Can I just call you Audrey...?
Yeah, of course.
[laughs]
Just Audrey.
We’ll have another interview from KBS.
Yeah, I don’t know whether it will be simultaneous, or...
No, I asked to...
Oh, you asked them to?
...because I would actually use the time to have...
I just want to run your profile briefly for fact checking purpose.
Go ahead.
You’re the youngest minister in Taiwan, is that correct?
At the moment, but not historically. There was a Minister of Youth Affairs, 鄭麗君. When she went into the Cabinet a few years ago, she was 34.
Right now you’re the...?
Right now, I’m the youngest, at 35.
You’re the first transgender minister in Taiwan?
In the world.
You never went to college?
Yeah, I dropped out of junior high school when I was 14.
When did you start your first company? At the age of...?
And your IQ is 180?
That’s just a rumor.
That’s a rumor?
My height is 180 cm. It often gets confused.
[laughs] What’s your real IQ?
For adults the IQ test only gets to 160, above which the WAIS Test cannot reliably test. I am 160 or more but I don’t really know how much more. I’ve tested twice and there’s no way to know.
...or more. I see. You started your Silicon Valley startup at the age of 19?
That’s correct.
What happened to the company?
It is still running. It was like an angel investment by a lot of Taiwan companies, and then we continued part of our operation in Mainland China and part of it in Taiwan. It’s still running (benqguru.com), but I’m no longer in charge of it.
You have actually a very impressive record.
Minister, after meeting Dr. 張大煥, he was really interested in your philosophy. Habermas come up with democracy problems who have...I’m sure you know, because you mentioned the deliberation as a keyword. Habermas said deliberated democracy, which is the most important idea.
You’re actually realizing that deliberative democratic system in Taiwan, so we are very intrigued about your progress.
The thing with the deliberation is that it’s magical, because when people enter into a high quality deliberation, everybody walks out changed. Then people respect each other and they will become immune to the rumors, to the propaganda, after deliberation.
The problem with deliberation is two. First, it’s messy. Meaning that every time you run it there’s no established way to get high quality deliberation. Second, it doesn’t scale. Meaning, if it’s 20 people, maybe you can get deliberation. If it’s 200 people, it’s impossible. If it’s 2,000 people, people just start shouting at each other.
Basically, listening, as the key to deliberation, does not scale. Meaning with more people it actually loose it’s magic. We use ICT, information communication technology, to scale the experience of listening. That’s our main contribution.
In that process, do you actually allow the netizens into listening process?
So about your program, your software. It has been developed in order to gather public opinion?
Yes, of course. There are these very important principle, in the idea that we use the Internet to deliberate people’s feelings, not their suggestions, not their decisions. Because if you jump into suggestions, people will not reveal the feeling they have behind the suggestion.
Even though five people say at least the same thing or they vote the same way, maybe they do it out of different subjective feelings. What we have outsourced before was people’s feelings, for example, around one concrete issue like you own a private car and you don’t have a professional driver’s license. Then you use this app, and then you get a stranger on your backseat and you drive them somewhere, and you charge them for it ‑‑ without mentioning any company name.
[laughs] Yeah, exactly. Then, our software is an open source software. Basically, ask people one yes or no question, and based on their yes or no question, they get clustered. After you answer a question, whether you think insurance is important, whether you think it’s public safety, your avatar moves on the screen so you can cluster with people.
This interface has two very important thing. First, it shows your Facebook and Twitter friends first, so you can see if people having very different idea than you, they are not anonymous enemies. They’re still your friends. It’s just you haven’t talk about this over dinner, so it’s not your enemy. That’s first.
The second thing is that in Internet forum it’s very easy for these to politicize, to dominate all the discussion, and these people are silent because they’d shout...
They’re always polarizing.
Yes. The thing here is that you can press consensus and see there’s nevertheless 91 percent of everybody agrees the importance of checking of the driver’s credentials are protecting the passenger, so there can be consensus among disagreements.
After you answer a few yes or no questions, you can also type in your own feeling, which then gets voted by other people. What we said before the deliberation was that if by the end of the process we have a set of agreements, more than 80 percent of people agree on something, then we agree to use this super majority consensus as our national agenda to negotiate with Uber and with other taxi.
May I interrupt, Minister. How do you gather peoples’ own opinion? Like you said, you didn’t listen to their suggestions, you didn’t listen to their rational opinions. You just captured their feelings or their innate tendency to a specific policy, so how do you gather information?
No. We just ask people to start the sentence here with 我覺得, which is, "I feel that..." and then that’s it. We’re not passively monitoring. We’re actively asking people to write their authentic feelings, and then, this is an open‑ended survey, so people can feel anything.
Within that process, people might really disguise or didn’t want to show their real feelings, so you might have to ask them twice or more times.
That’s right. You can change your positions anytime here, and then you can state a more nuanced feeling, because we only take as finding things that you managed to convince everybody. If it’s a disguised or if it’s a non‑authentic feeling, if people don’t share it, it doesn’t have any effect.
People are basically competing to first convince people who think like them, and gradually to convince people who doesn’t think like them, and we only take the supermajority as the agenda.
How do you call that system?
This system has a name called Pol.is, P‑O‑L‑.‑I‑S.
You worked hard to develop. You say it really taken a lot of resources to develop...
It is actually a community effort started by some Seattle hackers, but it is open source so everybody can contribute, so I also contributed to its development. It was contributed usually by people who participated in an Occupy or a protest movement before.
After the Occupy Wellington, the New Zealand people would join, and after the Occupy Wall Street...Every time after an Occupy, people start thinking, "OK. Democracy can take another form," and then these people will join this development.
You’ve been a month before your team in Taiwan was in the 2014 Sunflower Movement.
That’s exactly right. We started this development before the Sunflower, but we were not many people, a hundred or so, but after the Sunflower it was just exploded in interest.
Just quickly, is this platform are you using for the Taiwanese government...
Are you using this platform for Taiwan’s government to gather public opinion?
That’s exactly right.
Can you give me just one example of the latest issue that you’ve got the idea from this from?
Mm‑hmm. The most high profile of course was the Uber case, but we also use the same for the Airbnb case, for Internet sales of alcoholic beverages, and so on. Just today, Uber announced that they will bind themselves to the consensus that’s made here, and they will change their business model starting tomorrow to play by the rules that were set in the deliberative fashion.
How many people participated in the Uber survey?
In the Uber survey thousands of people participated, but only around one thousand or so logged in so that we can see their avatars, and they have answered more than nine questions so we can determine their position. It’s about, I would say, about one thousand active players, but about nine thousand or so are peripheral players.
Actually today I’m interested in two of the initiated from your department, and one of them is the eSport. I’ve heard that your department’s planning to replace the mandatory military service to playing computer games. Is that right? Can you tell also more about this and why you are doing this?
Right. First, it’s not about computer games. eSport is a sport. It just happens so that the stadium they use is a virtual one, not a physical one, but the point here is the computational sport aspect.
My favorite argument in this conversation is that I think Go (圍棋), the board game, is now a eSport, because most of the competition and training happens online, and also because machines now play better than human, [laughs] so this is mostly for two human beings to enjoy each other’s company and play for performance, and in this sense, it really has nothing different from other eSports.
As for the alternative military service, it is true that it’s already the case for Go, for 圍棋 players in Taiwan. They can have alternative military service if they win some international or national competition awards. The same applies for any other sport as well as like symphony conductors and other cultural performers.
An e‑Sport athlete being a cultural performer, and which has international competition and so on, shouldn’t really be discriminated against. That was just the basic idea, so we’re not actively promoting in. We’re just saying this is like other sport.
The other thing that we used during the e‑Sport thing is that we have on our previous website a accountability trail where you can have this QR code that lists the entire decision making process for how exactly this policy was formed.
Starting from the public hearing, down to the classes, down to the military service, down the national tax breaks, and so on, we have all the ministries and all the decision makers in our internal meetings taken as a transcript ‑‑ was real time transcript ‑‑ and everybody, after editing it for 10 working days, agreed to publish it to the general public even before we have set our direction.
This is very important, because then everybody, that use for athletes or the stakeholders, look at our conversation and deliberation records and let us know over the Internet where we have missed, where there’s legal precedent for that, and so on, and so they were contributed, and I quote them on the next meeting and so on.
Everybody knows that’s when all the three lights here goes from yellow, which means it’s in progress, to green, which means it’s resolved. We have resolved this policy case, and it’s one piece of the puzzle, it’s the military service, but there is also the educational aspect and also the industrial aspect. These are a complete picture, this is not just one isolated policy making.
What are the two most popular e‑Sports game in Taiwan, and how far you have to perform in order to be replaced the mainstream military service?
There are 10 spots with a firm criteria. Of course, if you win an Olympic level competition, that automatically qualifies, but even from a sport like LoL, there is actually no frequent Olympic level competitions.
The next thing of course is the alliances between the different countries. We’ve set a basic criteria that says it has to be regular, and it has to have at least this many people in its regional and this many people in its national level, and so on.
If you qualify for that ‑‑ there’s a few competitions for LoL that qualifies for that ‑‑ if you win the first three places or so, that also qualifies. Then, after that, it was about a national level competitions which still has to have some kind of structure.
This is a three‑tier system, but still, only the professional players get to continue their work during this kind of alternate service, so we’re not rewarding video game players. We’re rewarding professional athletes.
Another topic we are interested to know is that your country is starting to educate children about how to identify fake news at school starting next year. Can you tell us why you think it’s important to educate that?
I think it’s important that we don’t use the words "fake news," because I think it is a disservice to journalism. [laughs] There are rumors which are viral, which means people who spread it doesn’t bother to check, but it’s not necessarily false or fake. Some rumors turn out to be true.
The point here is that people need to learn to do basic source checking, fact checking, and balance of a diverse set of opinions and so on, basically critical thinking and independent thinking. In our curriculum, we call it information technology and media literacy.
This idea is not about spotting "fake news," but about being able to make decisions and make opinions of oneself during a learning experience so that people don’t guess away one way or another based on rumors. That’s the basic idea.
The other thing of critical thinking is being able to listen and change one’s mind or opinion when new facts surface. Again, this is kind of against human nature, so we need to get started very early on and say, "Standing to be corrected is actually a virtue. It is not about losing face or anything." This is the other aspect I think are very important.
Starting, what age would Taiwanese students...?
The entire K‑12 system, that is to say starting from the first grade all the way through senior high school, because this is not a class. This is one of the literacy that need to be merged into all the classes.
All the classes, the teachers and the textbook makers and schools need to find some way to find media literacy and critical thinking into the way they are teaching, and not just taking it as a class, as a examination.
I’ll have just one more. What was your reactions in your country when you were first appointed? Number one, you’re the hacker, so people may be worried that. "How can a hacker be working for the government? Hacker can attack." A hacker is usually seen as somebody who attacks the government.
Like lawyers?
Yeah, or maybe lawyers...
(laughter)
Number two, in South Korea, we had not a high government official to come out of closet, not to mention publicly transgender, it would instill huge outrage with the constituents. Do you have those outrage in Taiwan as well?
Not at all, because first, I started working with public service in 2014, after the Sunflower Movement. For many career public servants, I’m already very well known as one of the people who helped them to communicate, to facilitate conversation with citizens.
In many cases, in people in administration, they call me kind of an understudy minister already. For a couple of years before I joined the cabinets working for the previous minister of cyberspace, Jaclyn Tsai.
What I’m saying is that if people had some outrage or whatever, it already happened 2014. It’s not waiting after I become a cabinet minister. That’s the first. The second thing about hackers. I think in Taiwan, because cyber security is very, very serious matter, people put a lot of understanding into it.
They don’t really have this magical thinking that think hacker are somehow magicians. They understand this is just one kind of technology, and technology is just like any locksmiths or whatever, has their pride, has their professional code of conduct, and things like that.
Just like when people know a lot about law, maybe they exploit a loophole of law and become very bad lawyers, but they don’t have to be, right? They can become lawmakers that makes a better kind of law that is not going to be exploited.
I think it’s much the same. I understand information technology, but it doesn’t mean that I will exploit its loopholes. It means that I will set up a system that doesn’t suffer from the same kind of loopholes.
Since you’re at the hacking competition, is it getting more important these days for the government to foster more hackers like yourself?
Yeah, verily. The first thing I did after joining the cabinet was recompile the Linux kernel to add secure computing to our internal government information facilities, which is the system that we use every day.
We then invited penetration testers, which are security experts, who tried to attack the system and to find the loopholes, so that we can rest assured that when we are using the system, we configure it properly and so on.
We take this very seriously, and we believe that active penetration testing and forensics, and all these tools, enables us to have a clear understanding of our information technology’s infrastructure and its limitations, so that we use it in a responsible manner.
Otherwise, all would be just voodoo and hearsay, which is pretty much the worst case, because then people will take some countermeasures that then cause side effects and so on. I think a clear understanding of the cyber security landscape is for everybody’s benefit.
What about the growing intergovernmental or allegedly intergovernmental cyber attacks? There is allegations of North Korea attacked South Korea, for example. There are allegation of cyber terrorism between China and the United States. Does that make this more important to have more hackers in the government?
Yeah, I think just like with any other terrorism, the impact itself is actually low. The thing that they hope for is to provoke some mysterious fear, and then have people overreact. The overreaction, actually, causes more damage than the initial attempt, that cyber terrorists are a real‑world terrorism.
It’s like a mosquito trying to get onto a bull, and then to break things into a store. The whole idea is to have a calm and clear understanding, and have cyber security experts explaining in a very clear, accessible manner what exactly is happening.
Then people would not panic, and they will not clamor for things that will actually cause more damage. I think literacy, again, is the most important thing here.
That connects with your education about the media literacy. Just one more about...I think there’s some Taiwanese people who believe that, the rumors are circulating in Taiwan, maybe behind that there is China.
Do you think that’s one reason that Taiwan should increase its education in media literacy? Because of China...
In Taiwan, there are many ideological camps. There’s pro‑unification. There’s pro‑independence. There are everything in between, and I think these camps are more than sufficient to generate credible rumors without being remote controlled by any other places or any other groups.
That said, I think it is important for people to get inoculated against this kind of virus of mind and deliberation, as you already mentioned, I think is the key for the inoculation, because any effect, fact‑checking, or whatever, after a rumor has already spread only has a containment effect.
If people have already considered carefully even the position of people they don’t agree with, they already have some kind of inoculation in their mind, so that they will not fall victim to rumors. I think really, this kind of early‑stage prevention is the key here.
Just by getting people to listen to the ideas that they don’t like, basically, develops your immunology system.
I think your view on this fake news ‑‑ I only use this term because that’s what my readers are familiar with ‑‑ it’s very refreshing. Could you comment on the US, US President Donald Trump, about how the way he communicates with the public?
Do you think the way he names some of the news media as a fake news outlet? What do you think about them?
When Twitter first got invented, people used Twitter in the way that’s very direct, and make full use of the 140 characters. Taiwanese people used Twitter more like regular blogging, because in 140 Chinese characters, you can say a lot, [laughs] but in English, not so much.
I think there’s a distinct style difference in the ways that we Twitter versus a English speaker use of Twitter. They had to be very brief, and basically just say even just one idea or half of an idea.
As things developed, because I was in developing social media networks, we see a lot of very creative uses in Twitter, like multi‑part tweets, like the tweets that are complementing each other using hashtags and conversation communities, and so on. They built a community structure out of those very short forms of communication.
@realDonaldTrump used Twitter the way when it was first invented, [laughs] used it as it meant to be, so it’s also very refreshing to me. I learned a lot.
What do you learn from his way of using Twitter?
Basically, he had self‑contained messages.
Self‑contained...?
Contained messages that doesn’t rely on context.
Can you elaborate on...? [laughs]
Can I elaborate?
I think that was pretty self‑contained. The idea is that if I have make a very long speech, where each sentence depend on the context of the sentence before and after, then when any one of those sentence gets taken out of context, it will be turned into something that I don’t already mean.
If I make all my messages self‑contained and short, then there’s no danger of being taken out of context. This is the basic thing that I learned, not necessarily from @realDonaldTrump, but it is the idea of Twitter at its beginning.
I don’t want to take all your time, Minister.
Some things in your career path is quite very impressive and somewhat special. I heard that you learned programming by yourself after dropping out of mid‑school. What would be the reason?
I learned programming when I was eight years old. That was seven years before I drop out of middle school.
Sorry.
Would you like to translate? Is English OK?
No, you just speak in English, and I’m going to record it.
That’s great. Yeah, I started very early on to take an interest in mathematics, and I started reading programming language books when I was eight. They described a way for people to very quickly get results with mathematics without having to do all the calculations by hand, and it’s much less error‑prone.
I got very interested. I didn’t have a computer, so I basically just took pencil and paper and drew keyboard and typed on paper, and then wrote what the computer would say and so one. Basically, I drew a paper computer.
This means that computational thinking is one way of thinking. It doesn’t depend on the machine, and if one structure oneself in some way, some part of the thought process can be automated.
For me, it is like learning a music instrument without having access to the piano or to the oboe instrument. The idea is just to learn the idea of melodies, and chords, and compositions, and so on, so that people can play this kind of music.
After a few months, I started programming on a real personal computer, and then I found that those notes that I learned as programming languages actually makes melodies that are shaping spaces, around the spaces in which people can interact with each other.
The first programs that I wrote are computer games where people can play with each other, against each other, and so on, so that also characterized how people behave around each other by creating a space of people’s interaction.
This is a very special kind of instrument, I would say, while learning it is just like learning any other music instrument. The notes it creates is actually a new kind of space where people can interact.
You’re a civic hacker since you are 33, right?
What kind of activities did you focus on, and what were your goals at the time?
I’m always a civic hacker. I started working on Internet’s human ride and freedom of speech movement at least since 1996. That’s during the Blue Ribbon Campaign and so on, so yeah, I’m always a civic hacker I would say.
The primary thing, of course, is the sharing of the knowledge and making the process of sharing also a common knowledge. As early hackers, we all believe that the tools that we use to shape our lives is more and more taking over the role that was initially taken by laws.
More and more we’re living in a case where algorithms, where computer code already determine what is possible and what is not possible even before the legal code kicks in. The problem being that computer code was not transparent.
And people using the computer code doesn’t really know why it’s went to here and has no freedom to modify it or to distribute its modification. I would say that my core concern is still this kind of Internet freedom, with software freedom being one part of it, but also freedom of speech, of assembly, and of other human rights.
Thank you. The last question would be why do think cyber security matters the most in the era of the fourth industrial revolution and if there’s any example you can give of this.
The security matters a lot, because if people are still using Internet for freedom of speech and of assembly, people must know for sure, not just blindly trusting that the speech they make on the Internet will not be used against them.
If the Internet becomes a technology that oppresses people, then when people are going to have assemblies or have to express their freedom, they will go back to pen and paper or to face‑to‑face meeting, and the Internet will lose its original liberating promise of getting people who feel that they are alone and they don’t have people who think like them into communities.
Basically, if we don’t do our job right on security on the Internet, the Internet becomes a fragmenting force instead of a community force. If it becomes a fragmenting force, people will have a lot of small Intranets, like a balkanization of Internet, and each Intranet will have an exclusionary policy that excludes people who don’t think like them.
You will not foster a cross‑community dialog and else will become a much more fragmented space where people live essentially in their own realities, and that is as undemocratic and as non‑humane as we can imagine if we extrapolate these trends further.
I think the utmost importance is to get people to still see Internet as a secure place by having the literacy of telling insecure communication with secure communication, and keep to the secure communication methods so that people can still form communities where it still becomes possible to talk with strangers, to learn from strangers.
This is another question about security, because we have now self‑driving cars and AI‑level at so many things, and he would like to say to us some words why is it specifically more important security in AI and self‑driving cars.
Security, this word in English is very interesting, because it describes both a property of a system and also a state of mind. I feel secure versus a secured system.
In Mandarin Chinese, in the language that we use, we say 資安, the 安 meaning "at ease," a calmness with information. This is important when I’m sitting in a self‑driving car, for example, because if the system is mathematically proven, it satisfies a lot of properties, all the engineers, everybody, vouch for it.
If their proof‑carrying code and all those advanced verification mechanisms cannot translate it to the language that everybody, common people, understands, they may achieve the mathematical definition of security, but there is no psychological or sociological definition of security, of feeling secure.
For me feeling secure is the goal, and this is just the mechanism, the instrumentation for it. On the other hand, there is a lot of what we call security theater, which means that not going through the rigorous mathematical proof, but just to enact on theater and trick people into believing, into feeling secure about something that’s actually not very secure.
If you go to those security theaters for...this a lot like rumors. If you go to a lot of security theaters, people lose the ability to tell a rigorous system versus a non‑rigorous system. I think this is of utmost importance.
Because if people cannot tell a rigorous system versus some non‑rigorous system, people will not ask technology what they want to feel secure. People will just blindly accept whatever the vendors push to them.
This creates a false sense of security, but then, when cyber terrorism or other attacks happen, people will panic because they did not know that their system, which they rely on, doesn’t have the property that they imagine that it have.
Again, as the AI and IoT technologies becomes part of not only our life but become part of ourselves, it is very important. Just like the medical knowledge enables us to know what is happening with my own body, my extended body also needs the same kind of self‑introspection and knowledge so that I can feel at ease with an extended part of my body. For many people, their phone or their smartwatch is already part of their body.
Do we have time to ask one more question?
One more question about these answers.
Go ahead.
Like you just mentioned, the autonomous car, the people...I mean the producers of the autonomous car will always say that this is a safe. It could be, like you mentioned, it’s a security theater, because we do not know that the autonomous cars will be safe or not.
We cannot really tell. Technology is so complex. Artificial intelligence technology cannot really interpret it through people’s daily arrangements. Perhaps you could understand it...
So how can we assure that the autonomous vehicle is really safe?
First we can do it through simulation. You can invent some simulation scenario, and try to put the algorithms in and see if it behaves the way you think it should behave. This is how actually we test real human testers.
Human testers describe the situation to human drivers, when they’re taking the drivers’ test and see whether they have the good...and this is essentially the same that we put the pilots also into in their cockpit simulations.
Basically, as the spirit of Turing Test, if AI can handle all the situations the human can imagine, and through at it, at a simulation, then we can judge it based on how it thinks and explains itself and their stimulation scenarios.
At the moment, there’s still always somebody behind the wheel to take over when the machine fears that they don’t have the confidence to handle the situation. Another very important thing is the human machine hand‑off.
We need to have augmented reality system projected on the window of the car to have the driver know what the AI are thinking about the situation, and not just very low bandwidth SMS‑like messages, because then it will be too late.
It’s very important to have this co‑piloting mechanism between the AI pilot and the human driver for the foreseeable future before we can even think about going to fully autonomous driving.
Do you mind talking about some personal life?
Go ahead.
What was your family’s response to your decision to go through the transition from man to woman, could you tell us about your plans for your family, whether you have a plan to start a family?
My family is really non‑rigid as far as gender stereotype goes. My mom was very expressive and was raised in a very androgynous manner, and so they don’t tell me what a man or a woman should or shouldn’t do.
Instead of transition, I would say that I’m just post-gender or post-genre, meaning that I don’t think there should be things that only one gender should do. I think anything that is declared is discrimination. I’m more against discrimination than against any kind of particular gender identity. That’s the first.
When I developed this kind of thoughts, when I was an adolescent, I think they think in more philosophical stance rather than a psychological thing. I think they’re pretty friendly, and we keep dialoging.
Even when I decided to change my name, I made sure that my new name is OK by my father, [laughs] and that he also think it’s a good name before I changed my name into it. As for family, I already have a family. It’s just not human children. [laughs]
I live with seven cats and two dogs, and have been living with them for 10 years now. It is a very large family, it’s just not human children, and I do plan to continue living with them.
Two cats and...
Two dogs and seven cats.
Seven cats. Oh, wow. I have two cats. [laughs] I consider them my children as well. [laughs]
Yeah. Any final questions?
Thank you for all the...very interesting.
Thank you. It was a pleasure for me as well.