Thank you for talking. I’m sure you’ve talked to a lot of journalists recently. [laughs]

I just went back from Berlin and Amsterdam, so a lot of journalists there, for sure, but not yet locally.

Really? I’m based in Beijing. I’m not really local. I came here because I thought disinformation on the Internet was a big problem in Taiwan, but maybe that’s not true.

You’ve read the two CPJ reports?

I saw the CPJ report. Yes. I thought it was a big problem, but then I’ve been here a few days, and I feel like people are actually pretty confident it’s not a huge problem.

It’s less of a huge problem when compared to the previous election.

Yes. You think it’s, compared to last year, better?

Yeah, definitely.

That just seems also because Taiwan has been dealing with Chinese interference for so many decades. People are pretty savvy about what information they get. Is this whole problem overrated? Are we talking about a problem that’s much smaller than we think?

I would say so. The magnitude of the problem can be seen in two ways. One is in the raw attempts and the creativeness of those attempts like spam email. It used to be a large problem. We talk about it less now. It doesn’t mean there’s no spam.

There may actually be more people putting their time and energy into improving ways to spam and scam people. It’s not, in raw numbers, less people practicing it. That’s one side.

One is, you just said, the confidence level that the society feels that whatever new mechanisms they invent and throw at us, there’s some way for the liberal democracy to respond in a timely fashion.

In terms of confidence level, yeah, the Taiwan society is cautiously optimistic compared to the previous election.

What’s changed from last election?

How do I start? The election this time is not tied to referenda. That’s obviously one of the very large divisive topic off the table. The election this time is also defined by three candidates who have, broadly speaking, compatible views of Hong Kong in similar issues, which is quite different from the previous election. That’s another point of divisive removed.

Just from the structure of the election, there’s, what we in the cybersecurity call, less surface for attackers to sell the spot. It doesn’t mean that there’s nobody that wants to sow discord. It means that the topics on which they can sow discord is a little bit more limited compared to the previous elections/referenda.

Do you think there’s much changed in terms of the actual information environment, in terms of the platforms, in terms of the companies that are involved?

The platforms have signed this counter disinformation self-regulation policy, which helps. It helps raising awareness, for one. They’ve also taken actions to attribute publicly, like coordinated inauthentic behaviors. Also because of Hong Kong, people in Taiwan have seen how a real operation looks like in practice, on social media platforms, with less self-interest attached to it right.

People become more objective, that people can see when the media publishes something about the Hong Kong issues. People can see with less immediate self-interest in their political ideologies, that it really is an information operation.

What does make you worried? Are there any specific types of vectors that your thinking still has the biggest potential to impact the election?

In terms of disinformation?

Is it YouTubers this time? Is it Deepfakes? Is there anything that you are worried about?

I’m personally not specifically worried about any surface, because what we’ve done really is a rapid response mechanism that reacts to new and emerging methodologies. Deep fakes, if we’re not prepared, would have been one of the differences, just in terms of the expense to generate a convincing synthetic video has dramatically fallen in terms of cost, compared of last year.

On the other hand, as you said, people here have seen 動新聞, the animated news, for quite some time now. [laughs] The very fact that there could be synthetic convincing videos is already in people’s mind.

You don’t need to make a deep fake, a very expensive deep fake. You can just make a 動新聞.

Exactly.

Sorry. This is my first time reporting in Taiwan. I don’t know the environment very well. Is 動新聞 still a big part of what people watch?

Yeah, there’s still various motion-captured illustrations in the media, like short animated videos in their not necessarily news coverage, but in their short films and things like that. It’s not always associated with news, but motion-captured video is a part of everyday landscape.

Were there examples of animated disinformation in the last election?

In the last election?

Or even 2016.

That’s a good question. I would say there’s no media using an animated doll like reenactment of something as the main vector of disinformation, but mostly because people understand that these are synthetic. They’re not really good source materials for remixes. Of course, deep fake means that any of those synthetic video can look arbitrarily real.

Why is a meme so powerful then? As you say, people recognize a 動新聞 is not reality.

It’s synthetic.

It’s synthetic, but people will believe a meme if they see one image.

They won’t neccessarily believe a meme. They do share the meme, different idea.

OK, but they’re willing to share it.

Because it’s fun.

動新聞 could also be fun. It could also be a vector.

These are orthogonal. 動新聞 is one of the ways using motion capture and avatars to lower the cost compared to hiring an actor. That has nothing to do of whether the message is humorously viral or not.

I’m just curious why, if we didn’t see any examples of animated disinformation, why not? What’s the reason?

Because people know that you can synthesize such things and would not share it as if it’s real. It’s much easier for the disinformation makers to share an actual footage and then make it what we call mal-information frame around it.

Like sharing there was a large protest in the presidential office yesterday – “All the media ignore it. It means that the ruling party has censorship capabilities.” – but neglecting to show that this video is actually three years ago. Of course, the media didn’t report about it because that was very old news. This kind of mal-framing is cheaper to produce because the footage is already there.

It’s easier to share because it’s more simple to understand without making any leap of faith. This kind of mal-information, third, you can try different framing and variations and do some A/B testing and see which one goes viral. Whereas using synthetic video, you have to at least rewrite a script somehow. It’s just cheaper, while being more effective.

If this time we’re getting lucky because the candidates are pretty aligned on the big issue, in the future that might not be true. Are there longer-term things that you think the government or civil society here needs to do to insulate Taiwan for the future, or are you confident in the long term also?

That’s two different questions now, actually. If it’s not disinformation, it could be other factors as well. It’s not something that we say, “Oh, the disinformation has been fixed.” Then we pretend that propaganda or information operation or whatever is no longer there.

Of course this whole system is still there. It’s just maybe disinformation stops to be a useful vector. They will then move to other vectors.

Just our relative cautiously optimistic saying that we’ve got disarm disinformation to a certain degree of resilience doesn’t mean that…It’s like we are earthquake-proofing our buildings. It doesn’t mean that it survives a typhoon.

As you said, we don’t talk about much spamming, although of course there is spam.

There are still a lot of spammers.

There’s still money to be made from spam. People will try it. Is there something that can be done technologically or in terms of the system in Taiwan that will make it less of a vector, or are we already doing everything?

Media competency is something that still need to approach more cultures in Taiwan. The basic education media competency website, which you can find easily, it’s a good start. If you look at the MOE website on media literacy, you can easily see that they’re still pretty much targeted at current K-12 level.

For the elderly and elderly of different cultures and different languages and so on, there is still a lack of inclusivity in the shared curriculum when it comes to media competency.

That’s something that I believe the community colleges, the university social responsibility programs, just MOE in general, and also Hondao Foundation or the foundation that works with the elderly, they are all working toward being more inclusive when it comes to media competency.

The website I mentioned is mlearn.moe.gov.tw. You will see that they still are targeted at K-12 and just certain part of the population. We still need to be more inclusive.

Is Taiwan’s market for information working? Is it working in the sense that the best information gets out there and travels the farthest, or is there still something that’s broken in the market, where misinformation or bad information spreads more than it should?

There’s always room to do more. At least the innovation in this area would not need a central authority to approve. The best ideas may not be reached yet, but everybody who try all these various mechanisms don’t need a central authority approval.

According to Civicus, we’re the only jurisdiction in Asia now that is completely open in this regard, that whatever the journalists want to work with civic technologists they have the most freedom to experiment. The other one, I think, is New Zealand. That’s it.

In terms of new innovations in, for example, collaborative fact-checking for the presidential candidate race or things like that, we do have really good room for the civil society to try out those ideas.

Wherever the market is broken, there is also opportunity for people to collaborate and make a better mechanism out of it. I would say that the self-repairing capability in Taiwan is very healthy, but obviously we’re obviously not at the optimal point. We’re not even at the Pareto-optimal point.

Whose responsibility is it to get to the…

Everybody is responsible. We’re a democracy. It’s all the citizens’ responsibility.

How big of an impact has Wang Liqiang had on this election so far?

That depends on how the Xiang Xin evolves. It’s an ongoing thing.

Is there going to be any judgment before the election?

I don’t know.

I don’t think so.

I don’t have any access to the administration. I don’t really know. So far, I think people are still paying attention, not necessarily to Wang Liqiang but to what Xiang Xin and the network is up to.

Is it safe to say that it’s already helped the DPP in some way because it’s another thing that makes people nervous about China?

It’s less about making people more nervous. It’s more about putting on the record something that people have known, sometimes firsthand, for some time. There was no names on it. What this does is that it puts some names on it.

As of the information operations, a lot of people already have friends or families that were being paid during the last election to do information operation things. It’s not exactly a secret.

In cybersecurity terms, how do you view China as a bad actor? What’s your model of Chinese influence? Is it getting more sophisticated? Is it still very obvious what it’s doing? Is it learning? How do you view it? How do you model it in your head?

In cybersecurity, what we’re looking at is, essentially, it’s an advanced, persistent threat. [laughs] You were asking about advanced part. It’s a two-way sign. I personally remember when the Great Firewall wasn’t great at all and it’s very easy to circumvent.

Especially because there’s a lot of people working on circumventing it, the Great Firewall become greater, because then people are paid to coevolve with the circumvention. Both the bypassing techniques as well as the filtering techniques become much more advanced. This is what we call generative adversarial networks. [laughs]

When the PRC try to – I wouldn’t say replicate, but introduce – similar authoritarian installations to other jurisdictions, sometimes the difficulty is that the people there just wasn’t at this constant struggle against people who circumvent it.

They sometimes don’t have the same motivation or resources for innovation because their authoritarian regimes may not rank harmony stabilization – whatever that’s called – as the top priority for the jurisdiction.

I would say the advanced part is not purely technological. It symbolizes that jurisdiction’s relative priority vis-à-vis other things that they could be spending their budget on.

China prioritizes interfering with Taiwan. [laughs]

“Social harmony and stabilization.”

In terms of interfering in Taiwan’s projects, right? It’s also a high priority?

Well, the same Great Firewall has been turned into a great cannon. That’s the “persistence” part in persistent threat. Just recently, it’s been used against LIHKG. Obviously, the main LIHKG is considered by the great cannon wielders, a higher threat than any website in Taiwan, because they choose to attack the LIHKG website.

Which website would they attack if they wanted to?

GitHub, maybe. [laughs]

That would make them a lot of enemies, not just Taiwan.

It’s a large collateral, but they did attack GitHub in 2015. That was some years ago. The point here is that I model them as a rational agent that perceives the existence of LIHKG now, or GitHub a few years ago, as a threat.

Instead of saying that it’s about any particular jurisdiction, it’s about the vector that they believe will destabilize the “harmony.” Sometimes, they made some damage and sometimes, they didn’t.

Their goal in Taiwan is not just preserving harmony. They want to take Taiwan back. They want to use the democratic system to bring Taiwan back. That’s a more complicated goal than just harmony.

It’s difficult to imagine how one can achieve that using purely cybersecurity means.

It’d be a lot easier if they could avoid a real war to the information war.

The critical infrastructure in Taiwan are either not even a part of the Internet or are already so well guarded and penetration tested and really, every day, battle-tested that it’s pretty resilient at this point.

What do you mean specifically?

There is a National Server Security Act that defines the critical infrastructure and the personnel and the budget required to protect that and the federated design and the resilience design that even if one of them falls, it doesn’t propagate back to the other parts.

It’s definitely something to be very diligent about and may be building industry in Taiwan, but it’s very difficult to imagine a scenario where whatever you’re talking about, the annexation is done in purely cybersecurity means. It’s difficult to imagine that.

I mean interfering with the elections, actually, supporting pro-China candidates or leading Taiwan gradually in public opinion…

Without backfiring?

…without backfiring. That’s the goal, I can imagine, right? I’m curious how good you think China has done so far?

As in not backfiring? It’s hit and miss. Some of it works. Some of it, when exposed and attributed, backfires. It’s hard to predict the effect, definitely.

There’s no scenario in which I would imagine that there’s just a series of information operations that are all hit and no miss, that eventually becomes what you said, a kind of democratic takeover. The attribution system in Taiwan would have been completely disabled for that to happen.

The attribution, at some point, always happens. At that point, it becomes exponentially difficult for the information operation after this point without worrying about it backfiring from that attribution.

When you talk to the CPJ, your suggestion about fake news and disinformation is that there should be rapid response, you flood with good information in order to fight…

Funny information. Not necessarily good. Just funny works.

Funny information. [laughs]

Humorous.

I think the issue is if people have lost faith in all information, the don’t believe the media. They don’t believe social media. They know everything has bias. They know everything is, potentially, manipulated. Then, your solution doesn’t really work. It still doesn’t restore faith.

Yeah, but it’s not an either/or. It’s not what you said, an extreme case. It’s like saying if somebody doesn’t have Internet connectivity, then you can’t reach them through social media.

That’s true but even for these people, not to mention that broadband is a human right in Taiwan, but even for those people, as I said, if you target the memes so much so that it’s humorous in their local culture, they will hear it from their friends and families anyway.

That’s how the media competency education works. It’s enabling the local people to become more informed information providers. They’re, essentially, citizen media for their friends and family. We empower them with the same set of tools that as a journalist, one needs to learn about source checking, about fact-checking, about balanced viewpoints, and narrative building, and so on.

Then, once they’re equipped with those tools, they don’t need the Internet to do such work. Journalism existed long before the Internet. That is, essentially, how this empowerment works. All the fun cat pictures are there just so that people become interested in learning something about journalism.

It’s not to say that they’re just blindly copy-pasting those cat pictures. That’s not what we’re after.

Do you think, in Taiwan, there’s a lot of trust in the mainstream media in what the government says also? That’s a separate question.

In institutional media and in government?

And in government, so if the government were to respond to fake news about President Tsai’s PhD thesis, they issue statements. They say no, her PhD thesis is real. Do people believe that in Taiwan?

People believe in the full context of accountability. People like to make a fully-informed decision themselves. If this is a message in the form of trust as we’ve got it right, then, of course, nobody really trusts that.

If this is in the form of this is the complete report and you can download the PhD thesis from the National Library and read it yourself. Then, people tend to actually go and read it and trust more the people they already trust and what they reported after reading it. There’s quite a few people that really run it into software that matches plagiarism and things like that.

When more people do that, the more the authenticity becomes established. This example, you’re saying, this peer-reviewing the academic setting and this is, ultimately, an academic matter. People, of course, trust the academic community more than the governance or the news workers’ community. It depends on the case at hand.

What about the institutional media? Do people in Taiwan trust mainstream media?

It depends. For people who have their information sources as diversified, as some of the younger but also increasing the older generations as well, they can make their own judgments of what the bias of institutional media tend to be and then, just pick the institutional media with a kind of error correction lens knowing that this institution is already biased in which way.

That is how media literacy works. For people who are media literate, I would say they maintain a healthy level of distrust but not overblowing it either. They understand there’s institutional bias and read and view the messages with that in mind.

For “The New York Times,” and the US, I think we are concerned that there are not enough people like that these days, that lose faith in all large institutions. I just wonder, in Taiwan, what stage is Taiwan at? Is it so bad here or is it better?

There are certain, like the public TV system, that people generally trust. At least try and trust enough to run the presidential debates and policy presentations, which is actually going on right now.

If you think about the disinformation problem in other countries, as you say, China has a very clear interest in doing this kind of stuff here. Arguably, compared to even Russia and the US, China’s motivation is much stronger. Why are you so confident that Taiwan is withstanding this threat…?

I’m cautiously optimistic. Not “confident.” [laughs]

What are Taiwan’s assets in resisting this?

First of all, it’s just like biodiversity. If you have cultural diversity, then there’s less chance for a person who want to spread disinformation to find the ideological backdoor that, in a very swift motion, affects a large swathe of people.

Everybody here individuates by looking at the ideological constellations and just picking some stars that feel closer and can inform an individualistic constellation based on the transcultural tendencies they’re working on.

As transcultural Republic citizens, people understand first that there really are various cultures. Anything can be viewed through different cultural lenses. That’s something that people here intuit very easily.

Second, people understand that despite these cultural differences, we can still move forward using democratic means that are increasingly accessible to everyday citizens, not just by voting but by petitioning participatory budgeting referenda and things like that.

The more people participate in democracy, the more context they have. The more context they have, the less likely that a single propaganda can affect their understanding of any particular matter they have participated with. That’s the open government part.

The transcultural part and the open government part both serve as good, sometimes good enough, inoculations against disinformation. It’s hard to craft disinformation that targets a large swathe of people when everybody is so diverse in the beginning.

Arguably, the US is very diverse. Do you think that helps the US also? Clearly, because I think one thing we have not seen in terms of, we look back at Russia in the US election, some of the disinformation is very tailored to specific groups, very narrow groups.

Even specific ideologies?

Yes.

Strictly in the minority ideologies?

Right and specific places. Is China just not good at that yet? Why haven’t we seen that here? It seems that disinformation is quite blunt here.

I wouldn’t say that. I would say that the main venue they used, thanks to the work of the analysis teams that there’s a public education, have been received by the large social media companies essentially by the reports. They patched those bugs.

For example, during the Taiwan election, it’s no longer possible to run a precision targeted social or political issue advertisement without disclosing exactly who gets affected by it. It’s just not possible anymore. It’s definitely impossible if you’re a foreign national. That closes one obvious vector.

You may want to take it to YouTube or Twitter. It turns out they don’t even allow this kind of advertisement anymore. It’s not like the people who sow discord this time is somehow not as competent. It’s just that the venues they could have been using have been attributed by the investigation into the Russia case and then preemptively closed.

That’s true. Some of those avenues have closed. I went to Taoyuan this afternoon and talked to a city councilor who had a Facebook page just for his constituents in Taoyuan. People tried to buy it. All those people offered a lot of money to buy that. That’s quite targeted. Also, you can imagine…

One can’t easily fight against the vector of someone handing you their password.

That’s a different thing. The damage is less acute because you can’t be as precise when you do that.

That’s true. I was not here for the last election. What were you specifically worried about in terms of disinformation last year?

Last year, for precision targeting, because it was the first election in which the raw data of campaign finance will be published by the control branch, we know that a lot of people, instead of going through the campaign donation and expense, knowing it will be fully public for the first time, instead just went to precision targeting advertisement, because they don’t have to disclose that.

That was my main worry back at that time.

Precision targeting for advertisements on social media?

Yeah, social media. Now, of course, the norm is such that people know how is it like to disclose campaign finance by the Control Yuan.

In the self-regulation of the multinational media companies, we just say, “This is the norm here in Taiwan. We would really like to see you doing what our Control Yuan does.” We couldn’t quite say that last election because the Control Yuan haven’t published their first raw reports of campaign finance and expense to the public analysis.

Now, they have, so we just talk to them saying, “This is a norm in Taiwan. Here, you are expected to either disclose it as transparent as our Control Yuan or better, or you can choose not to run political and social advertisement during election.” By and large, they understand that this is social norm here.

What’s the norm?

The norm is for the Control Yuan to not only do the auditing of campaign finance and expense by themselves, but rather publish the raw data of each individual record on a public platform for everybody, including investigative journalists and data scientists to independently analyze.

They would publish advertising also?

If it’s billed as such. The previous election, what we’ve witnessed is that instead of going through campaign donation and expense to advertisement, they just simply paid for advertisement, therefore not entering the campaign finance system. It would have been hidden from the investigative journalists and data scientists.

What we’re saying is then, “This is norm in Taiwan.” If your Facebook, your Google, or whatever does sign on the self-regulation, we will say, “This is what the Control Yuan published. You’ll need to at least publish this much. Otherwise, it creates a perverse incentive for people to go to your system instead of the campaign donation one.”

That’s what Facebook has done. That’s what Google and Twitter now say, “Well, we’re just not running political advertisements this time.”

Did that norm change since the last election?

Yeah, because the last election was the first one that the Control Yuan published this publicly.

How did the norm change? Was there a specific…?

There was a law change.

When was the self-regulation agreement between the platforms?

It’s actually informally just agreed upon right after the previous election. It’s written into word as something that could be cited. It was June 21st this year. This is the keyword if you want to look up online.

Who is part of the agreement?

The initial signatories were Facebook, Google, LINE, Yahoo, and PTT. That was the initial five.

Is PTT less influential in politics than it used to be?

I don’t have that metric here.

Anecdotally, do you have a sense?

They stopped accepting registration of accounts. I guess as time goes by, that also means less young people are on PTT.

Is that going to change if they start registrations?

I really don’t know.

Why did they stop?

They couldn’t quite sort out newly registered fake accounts. Their ability to sort out fake accounts is dwarfed by the massive attempts of registering fake accounts.

Was the Kansai Airport incident a major factor in last year’s election? Was it important?

It was important in raising awareness of the necessity of fact-checking networks. That was literally one of the first cases that TFCC processed. It really helped to show people how TFCC works and how, if it worked relatively well before this incident happens, it could have prevented part of the incident.

You think so?

Yes.

TFCC is government-funded?

No. They don’t accept funding from governmental parties. The individual donations are only up to NTD one million each. Anyway, it’s crowdfunded.

Why do you think that incident made such a big impact at the time? Obviously, it was tragic more than most fake news. Was there something else about it that…

I think it’s mostly that because of TFCC’s work, that really outlined exactly what happened. What prepotently have happened. People really see information operation in a higher detail. This is step by step how information manipulation works.

A full attribution, of course, always is educational. I think it was subsequently uneducating. Do they report on it, where they added more attribution work? People learn a lot about that. I think that is an important part in it.

Ultimately, correct me if I’m wrong, TFCC has not also attributed any of that to Chinese government, right?

You mean this thing?

The origin of the information, the bad information.

The thing is that they have said that what they have cited by the local media was originally…

On Weibo.

Yeah, on Weibo. These were copy-pasted a source on Weibo, but they don’t quite say it’s sponsored by which entity or organized by some entity. Obviously, the initial round of reporting was simply turning some Simplified Chinese on Weibo into Traditional Chinese. That part was part of the attribution.

Why do you think this case is having such an impact now? People are still talking about the…

Because it’s connected to something that people remembers. That’s tragic.

Do you think it’s changed people’s impression of DPP also? It’s exposed something about DPP?

As a nonpartisan, I don’t really know how DPP thinks about this case or how the candidates think about this case. For the record, I don’t know how KMT think about this case, either.

By and large, people understand it’s something tragic. Anything that assists in the tragic progression, no matter which sources work before or after her is seen as a contributor to the tragedy. That is natural.

Do you think the parties are responsible for creating a lot of disinformation?

The parties? As in political parties?

Political parties.

In Taiwan, there’s so many political parties that it’s impossible to make a general…

The major parties.

I personally only associated myself, once, with one particular political party, is the Can’t Stop This Party or literally the unstoppable happy party, the 歡樂無法黨. I sent them a basket of catnip on the formation of their party. That’s the only party I associated myself with.

I would say that the more prominent party members, there’s four of them: Brian, Froggy, Chih-Chyi and Retina of the EYECTV. They’re all YouTubers who use humor in their response to further public discussion about public events.

Although, of course, many of their content is subject to debate, I wouldn’t say any of them engage in intentional harmful untruth, which is our legal definition of disinformation. The only party that I associate with I think it’s pretty OK on the disinformation front.

I don’t know about other parties.

Separate from the China issue, this is something a lot democracies are dealing with. Digital campaigning is now very, you can do a lot of new tricks. You have a lot of tools now.

It’s very memetic.

I don’t know in Taiwan how bad or how widespread any acquisitions of official untruth. I just don’t know how big of a problem it is.

It’s not unheard of, but I wouldn’t say it’s a dominating factor.

Tell me what a day in the life of the digital minister of Taiwan is.

Depends on the day of the week.

Give me a…

On Wednesday, I meet with nine batches of people, because today is my office hour. [laughs]

That’s your office hours. What about non-office-hour day?

Yesterday, I was in Wulai because on Tuesdays, every other week or so, we tour around Taiwan to listen to the rural, indigenous, or remote places and connect back to Taipei through video conference. That’s a typical Tuesday.

A typical Thursday, I attend the cabinet meeting, the office of board of the science and technology meetings and do meetings. That’s the only day I have to be in the Executive Yuan building.

On Friday, usually, every other week or so, we have a collaboration meeting with stakeholders anywhere in Taiwan, as long as they can get 5,000 people petitioning or respond vigorously on a regulatory preannouncement, both take place on the Join platform, which has 10 million visitors, which is a lot.

That’s Friday, is summoned in two places. Every Monday, our team have a lunch together. That’s pretty much it.

Sorry, the Friday is for a specific digital petition?

Yeah, digital petition and…

The government has to respond to it if a certain number of people…

After 5,000. Yes.

What kind of petitions do you hear?

There’s a listing of all the petitions and regulatory announcement we’ve worked at about in oc.pdis.tw . If you type this down, you can easily see all the 58 cases we’ve worked with. There’s some major trends, like animal welfare, transportation, improvement of digital services, public sector work environment, education. It’s a full spectrum.

What did you hear in Wulai yesterday?

Lots of things. We were working with co-ops, which is a long and proud tradition, especially in indigenous lands. Increasingly, they’re facing a lot of regulatory challenges and differences as compared to modern companies, because Taiwan just changed our company act. Each company can have a certain sense of autonomy like a co-op, like special voting rights and things like that.

Especially the labor co-ops, which is very strange in the eye of many regulators, because they don’t have any capitalist side, it’s entirely they’re worker side. It’s difficult, for example, to talk about the insurance, the procurement, the labor law enforcement which all are designed to reign the capitalists. In the sense of a labor co-op, there is no capitalist in the equation.

How exactly do the regulations work? The public procurement, public construction commission designed a check form for people who are in the public procurement in various different municipalities and ministries to check that labor co-ops need to have their co-op charter agree with the public procurement. They cannot just get all sorts of different cases. They have to at least align with their charter.

The design is so that if you don’t click that button, if you don’t check that checkbox, then the entire procurement become illegible for worker co-ops, especially worker co-ops to apply. That becomes a discrimination of sorts. Our constitution said specifically that state is required to encourage co-ops. That creates a counter-constitutional situation.

We just work with the public procurement agency to just change that check box into a reminder text. Of course, there was 20 different cases. I’m just saying one that requires cross-literature collaboration.

Who goes on Tuesdays? Is it just you or other ministers?

All the relevant ministries.

Really?

Yeah. It’s convened by reverse mentors of the cabinet, 行政院青年諮詢委員, the youth advisers. They can pick and choose which agencies they feel are related.

Yesterday, there’s at least the ministries of culture, the indigenous council, the ministries of education, economy, public procurement obviously, finance, national development. I probably have missed a couple more, at least eight or nine ministries in the meeting.

To return to disinformation, your proposed solution that you gave to the CPJ, how much of that is implemented in Taiwan right now in terms of rapid response?

All of it. It’s what I said, one hour is quantitatively shown as the most effective. The current requirement is to respond in two hours.

Who responds?

The cross-disciplinary team in each ministry. They all report to Kolas Yotaka, our spokesperson.

When did that start? When did the system begin?

It’s proposed early 2017 in a cabinet meeting. It’s implemented in some form when 徐國勇 was our spokesperson. Having a clear-cut, quantitative measure within two hours, 200 characters, two pictures, and things like that, that was since Kolas took the role. July 2018 was when she became our spokesperson.

By July 2018, the quantitative metrics of the…

Are starting to form. Kolas improved on it. As the cross-sectoral collaboration progresses, because that also gets a boost when, for example, LINE Today agree to syndicate the clarification pictures.

LINE Today only started working on that as their digital accountability plan this year, this July actually, after they signed on this counter-disinformation code of practice. There’s a year or so of preparatory work. It gained a real boost starting this July.

That was when LINE Today agreed?

Yes.

Cautiously optimistic, what makes you pessimistic? What is the flipside of all this for you in terms of disinformation, in terms of Chinese influence?

I’m not pessimistic about anything.

Why cautious? Why not be fully optimistic? [laughs]

Because we understand that the people who are somewhat discord are also building a community, that they are also learning from each other’s playbooks. Just as how the Great Firewall becomes great is done by learning from the people who counters them and then innovating.

We never are not conscious [laughs] of the fact that there are new innovations happening literally every day. With machine learning, it’s not only human that innovates. Of course, we need to stay cautious about it.

Were you a very political person before you became a minister?

Like Internet governance? Of course, I organized large open source projects. That’s a political work.

What about more traditional political issues?

What do you mean by traditional political issues?

Like the outcomes of elections and Chinese…

Oh, representative democracy.

Yes.

Even at this moment, because I’m not…

I guess you’re non-partisan.

…a member of any party, I am not really working directly with representative democracy.

We work on ways to make a more informed citizenry, which in theory makes representative democracy also better and complement the work of the representatives, but I am not directly working with, say, the political party system even now.

What I’m trying to understand is what differentiates Taiwan from many other liberal democracies is the existential questions that hang above Taiwan. The future is just totally up for grabs in Taiwan in a different way than it is for…I don’t know. It’s comparable to what Israel faces. How does that color…

All the jurisdictions are feeling some of that now, thanks to climate change, but yes. [laughs]

Climate change is a good example. Liberal democracies are not very good at thinking long term like that. For you, what is your picture of Taiwan’s future relative to China? Is it something that you can actually…Does it impact your day-to-day life?

When we are working on digital democracies, we are thinking about the whole community of Homo sapiens and of future generations and of mountains and rivers and the ecosystem, so not just specifically about any nearby jurisdictions.

That’s what digital democracy offers, in that the proximity is not territorial. The proximity of Taiwan to, say, Switzerland is very close after we implement the referendum act, to Iceland very close after we implemented e-petitioning and things like that.

We have more natural neighbors in the digital democracy scene. That is different from the more traditional West value and geopolitical thinking. I’m grateful that I get to work on this layer. I’m not usually defining my work in this layer.

As a Taiwanese, you must think about China. You must think about the future in terms of what’s going to happen with China.

I’m working with Taiwan, yes, but that doesn’t mean that I…

You live here…

Sometimes, yes. I travel also.

You don’t think about China?

I sometimes think about it, but certainly not in a day-to-day hour-to-hour basis.

I think about Taiwan most because the system that I helped design and facilitate, people in Taiwan embrace it a lot. Then, I’m also board member of international NGOs, just today in Amsterdam now, and then, also in New York and also in Spain and so on.

The system that I helped design and facilitate do not necessarily have Taiwan as its first users. I’m helping anyone who wanted to do digital democracy. Taiwan is just one particular large, liberal democracy that is willing to try out this new invention.

I’m not designing this for Taiwan… I’m designing this with Taiwan.

Is there anything about Taiwan’s particular geopolitical position that you have to adjust for in digital democracy or is it truly universal?

I think, of course, Taiwan has broadband as a human right. That makes the legitimacy theory much more robust.

Taiwan has a broadband as what?

Broadband as a human right.

Oh, yeah.

As in that anywhere in Taiwan, you’re guaranteed almost, 98 percent now, to have 10 megabits-per-second Internet connectivity. In a digital democracy, it’s complementary to face-to-face democracy. It’s one more venue to include, instead of excluding, people.

This makes the legitimacy theory much easier to build as compared to other sometimes larger jurisdictions where there’s a large class of people where the best way to deliver digital democracy is still through SMS or even a phone, landlines, or even through ATM.

It’s the broadband as a human right that really helps. That’s the only thing.

You don’t think the fact that China is in this position relative to Taiwan makes a difference with digital democracy?

Not really. The work that I do working with people in, say, Estonia, they also have their backup plan in terms that their neighbors do something.

Giant neighbor.

So it’s not like this is a uniquely Taiwan problem. Many people in Taiwan, of course, are paying a lot of attention on, as you say, the adversarial situation. It’s just my particular line of work don’t need to worry about it as much.

I’m conscious of the time. Anything else in the current election cycle with disinformation that you feel we didn’t talk about, that you’ve thought about recently?

Not really. It will be very interesting to see the year after that, the August referendum day, how that works.

That’s 2021?

August 28, Yes. Because then we’ll have a new tempo, a new norm really, of a representative year and a deliberative year and then a representative year and then a deliberative year.

It will be interesting to look at disinformation in the landscape of a purely deliberative referendum-only year, and how that evolves and whether the work that we’re now doing makes a more informed discussion between the stakeholders or whether this way is still…

Like being the polarizing effect of multiple referendums at a single day still outweighs whatever the informed discussion that people had before the referendum day. That’s a large hypothesis-testing time. We don’t know the result. We’re working toward that, but we don’t really know what will transpire.

Until we move to the day when social media is open source entirely and there’s protocols that are available for everyone to build on top of…

Like democratized…

We live in a world where social media is very anti-democratic. It’s controlled by these big companies.

I don’t know… PTT is open source.

PTT is only one of many. The others are all very commercial. They’re all profit-seeking. In this world, what can the government do? Is there a case for breaking up these companies? Is there a case for regulating them differently? This is a debate in the US right now. I’m curious what your…

I know. States Warren and…

First of all, I think the Taiwanese social media companies are doing fine. Most of the innovations in Taiwan, like Gogoro… They are not a media company, I’m just using it as an example.

Like who?

Gogoro.

Gogoro.

They do electric scooters, but they’re also a pioneer in renewable energy and energy storage and also in our carbon reduction. This kind of triple-bottom line thinking in Taiwanese startups is very common, so common that the social innovators and the startup people now coexist in this space and even share the same incubating program.

It’s very rare to find a purely profit-seeking startup. They have to at least design some of the public benefit into it or social accountability, partly because the social sector in Taiwan is really strong, really legitimate. Social sanctions are really bad [laughs] for any company.

Also that people understand that we’re, as I said, a transcultural republic citizen. Your idea must not overtly decimate other cultures. All of this makes a startup scene that are much more sustainable, a little bit more long-term thinking, and makes, say, impact investment and so on a more acceptable idea. All of this is different from Silicon Valley.

Unfortunately, Silicon Valley still rules the world. [laughs]

I don’t know about that. [laughs]

Facebook’s influence here is also enormous.

It is enormous, but I would not say it “rules the world.”

This I will grant: FB is a co-governor. It is a governor in its own right. I would even go as far as to say they are now acting sometimes semi-sovereign. Still, it’s a large distance between that and saying they rule the world, which sounds like hegemony.

Is there a case for government intervention? Or is Facebook inevitably, in its DNA, going to facilitate bad things because of its profit-seeking, because it’s hegemonistic?

The oversight board that they’re building, which is a kind of rule-of-law system, a judicial branch within a public corporation without a legislative branch, is something that’s quite novel.

I don’t think Zuck is doing this purely for PR. I do think he is giving an earnest try in building some sort of accountability apparatus by learning from the legal theorists as a co-governor.

Whether that will work or not, or whether, as you say, that it’s in its DNA as a shareholder-profit-seeking company to be not limited by public accountability, that’s literally the question of the coming year.

How it gets answered, not only by Facebook, but by all the large Silicon Valley players, will ultimately, I think, determine how the US government acts toward it.

Yeah. You’re impressed so far with Facebook’s efforts? It’s not PR?

No, I wouldn’t say I’m “impressed.” It’s so early in its days. I’m just saying they are giving it an earnest try, but like any of their new products, it may or may not work.

It’s not at the stage where I can be impressed or not.

All right, thank you, Minister Tang. Thank you.

We’re good?

Yeah.

Thank you.

Can I ask, in our stories, we say Mr. X, Ms. Y. What pronoun do you prefer?

My gender is literally whatever. Some people use Mx. That’s fashionable now.

Yes. We can use Mx. if that’s what you prefer.

Sure, of course.

Mx.?

Mx. is good. There was a Hebrew-writing journalist. All the verbs are…

In many languages it’s very tough.

All the verbs are inflected. I think he alternated between feminine and masculine forms, like every other pronoun.

I see. We can do Mx.

Mx. is good. Cheers.

Thank you so much.

Thank you.