We wanted to begin with your philosophy, the philosophy behind some of the things that you’re doing. You have described yourself as a “conservative anarchist.” What does that mean?

Certainly. To conserve is to respect the traditions. In Taiwan, we have more than 20 national languages, each with their own cultural lineages. To make “progress” in one culture may very well be a kind of epistemic injustice against the other 20 different national languages and cultures.

By “conservative” I mean to adapt the technology to the society rather than asking the society to adapt to technology. That’s what I mean.

It draws a lot from the indigenous nations’ notions of sustainability, and indigenous people’s sense of a spirit, like the Saviah or Patungkuonʉ, the top of Taiwan, the mountain is itself a spirit, and we are ephemeral actors who work to conserve the spirit. That is what I mean by the conservative part.

The anarchist part is more related, of course, to Internet’s end-to-end principle, meaning that the wisdom is at the edges, at each individual end. No one in the middle can or should have a say in what the edges have invented or innovated to communicate.

It’s basically switching from coercive power to an end-to-end principle, which is essentially a power that’s based on rough consensus. It’s based on deliberation, based on the general understanding that the earlier innovators must not restrict the innovations of later innovators.

Be liberal in what you accept, be conservative in what you send. Liberal in what you accept. That are all core, the Postel’s law, core Internet principles that shaped the early Internet. Of course, it’s relevant. It’s still very relevant.

This is Peter: We’re running this interview together in our own kind of Taoist exercise, I suppose. How’s it been like taking those principles, especially the early Internet philosophy, and then moving that into actual policymaking? Is it possible to translate those ideas?

Maybe I don’t know the…I know the philosophy of the early Internet superficially, just from John Perry Barlow, the former Grateful Dead singer and stuff, and it’s literally anarchist sometimes.

You now work in government, and how does translate into…

With the government.

Excuse me?

Working “with” the government.

With. Sorry.

What do you mean by “with” the government?

Meaning that I call myself digitalminister.tw, but it doesn’t mean that I command anyone. All the colleagues in my office, they are secondments from other ministries. They report to their own ministers. They don’t report to me.

They basically are given a space to form rough consensus, to work out loud, to publish not only the end result of policy-making, but the process of the making of the policies by listening at scale to citizens, and so on.

At the end of the day, I do not give them command, and they do not give me orders, either. This is what I mean by working with the government on one side, and of course the social movements on the other. I’m kind of like Lagrange point between the social movements on side, and the governments on the other.

Fine. That already sounds a lot like the philosophy that we’re talking about, but in terms of vision of the Internet, is there a role for the state in it then? How does one make sure that that role is never coercive and is always…?

Also, by setting up these projects, by setting up Polis or vTaiwan, isn’t that an intervention of the state, or how do you describe what you’re doing there? Is that, you’re creating a framework…

It’s a people-public-private partnership. The order, very important. Always, the norm is set by the social sector. That is to say, the people, variously called the voluntary sector, the third sector, the civic sector, the civil society, etc., but I will say social sector for the sake of the conversation, because it rhymes better with social innovation.

Anyway, the social sector sets the norm. For example, around countering infodemic, the fact-checkers set the norm. Around, for example, the targeted advertisements on social media, that actually has a antisocial externality when it comes to election, and the social sector sets the norm of radical transparency of campaign donation expenditure.

I can go on, but the point is that the social sector sets the norm. Then the public sector amplify the norm by essentially saying that, “We found that the social sector now has a norm that says, for example, during the elections, all the campaign donation expenditure are to be published as open data, not just open data, but rather open API as quick as possible, as structured as possible.”

That enable us to essentially do a negotiation with, say, Facebook saying, “If you don’t want to be labeled antisocial media and face social sanction, it is actually in your best interest if you publish campaign advertisements exactly like campaign donations.

That is to say limited to only domestic contributors, and also publish the exact terms, like algorithmic transparency in real time for independent social sector journalists to analyze.

If they don’t do so, we don’t really have jurisdiction over them, but they will face social sanction. This is the state’s role to amplify the norm that’s set and agreed by the social sector, and to make sure that it can also reach other stakeholders in Internet governance so that they can act in a prosocial rather antisocial way.

That actually, what you’ve just described, was tried in Europe during the last European election campaign, and it didn’t work.

The EU vs Disinfo.

It’s part of the code of conduct for the parliamentary elections that Facebook was going to do an ad library, and they ignored it, basically. [laughs]

I know.

They ignored it. It was very hard to access, there was no way to prove that they were doing it openly. It wasn’t even clear that the public was that interested in it when it appeared. A few little NGOs worked on it, and it had almost no impact.

We’re aware of that. Actually last December, when Facebook put a lot of resources into the civic integrity, we’re aware that Taiwan is probably the first jurisdiction that they actually did what they claimed that they would do in Europe.

Why did they do that in Taiwan and not in Europe?

Otherwise they will face social sanction here.

What do you mean by social sanctions?

Like boycotting, like people actually going to, say, PTT or the vCard, or any of those alternatives run by other players who, like PTT, have already signed up to these social sector norms. Basically they would be seen as the odd entity out.

If everybody else agree that this is the norm, then people would just maybe publicly delete their Facebook accounts if Facebook continued to not respect the local norm. Our government – the control branch is a separate branch in the government – published the campaign donation expenditure previously also only as statistics in a very difficult to understand format.

Thanks to the citizen activist, hacktivist even, that went to the Control Yuan, brought out photocopies and did a OCR, a massive crowdsourcing project to turn these into CSVs, into independently reportable numbers.

That also partly enabled the independent investigative journalist to discover that the majority of campaigns did not file their social media advertisements budget as campaign donation or expenditure, essentially creating a loophole in the political process.

When their petition and activism became the new law in 2018’s election, that’s the first time that the control unit adhered to this social sector norm. Everybody see that it actually made things much more transparent, and then it left people asking why is Facebook not catching up?

The fact that the social sector sets the norm and pressure the public sector to accept that first as an example, and then later on, the social sector, social media, all joined and would indeed single out Facebook if they are the only platform that didn’t do it, it create enormous pressure based on the social plus public sector norms.

The problem in Europe was that we didn’t have citizen activists who were that attuned to the problem and weren’t paying attention.I’m trying to understand what’s different between Taiwan and Europe. Is that the problem, that we haven’t got this same public…

There are alternatives also, because the PTT or, really, vCard, or Plurk, or whatever, they are real competitors to Facebook. Also, like the PTT, although we call it the Taiwanese equivalent of Reddit, it’s actually different because the PTT is run essentially by National Taiwan University students in their spare time as an exercise in open source governance.

The entire PTT regulatory framework is open source, and the code for PTT is also open source, enabling participatory accountability. That is the norm in Taiwan. Anything that falls short of this norm automatically have less legitimacy. That probably cannot be said in the whole of EU.

Also, I’m thinking like, you’re talking about the social sector as a unified body that everyone seems to trust and represents everyone. I can just imagine in America, where civil society is already a polarized term, civil society is seen, sadly, as only representing one part of society.

It’s like a degree of trust in the social sector in Taiwan that also helps this or is that polarization issue?

Yes, because the social sector really grew in Taiwan even during the martial law, working on, of course, charity projects and so on. Because we have lots of earthquakes, typhoons, and so on, and that really mobilized the entire social sector to essentially fulfill the role of governments even before the democratization.

During the democratization, you can see it’s the co-ops, the local community colleges, and so on that essentially prototype how democracy could actually work so that we have a very peaceful democratization.

Because a lot of those fair, trusted NGOs cut their teeth on the social issues in the ‘70s and ‘80s, but our first presidential election was in ‘96, so that means that the state has a massive deficiency when it comes to the time that it takes to earn legitimacy.

Even now, if there is a typhoon, and a large NGO publish a number of damages and so on, and the local municipality publish a number, people tend to trust the charity’s number more.

In America, actually, there’s new research that shows that one of the civil society bodies that actually overcomes polarization in America are megachurches. Very surprising, but they actually have both sides onside.

Exactly.

Maybe Facebook reform has to be pushed by megachurches, and then everybody in America could get behind it. That’s very interesting. I wonder how one could replicate the mutual commitments and trust in civil society.

There’s even a playbook for that from the Pope called, if I am not mistaken, “Fratelli tutti.”

Right, yeah. The Vatican encyclical.

The Vatican, yeah. That’s actually where you can begin, because it talks in very concrete terms how to rebuild those trusts. It’s an encyclical letter.

The trouble in the US is that you have a chicken and egg problem, which is that you have this very profound polarization which is driving social media to increase the polarization. It’s very hard to know how to end that bad cycle.

In other words, the polarization is real. That is then reflected in social media. Social media then amplifies the polarization, and in the US, we have an actual moment of epistemological break, where people now live in literally different worlds and believe different facts.

I thought you are now recovering from the epistemic break.

No, I think the…

I think people are realizing, like the fact that you point this out, I think, is a really good beginning. For the longest time, we see people in denial, like people were saying, “No, this is perfectly fine. There is no epistemic crisis.” Admitting there’s a crisis is a good first step.

It is, but in America, I’m not sure who’s meant to heal the crisis. In Europe, we have a tradition of public service media. There is an idea about who could do this. In America, they trust innovation and a tradition of civil society which I think is quite drained now.

As we’re on polarization, maybe we should get into the innovations that you’ve been pioneering or you’ve been helping pioneer. We’ve read a lot about them. We don’t you to repeat for the millionth time how it works.

We think we understand it, even though we’re not technologically that savvy. Could it just as easily work somewhere else? Is Polis and vTaiwan, is this like a new model that we can start overcoming polarization?

Polis is, of course, already quite well-adapted. It’s not just in Taiwan. If you go to pol.is, there’s a very comprehensive knowledge base about the case studies, best practices, and so on. From what I’ve read, it’s not just in Taiwan.

I think a lot of European jurisdictions are now actively looking into using Polis. I was just consulting with a group that was doing early experiments in Amsterdam, but there’s also, for example, the Demos, London, UK, which may or may not be Europe.

Anyway, New Zealand also, which may or may not be Europe, and also the Alternative Party and so on.

We work with Demos together, so know how they’re using it. Anne, I think you had some very specific questions about Polis and how politicians might react to it. I don’t know.

I’m married to a politician. I was describing to him how Polis worked. His first reaction was, “Well, if that works, why do you need us?” Meaning we politicians.

You can tell him he is the second diamond. Polis is the first diamond.

[laughs] Have you had resistance to these ideas in Taiwan from the political class?

Yeah, which is why we always say in design thinking terms that Polis is just for crowdsourced agenda-setting, nothing more, nothing less. This is for the discover or exploration phase in design thinking terms, in the ideal double-triangle.

It does a little bit of define because it defines the rough consensus, the resonating affect, the resonating ideas, but that’s all it does. Everything after that, like actually discussing over that agenda set by the Polis, is still for the professional politicians to convene such multistakeholder forums.

Of course, Polis can help, because you can invite the people who propose the best ideas or the most controversial ideas to the table. Basically, make the plurality of voices more plural, but everything that follows afterwards, after the define stage there comes the development and the delivery.

That is still traditional political work. What Polis does is essentially scaling out the initial exploratory phase of listening.

How do you screen out bad actors and people who are deliberately participating in a false or undermining way? That’s a problem that almost no social media has solved.

If you take away the reply button, then it’s automatically solved. The reply button allows the troll to grow.

If you take away the reply button, as many systems nowadays do…Like Slido is doing this by default, or our e-petition platform, Join, learned from the Better Reykjavik system, where you can post pro and con statements and to upvote and downvote, but there is no reply across the aisle and so on.

If you design the ideascape, for lack of better term, that makes sure that the trolls cannot waste each other’s time, then they have better spaces to work on, like more satisfying space to work on. We also have a sense of exactly maybe one account per SMS number.

In a sense, this is not real name, but you do need to authenticate through SMS to get into the Join platform in order to start upvoting, or participate in quadratic voting, or things like that. There is also a measure of authentication when we make sure that people act in a way that behaves responsibly.

They, of course, often can choose pseudonyms because we also believe that pseudonyms can ease the power imbalances.

This is actually one of the themes I was going to ask you about, which was anonymity. In so many fora, study after study is showing that people who are anonymous behave much worse than people whose identities are real and who have to take responsibility for their speech.

I didn’t say anything about anonymous. I say pseudonymity, meaning that you can choose a nickname and stick with it.

What’s the difference between that and anonymity? That’s the definition of anonymity. People don’t have to take responsibility for what they say because they’re hiding behind a pseudonym.

No, it’s different, actually. A pseudonym that persists over time, of course, has its own identity. It’s just you can’t go back and trace to the real person behind that pseudonym. It’s one degree of protection.

For many people, actually, their pseudonyms are much more widely known than their real name. For example, _why is a great example in the Rails Rudy community. The full name is why the lucky stiff. You can google that and learn the story.

A pseudonym is in no way temporal or temporary. It can be as long-lived as an actual human being, and we promote that. We also understand that, if people have to participate with their legal real name and with their traceable-to-actual-person identity, then there’s a lot of things that people who are currently suffering power imbalance would not dare to bring up in public consultation.

This idea of radical transparency that you talk about a lot, how far does it need to go, and are there any limits on it? In this case, we’re talking about whether people do have the right for a little bit of a lack of transparency in order to protect themselves.

How do we get that right, or what is the real…? We get transparency a lot as the solution, and I just wondered, as one thinks it through, when does it end?

Radical transparency means transparency at the root. What I mean by it is that the state need to be fundamentally transparent to its citizens. I do not mean that the citizens need to be transparent to the state or to each other. This is the direction of transparency.

The other form of transparency, although it’s sometimes also called transparency, I call state surveillance, because that’s what it is. For me, radical transparency means as the digital minister, I’m recording only myself on my side, and you are free to opt out.

If you opt out of radical transparency, then at the end of the day, our transcript will just me monologuing. I also made sure not to quote any anecdotal, third-person stories that you may have told, because at the end of the day, you might want to redact those, because you have not cleared that to your friend that you are at the liberty of mentioning here and so on.

There’s two principles here. The first is that the state actor must be transparent, but the citizens may or may not. The second is that transparency at the root means transparent by default. It takes some effort to take out your side of the conversation. It’s not a lot of effort, but if you do nothing, it’s transparent.

I suppose that gets into the next question, a question about the other big debate everybody’s having, is about data and data rights. What is privacy? Obviously, we have this idea of surveillance capitalism.

We have the Chinese model, which is all about giving up a lot of your data. But without data, I guess the Internet loses its power as well in terms of its problem-solving capability. How have you solved this?

Given that lots of data could be used to solve healthcare crises, including the one we’re having now, how do you balance between the goods and the bad uses of mass data collection tools?

Data, the term, is too generic. It’s not helpful. This is like you’re writing a book, or maybe you’re doing an interview. You’re saying probably, “Oh, I’m doing an interview. I’m doing journalistic work. I’m a journalist or I’m an author.”

You don’t say, “I’m a text collector. Give me some text.” Or your editor probably don’t say, “Hey, I’m just a text processor,” like, “I can process five megabits.” Well, it’s not the text. “Five megawords of text per year.”

You don’t call yourself text workers, and neither should the so-called data workers call themselves data workers, because it’s simply too generic.

When we say it’s journalism, it means a strict subset of text that has its own social norm around it, around fact-checking, around framing, around biases, around the integrity of the journalistic profession about the internal ethics of the journalist as part of their training and things like that.

This, for a lack of better term, I refer to as competence. Just as a journalist would have media competence – that’s a better word than media literacy, by the way – a so-called data curator or data producer need to have digital competence in the sense that they are aware of the limits of the technologies.

They aware of the negative externalities of these technologies, and so, when they are doing measurement, when they are contributing to the commons, and so on, they think in these terms, instead of in megabits of data terms. It’s as meaningless as megapages of text.

I suppose I have got stuck in thinking about data as a block. Practically, would that mean people would have more rights around these different types of data?

Right. For example, is this about the common sensing of the environment? If this is about us measuring the air quality together so we can have a more clear picture about not only pollution but also Climate Change, then we just call it exactly that.

We have maybe a crowdsourced collective intelligence system, or as we call it, the civil IoT system, sci.taiwan.gov.tw. We call it by its name. Then people will not confuse it with, for example, the health records, because there is nothing in common between the PM2.5 values measured by the AirBox and the health record that you put into the trust of your doctor.

There is literally nothing in common, except both may be called data. That’s quite beside the point.

What about the kinds of data that are captured by social media companies, by your interests in pop music, and the kinds of ads that you click on.

Yeah, profiling, yes.

The profiling data. Do you think we need a regulatory regime for that kind of data? Should there be some controls on it, or is that just up to the companies or up to consumers to react to it? How do look at social media data, profiling data?

Again, I would think in terms of not consumers, because in a sense, we are not consumers. We are producers. It’s just through those intermediary platforms, different producers are paired into each other’s so-called news feed, having no journalistic…

They keep redefining those words, so feeds. I will just call them feeds. Those producers of media are paired to each other’s feeds, so we can share what each other have produced. In no point here is a traditional consumer relationship defined because first of all, I didn’t pay for any of these things.

The second thing is that I’m not tuning into a radio or a television. I’m not merely receiving things, I am constantly interacting. Indeed, it is the interaction pattern that allows the profiling. I don’t think the term consumer is at all useful when we talk about Facebook.

Now, with that in mind, I do think that Facebook, like any co-governor – this is the term that I use – they may not have sovereignty because they have not yet printed their own money, although they did certainly try. They are a co-governor when it comes, for example, to election campaigns because a lot of money flows into it to get visibility.

In that sense, they are a campaigner, or at least a co-governor that can regulate how campaigners work. Because they have their own jurisdiction over the algorithm, even though there’s no appeal process…Although there is the oversight board, but maybe one percent of the things go to the oversight board.

The other 99 percent is completely arbitrary, so we need to essentially treat it as a trade negotiator would in the sense that there is a rule-setter. It has overlapping jurisdiction on many cases.

What kind of demand can we give it that is backed by the social norm in my own jurisdiction, and what kind of demands are there from them that will benefit them, but at an expense of the general society here? Basically, I just treat it as a foreign power, which is what it is.

You think that has to be done jurisdiction by jurisdiction? There are some dreams which aren’t actually at that great a distance, about democracies creating common regulatory norms around algorithmic transparency …

Having ombudsmen who can oversee or co-create them…

Yeah, and I’m all for it. I’m just saying that ombudsmen in one jurisdiction doesn’t necessarily translate into ombudsmen to another jurisdiction. You need to do some translation work, but it’s worthwhile work, like GDPR.

We are in the process of translating GDPR to our jurisdiction. There are, of course, a lot of structural differences. For example, what we call the, as I mentioned, civil IoT system is called called joint data controllership in GDPR.

The GDPR clauses on joint data controllership is quite thin and ill-defined. That it’s actually hard to see how the existing civic union work on data gathering and so on in Taiwan can map cleanly to the joint data controllership clauses.

It’s, of course, still useful, because I can talk orange to orange. Also, I’m aware that there’s this data governance that talk about trusted intermediaries, or trustworthy intermediaries, or whatever term that ends up getting used. That is much closer to what we have in mind here in Taiwan.

In Taiwan, have you thought about this process of pressuring, or demanding, or negotiating with Facebook in order to get insight into their algorithms?

Oh, yeah, we do that quite openly. If you look at my SayIt website, there’s at least four such public negotiations, not only with Facebook, but also with Google, with Line, and other players.

I’m wondering if enough democracies could club together around common principles, essentially we’d have trade unions. We’d be like trade unions bonding through the algorithmic rights.

Exactly. That is actually quite powerful in international trade. If people agree on common principles, that allows them to form a trade partnership on particular goods and services. Indeed, I’m not saying that this is purely a trade issue, but internationally speaking, this could be modeled as a trade issue.

It would first require democracies to understand that this is necessary.

There needs to be progress to find consensus among the democracies that we think are of the key movers.

Yeah, definitely.

Audrey, we’ve been talking a little bit about these American companies, but of course, the other big challenge is China. You’re right on the edge of that. To what extent do you frame or do you think about your vision of how Internet and democracy should work together vis-à-vis the Chinese challenge?

It seems to me quite an insidious challenge, in many ways, not least because, in its own weird way, the Chinese internet is an efficient Internet. It seems to be, anyway. It seems to deliver services to people. It’s not open about its oppression. You’re much closer to it. What are its strengths and weaknesses?

One of the things that I think is quite clear is that they use some of the same words that we use to mean opposite things. I already talked about transparency of the state versus transparency to the state.

When we say that we operate on a rule-of-law principles, meaning that there’s a justice system that upholds the human rights and so on, they say they also have laws – actually, privacy, cybersecurity laws – but it’s interpreted in a very strange, some scholars call it, rule-by-law system, rather than rule-of-law system.

Yeah, I’ve just written about this.

Right, so that’s another thing. There’s laws. There’s rules, but it means different things. Also, for example, in Taiwan, when we say something, that this person have a lot of social credit, it means nothing about the government, but rather about essentially credit unions, which is a thing in Taiwan.

So that this person is trusted financially by a lot of their nearby neighbors, but they use that use that term to mean that people are willing to be tracked in their public behaviors to gain preferential treatment in education, employment, and household registration.

The people who have a low social credit, rather than people who don’t, I don’t know, agree on their debts, they mean that people who need to be shamed by disclosing their names, denial of airplane travel, high-speed trains, and each province may actually add a lot more to this public shaming, which was not a part of what the term [non-English speech] used to mean until they just reappropriated that term, I guess.

We are right next to it. We do see that these totalitarian norms are being put and sugarcoated in previously harmless words and then made into new norms that an unsuspecting public may look at it and say, “Hey, they are transparent. They have law. They have rules,” but it doesn’t quite mean what you think they mean.

China seeks to make its system appealing by changing the meanings of words, but also by creating a very efficient Internet which claims to be able to control crime and so on; How does that affect the way that you think about what you’re doing in Taiwan?

Do you think about that as you’re in opposition to that or does it not bother you? You’re so close to it physically and in other ways.

Sure. We’ve not had any procurement of PRC – that’s the People’s Republic of China – regime components since 2014, where we literally occupied the parliament for three weeks, protesting, and then later on, demonstrating how to talk about trade deals.

Involves half a million people on the street, and many more online. The 20 participating NGOs each talk about different aspects. One aspect is precisely this aspect. If we make the telecommunication deals with Beijing to make sure that in our then-new 4G infrastructure, we use these efficient, inexpensive components…

Basically, the same conversation that people around the world are now having about 5G, we had that in 2014. We said quite clearly then that the amortized cost, if you take into account that for each systemic upgrade, firmware upgrade, or protocol upgrade, you have to reassess whether their party have maintained now de facto control over the previously so-called private sector.

If you have to keep doing this systemic risk analysis, then amortized, it’s actually much more expensive and ineffective as compared to, for example, if we work with, say, Nokia or Ericsson, who probably will not get a de facto state takeover through swapping leadership anytime soon.

We eventually said no to any government project, not just telecommunication. Things with sensitive data – now, I get to use data, because it’s telco context – and also national security, including cybersecurity concerns.

The tender document must explicitly reject PRC suppliers, and it’s been true for the last six years. Of course, nowadays it’s been branded as clean path or something. Anyway, we just said that it is actually much more economic to develop our own clean devices and our own cybersecurity standard.

You framed that around an efficiency argument, but it’s clearly a security argument as well. How does one find that balance between security and openness? You talk so much about transparency, but here was a really quite hard geopolitical decisions.

How does one make sure that those decisions are also based on rights and democratic values, or is it just like them and us? We know that they’re threat, therefore, they don’t come in. Is that how ultimately that’s decided?

The idea here is that if they actually have a rule-of-law, or I don’t know, rule-by-law system, this is the entire conversation about the extradition in Hong Kong. Why would they protest against extradition if there is a true rule-of-law system that’s going on?

If everybody can get a fair trial, then of course, there’s no argument against extradition. Because we know it’s not the case, thanks to the work of text producers – sorry, journalists – we understand that this is not the case.

Basically, it’s the risk assessment. Of course, we need to continuously reassess the risk. Currently, the risk is still very high.

Basically, we’re saying, with non-democracies, which is a different set of…That has to be one of the standards that promotes openness. Openness goes together with other open systems.

Or in other words, the idea of openness is based on the idea that people who say that they will follow certain protocols have an alignment to that protocol’s value and is accountable if they do not follow such protocols.

I say this in kindergarten terms, because it’s supposed to be like, goes without saying. What I was trying to say was, during the Sunflower movement, people realized collectively that PRC is doing exactly the opposite.

That seems to be a more solid principle to unite democracy and security. Of course, China is also, it’s not just about ownership. It’s also about various disinformation campaigns. I am very interested about how you try to get the balance there as well.

If I understand your position towards foreign information operations versus the domestic one, it’s about competing rather than takedowns. Is there a legal definition of disinformation?

Yes, it’s intentional untruths that cause public harm.

How do you define public harm?

Anything that harms the public instead of the image of a minister, which is just good journalism. It basically needs to harm, for example, the public health, the election itself, the attack on the democratic institutions, and so on.

For example, there was a trended disinformation that says Hong Kong thugs gets exploited by people paying them $20 million to kill the police. Of course, this is info op. We traced the beginning of this info op actually to a Reuters photo.

This is actually a true Reuters photo. It’s by Reuters, but it says nothing about being paid or murdering police. It simply say there are young protestors, and then it gets recaptioned to be very sensational. Interestingly, this only spreads in Taiwanese social media, but not Hong Kongese social media.

Then the fact-checkers, of course, traced this back to the Central Political and Law Unit of the CCP, of the Chinese Communist Party, in the PRC.

We did a notice and public notice, meaning that we didn’t take anything down, but when you share this, for example, on Facebook, Facebook will tell you that this is probably sponsored – maybe not probably – by the [non-English speech] , by the people in Beijing.

This is state-sponsored propaganda. There’s, of course, people still are free to share it, but they need to frame it in a way that doesn’t harm the Reuters journalists, and also that doesn’t harm the democracy, the democratic system, which would probably get harmed if we do a takedown.

Two questions.

Go ahead.

When you talk about deception there, you’re talking about it has to be deceptive content or deceptive behavior in this case? A lot of Russian and Chinese ops are the same. They just say whatever. They just say, “You can’t trust this thing.”

The content itself isn’t necessarily untrue. It could just be emotional in some way.

In which case, it’s not disinformation. It’s just emotional content. The fact-checkers in the Taiwan Fact-Checking Center use the same definition as the International Fact-Checking Network, the IFCN. There is currently two IFCN member organizations in Taiwan.

They can also cross-check themselves. Basically, I frame it as journalists holding their own journalistic norms. When people claim that something is factual, or is news, or is reporting, it’s held to a higher standard.

If something is framed as, “This is my personal take, my personal feeling about this,” it’s not held to the journalistic standard.

Just so that I understood that, there, you said the Taiwan fact-checkers. Does that mean you have Taiwan government fact-checkers?

No, this is social sector, like they are not paid by the government.

They’re social sector fact-checkers, but Facebook accept their fact checks?

Yeah, MyGoPen and Taiwan FactCheck Center are the two fact-checking organizations that Facebook just simply gets the ratings and the public notice messages from. There are more fact-checking organizations, but Facebook currently work with two of them.

Facebook then agreed to put that fact check on all this content that they have identified?

Of course. It just like this.

First of all, who makes sure that they actually do that? Secondly, how does that affect people who read it? We were seeing in the United States that the use of, for example, Twitter’s use of fact checks on Donald Trump’s posts have almost the opposite effect.

People just discard them, don’t believe them, or say that really just proves it’s true. In other words, why would fact-checking work in Taiwan when it clearly doesn’t work in other places?

Two things going on here. The first is that the journalistic norm around reporting means that, for example, if Facebook, who publicly signed on this counter-disinformation accord, along with other social media organizations, companies, and so on, didn’t do so, it will be very visible.

There’s any number of public dashboards tracing not only what Facebook is doing, but also how Line, an end-to-end encrypted media that nevertheless is doing the same to allow for fact-checking to go even into end-to-end encrypted conversations.

If you long-press a message that you receive from me and send it to fact-checking, Line will do that. It is like flagging email as spam. It doesn’t encroach on the privacy of the email. Basically, by reporting something as spam, you are dedicating this to the Spamhaus, I think, the international social sector for analysis.

That’s the first thing. It’s based, again, on watchdog organizations in the social sector and the implicit threat of social sanction if they do not do what they claim they are doing. That’s the first thing.

The second thing is that if we see the fact-checkers’ notices, of course, that does not automatically win people over. Another strategy that we deploy is called humor over rumor, making sure that people also have the information presented to them, the scientific information sometimes, saying that, “Wear a mask to protect yourself from your own unwashed hands.”

This went literally viral and convinced a lot of people to wear the mask and also keep physical distance measured in dogs – Shiba Inus, to precise – three Shiba Inu indoors and two outdoors. What you’re pointing out is that fact-checking does not, by itself, of course, change people’s minds.

If we make sure that within the same news cycle, people get exposed to both a very humorous message that promotes joyful sharing, which interestingly have a similar R-value to the messages that trades on outrage, but it’s a one-way street.

If you laugh about something, it doesn’t go back to outrage, but when you’re outraged about something, a very well-placed spokesdog can help you to share it in a constructive way. Then it serves as vaccination of the virus of the mind.

I get the fact-checkers are doing their fact-checking thing. There’s a limit to that because there’s not that many fact-checkers in the world. This other side, this humor over rumor, is that like dedicated journalists who have now committed themselves to this? That’s their jobs, basically? It’s a very specialized…

Comedians, yes. It has a history. If you search for 動新聞 or animated news, I think, in Taiwan, you will see some of the very early attempts at using essentially computer graphics. That makes it more humorous and more likely for people to understand.

On the other hand, in the government we also have professional career public service, the participation officers. For example, the Shiba Inu that I just showed you lives with the participation officer of the Ministry of Health and Welfare, so that everything that is announced by the Central Epidemic Command Center gets this spokesdog treatment by the participation officer just literally walking back home.

They live quite close to the ministry, and take new photos, and then posting online. There’s professional teams doing this in ministries as well.

That’s good, then. We’d call that government comms. The comedians and stuff, this is a full-time job, especially like now. Is that the case?

They have their own political party in Taiwan. It’s literally called the, I think its English name is the Can’t Stop This Party. That’s the English name. Literally, it means the Unstoppable Happy Party, Unstoppable Joyful Party.

Their party is a real party. The party logo is the YouTube logo, but turned downward in the triangle. Its party members are very popular YouTubers. They dedicate their time for public benefit, making fun, basically, a more trending emotion, rather than what they call expedition, meaning personal attacks.

They’re doing this pro bono. They’re doing it out of their sense of social service, not like a public broadcaster.

It’s their party platform. They do have a Taipei City Councillor, so it’s a legit political party.

Right, but what Peter’s asking, they’re not in some way inspired, or connected to, or paid by the government?

They’re not paid by the state. No, they’re not paid by the state. Of course, for example, I publicly blessed the Unstoppable Happy Party, but I also made it clear that it’s not an endorsement. I’m not joining their political party.

I’m just saying that the digital communication strategy need the social sector to act in a particular way for that to happen. Not only do the government need to show restraint by, for example, not running government-owned fact-checking services.

Also, if the social sector can actually make things fun, there is less pressure for the government to adopt any authoritarian measures.

Audrey, all this, it sounds exciting. In this brief conversation, we’ve gone through regulatory measures. We’ve gone through security issues. We’ve gone through content creation. Clearly, your philosophy, I can see how it works in all those things.

It’s a ray of sunshine listening to it, because our situation is so dark and so negative. How sustainable is it? How vulnerable is it? Even as we’re talking, I could see so many ways this could be undermined. Will the next government do it? Is this just you?

It’s definitely not just me. All the four political parties – not including the Can’t Stop This Party because they don’t have parliamentary members – all the four main parties in our parliament have signed on the Open Parliament Accord and are working quite closely with the open government principles and so on.

Like the previously, in 2014, right after the occupy, all the mayors that supported this engagement, open government idea got elected, sometimes surprisingly. Everybody who didn’t, didn’t. People compete on how more engaging and radically transparent one can be. To my knowledge, none of the four major parties is against this trend.

Can we embed it more in some way? Can we codify the use of Polis to not just be…

That’s what we are doing. If you look at po.pdis.tw, or just go to digitalminister.tw, and click open government, you can find a set of ready-made regulations, if you want to transplant it somewhere else.

For example, the Tainan Municipal Government did just take that from our central government and realize it in a more municipal fashion, but I’ve really got to go.

You have to go? OK, thank you so much. We appreciate your time.

If we need to double-check something, do we just email you?

Email me, and I will send you my side of the recording, just for your bookkeeping purposes. Also, maybe we can make a transcript, and feel free to just delete everything that you just said, but I will publish my side of things.

OK, great.

Thank you very much.

Thank you, thanks. Peter, will you stay on?

Yes.

OK, thank you. Thank you, bye.

Bye.

Bye.