr. Otto Scharmer, the U Theory person, came and we co-designed the social financing workshop to mobilize the family investors to further the sustainable goals.

On 10:00 AM, we had a conversation with some community called Redroom, which is a community enabling expats and people just traveling and working in Taiwan to make a better bilingual culture. I invited them to become incubatees of this space.

Then in the 11:00 AM, I had the talk with somebody who use AI to generate GRI-compatible sustainability reports for public listed companies purely on quantitative data and the mood you want your shareholders to feel, which is a very interesting application.

After that, I talked with Dell Technology. They want to emphasize on their upcycling ways and their carbon reduction. They also want to talk about how to change their surveillance camera technology to be more human rights preserving in the Taiwan’s atmosphere. Smart citizens, not smart cities. We had a philosophical conversation.

After that, I had a conversation with a PhD student on social anthropology who really wanted to know what’s the mechanism designed to enable people to contribute to the common good.

After that, I had a conversation with Sense. It’s a Asia Social Innovation Award winner and also Asia-Pacific Social Innovation Partnership Award Winner. I gave their award. They use eye tracking, like Stephen Hawking, to enable people to communicate quite easily, but they have some regulatory issues.

One of them is about to qualify to reimbursement, they have to simulate a mouse left button and right button. The regulation is very old. Nowadays, nobody has mouses anymore.

Yeah.

The regulations really need interpretation, and I promised that I will look into the reinterpretation. Actually, I think it can be done without any regulatory change. It’s just an interpretation of a touchpad tap as a mouse left click. That’s all they need to do.

Then after that, I had a conversation in the United Nations Innovation Conference with the Stallman Foundation for Freedom. That’s finally, the interview with you. That’s what my day looked like.

That’s some pretty full day, I’d say. Is every day of yours like this?

Every Wednesday is like this.

OK. Thank you a lot. With the transcript, we will do as we said. I think you’ll send me a link or something.

I’ll send it for your co-editing. I publish after you do. Feel free to edit to make clarifications, or add links, or whatever.

OK. I’m not entirely sure if this will be online or in the printed magazine. Does that matter for you how fast it will be or is it…?

No. Usually, we publish after 10 days, but if you need time, you just let me know when. As long as it’s some time before 2030, I’m sure that we can accommodate.

Awesome. Thank you a lot.

Thank you. Cheers, bye.

Bye.