The French equivalent, you mentioned her recently, and her predecessor, Fleur Pellerin. They mostly seem to make the link between the government and technology companies, including small startups and stuff like that. It’s an economy-focused position. This doesn’t seem to be your case.

I do that, too. I’m the digital minister, but I’m also in charge of, in addition to open government, social enterprise. Social enterprise, by definition, involves the linking between the public sector and the private sector because they want to be sustainable, but also the civil society, because they want to solve problems that actually affect the society.

It is a link, but it includes the civil society at least as important as the private sector. I strive to balance between the social and the economic interest, but I wouldn’t say that I don’t touch the economic.

It seems that a lot of what you’re doing is open data and...

Open knowledge, open collaboration, open spaces...

Trying to create a public discussion.

That’s right.

That comes also from the self-awareness? There is a strong public desire for that.

Of course, there is a strong public desire, and also because Taiwan is pretty unique in that we have one of the highest network readiness, meaning that pretty much anywhere in Taiwan, if you want broadband, you have broadband.

How much of the population, you know?

It’s around 98 percent, if you remove the offshore islands and so on, but this is network readiness. It means that broadband is there if they want to, but it doesn’t mean that people actually go and use it. For that our numbers are pretty...I think about 78%...

78% of what?

Of the population, which is roughly similar...

Has broadband or...?

Uses broadband, which is roughly the same as France, far as I understand.

That’s right.

That is problematic when you want a public debate that happens only online.

Yeah, of course. If they are three years old, I wouldn’t worry but [laughs] if they are not, they are adults, then, of course, it’s a case of concern. For Taiwan, we have developed what we call assistive civic technologies, meaning that, for example, for one of the deliberations around Taipei’s social housing distribution.

The mayor said that the stakeholders themselves, must determine the agenda. The stakeholders are, of course, homeless people, mentally and physically handicapped people, single parents, people living with HIV, and so on. They could not get landlords to rent to them even if they have the money because they were stigmatized or something.

By saying these people must determine the agenda, it of course doesn’t mean that we need to set up a website and ask all the aboriginals and everybody to go on their website, that would be a catastrophe.

What they did, in the Taipei City, and I observed and helped out a little bit, is that we used a traditional rolling survey method to discover stakeholders to draw their family diagrams in standard, social workers fashion, and then try to go beyond the proxy that purpose to speak to them, but actually reaching them.

The problem is, of course, we can’t get everybody like the one who are paralyzed and so on, to get into the same city hall for a discussion, which is why we still use technology. We use live streaming. We use 360 video. We use sign language interpreters, real-time captioning, and also for people who can only type and not speak, there’s the other way of channel as well.

These are, of course, the very basic to get them to understand that it is really the Mayor’s will to be bound to the agenda that was set by the stakeholders and not just random, general public, who may or may not, understand what they are about.

It’s quite magical because after a few this kind of virtual, or real face-to-face discussions, they come to see each other as comrades instead of someone who won a slice of pie bigger than they are. We even get an aboriginal mother who said that those artist people, they deserve this more and they are willing to pick their fights elsewhere and so on.

It does have a good, coherent effort. What I’m saying is that technology...

It unites the will.

It unites the will. What I’m saying is that technology is a necessary but never a sufficient foundation. For things like these socially disadvantaged people, the technologies must behave in an assistive way. That is to say, to enable them to speak for themselves, instead of creating technology as a proxy to speak for their will, which never works.

Quick question. What do you think of online voting? Voting for general elections?

I think voting for policies is easier online, because it’s easier to have a fact-based or evidence-based discussion on policies. I don’t know about voting about people. Voting about people, in my experience at least, is seldom based on an actual, real connection to the person.

Of course, you can have any number of Q&A forums and so on, but it doesn’t really mean that you really know this person. I’m much more in favor of voting to determine their priority of policies and so on.

That’s the idea of POP, that even if you vote for a party or a person, the agenda sets only part will correspond to what you want. Then, once they get in power or not, they have to change their agendas to fit with other politicians. Then again, maybe you can elect someone who will enact none of the policies you wanted...

Right. It’s an intermediary problem. We know this very well.

Not about how it’s used, whether it’s for policies or for people, the very interface. The fact that the voting could be online. Do you think it would be dangerous with regards to...It has two big problems. The first is that 20 percent people who don’t really have Internet access might not...

Yeah, might not participate. There are ways around that. We’ve heard that there are countries who’s considering automatic teller machines, ATMs, as a voting booth. That increases the reach of Internet a lot. If you can use an ATM, you can watch the policy description on the ATM screen and then, press buttons to vote.

It, also, doubles as a cheap, not exactly secure, but cheap authorization and authentication mechanism and so on. What I’m saying is that if we introduce technology and the technology, strictly speaking, expands people’s reach, for example, people who cannot go outside can now also vote.

That doesn’t take things away from the existing people’s mechanism. It’s not like we’re taking those paper ballots away anytime soon, right? As long as it’s complementary, I don’t see any dangers in introducing new vehicles but I wouldn’t say it replaces the original.

The problem is that you do have the security critical point...You know, what happened in Estonia. Their voting system is catastrophically vulnerable but they’re still using it.

They say that they have a few defenses. You can always override your previous votes. You can always go to the physical booths and override your online votes and so on.

Yeah, but even then, you remember the team, the...

I do.

Yeah. [laughs]

What I’m saying is that, of course, it has its security flaws but it was designed so that flaws in one layer doesn’t completely turn over the legitimacy of the whole system. It was designed that way. I’m not saying that we’re adopting the Estonian system.

Because the thing is that, actually, to answer some of your questions about RSV, that’s one of the nice features is it’s an online system that doesn’t have to be online. I don’t think you have time to read the documentation on that because it’s long.

I have some basic ideas. There is sampling. There is very interesting fake tickets. There is this washing of...making it not pay to bribe. I got that main point. I haven’t gone through the mathematical descriptions.

Really nice mathematical feature. Are you familiar with ThreeBallot?

Mm-mm.

The mathematical feature is that it’s mathematically secure, that you can’t hack the system. If you manage to hack the system, the best you can get -- and it’s hard to get -- is the identity of the people who vote, and not what they voted for, and you can’t change the result. It’s end-to-end verifiable.

I understand that much.

At least we got the security down. The thing that we tried recently in practice -- and it works quite well -- is that you can reach the people who don’t have Internet. We were at a conference, and a lot of people there didn’t have Internet.

The thing is that you can vote through a proxy but without actually telling them what you voted for, and while being able to verify that they voted later. Voting by proxy could be a very nice way to access the last 20 or 30 percent.

That’s good because then you don’t have to trust the dealers of automated teller machines.

Yeah, you just tell the people, "Well, if you want, you can vote online. If you can’t, you can ask anyone in your family, or even come to the City Hall, and anyone will help you vote, and will not know what you voted for, and you will be able to check that your vote was counted."

As a way to do that, it could be the future. The problem that we have is that how do you get people to trust that?

Exactly, because it’s a sociology problem. The usual way for a high stake election is that you get extra for all the parties, and then they jointly witness the process and earlier process beforehand. It’s delegating the trust to the parties, like five or six parties.

It doesn’t quite work if you...for example, I have one large party, one medium party, and several small parties. There is always room for manipulation.

Do you know the error rates in the paper ballot voting?

I don’t know.

France has one of the best systems. We really are secure and everything. In the end, we still have about 10 percent of our counting offices reporting errors. But it’s each time one or two ballots, so really doesn’t change anything. People really overestimate the correctness...

Here, we allow live streaming of the opening of the ballots, the reading of the ballots and by individual verifiers, so that doesn’t happen as much.

That is nice. When did you implement that?

Last election.

Which was...?

Which was last year.

I don’t think we have that. I think it’s too late to implement it. The election is in a few days.

I’m not interfering with any other country’s elections.

Anyway the margins -- the compounded errors -- less than 0.1 percent, so it’s nothing.

Fingers crossed. It may come to that. [laughs]

That would be surprising. It would.

The problem is that for online voting, people could be convinced, I’m guessing. The fundamental of RSV is that we take a sample group. You take 10,000 people. That’s what allows you to have policy questions every day if you want.

If you ask 10,000 people each time, and they’re random, in Taiwan, they will get one question every one year, two years. It doesn’t matter. You can actually fiddle with it so that people have a reasonable amount.

There’s probably a sweet spot before people have election fatigue, but also not too long away between two votes so that they don’t feel disenfranchised or left out. The problem is, do you think that people will be ready to accept that?

They’re doing that every day on social media anyway.

What?

They’re doing this kind of discussion on policy on social media anyway, with their friends and their friends of friends.

Yeah, but letting some group of people chosen randomly choose for the rest? It wouldn’t be a referendum. It would be...

Of course these people’s families, and friends, and so on, will also be involved, because these people will ask their friends and so on. It still ends up being a society discussion.

That’s the goal of POP, is to say even if we use RSV -- we’re considering it -- if we want to have the legitimacy of 10,000 people choosing for the rest, then the only way to have that is to have the rest informed, it turns out.

How do we get that? We get that by a public debate. You’ve been using Pol.is?

That’s right.

That’s one of the things that’s really interesting. I looked how to set it up and...

Identify the principle component of a diversity of high dimensional opinions.

It works well in practice?

It works very well in practice, scales to thousands.

Hmm?

Scales to 10,000 very easily.

Yeah, the problem is, does it scale to 20 or 60 million. [laughs]

Maybe. We haven’t tried that.

That would be...

The visualization is designed so that it does clustering, and also allows you to seek consensus among clusters. Because the system rewards people who can get more consensus in their groups, and also rewards people who can propose things that are consistent among groups.

How does it reward them?

By virtue of being shown, it’s just like Reddit. It’s really not that magical. For example, this is the Uber X case. As far as we can see, the first thing it shows is the global consensus. We can see that it’s accepted by pretty much every group except for this one, which has a little bit of dissent.

Here is where I am a bit sad not to speak Mandarin, but [inaudible 17:52]. [laughs]

Sure. Anyway, the point is its interaction, right? If you click to a group, then you can see this group’s consensus. Or, this group, which has a consensus that pretty much doesn’t agree by this one and this subgroup. Then, this subgroup. Within each group, you can see its consensus.

The majority opinion allows opinions that are, pretty much, everybody agrees or disagrees with. This experience social media because the social media trying to reward things that are fringe, like these or these opinions. They become very loud. This rewards the majority because it’s...

There’s actually a metric in there.

Yeah, sure.

OK.

Yeah, it’s very easy to calculate because everybody can propose a yes or no question.

I’m just wondering what the metric on that graph...?

Basically, what it tries to do is a dimensional reduction between everybody’s point inside that dimensional space, and then reduce it so that a principle component of a vector gets most representative of the most divisive point becomes the X-axis and the one most orthogonal second sub-component becomes the Y-axis and that’s it. It’s very easy. There’s no magic in it.

There is one piece of, could be magic like any sufficiently advanced technology is, which is how it finds that biggest vector.

Right. The source code is there. It’s written in clojure, I think, but the math is pretty easy to go through and it’s open source, so feel free to...

Yeah, I read a bit, some documentation... it’s a representation of high dimensional data in low dimension, but is the vector basis the best one? Because, of course, they have to estimate the bases through algorithms, which, I don’t know how they do it, obviously.

Is that basis determination algorithm legitimate? I mean, it seems to work, so... [laughs]

It seems to work. It’s a game, of course. It seems to work. If somebody proposed a new question that polarizes everybody, then that becomes one of the new vectors in the principal component.

You can see the group shuffling just by somebody proposing a really uniting or a divisive opinion. It’s very dynamic. When this thing is ongoing, you see everybody’s position changing all the time, which has a good psychological effect, at first. It’s all your friends that’s showing here, so they are not your enemies.

Yeah, I saw the preference list and...

They’re not your enemies. They just didn’t talk about it over dinner . The other thing is that it shows that it is possible to have majority opinion even among people who are very divided. This part is the thing that we don’t usually see in social media.

Yeah. It is. I am quite in favor of approval systems rather than ranking systems, because what happens is that you find the consensus instead of finding the...

That’s right. The best or the highest ranked.

As long as people are honest.

Right. Of course, these are not anonymous votes. If you log in, your position is shown to everybody else, so they have a kind of moral peer pressure to vote honestly, because their Facebook friends see it.

That was the position I explained in a paper that I should submit now, by the way. [laughs] That one way you’re getting rid of the inherent problem of manipulability in approval systems instead of ranking systems, is just getting rid of anonymity.

Exactly.

Are we ready to get rid of anonymity?

For cases like this, of course, because it’s essentially idea gathering. It’s brainstorming stage. It’s not decisional and it’s not electing a person. For brainstorming, of course, we can be...

Do you know, I read that recently about the French parliament? That the people don’t have secret ballots except for a single reason. There’s a single type of vote where they have secret ballots and it’s to elect the leader of the Assembly for...

...and it’s, as long as you vote on policies and not people, it’s...

It’s reasonable.

Yeah. If someone votes on something and it might have been influenced or corrupted, then, at least, you know that they voted for that.

Exactly.

You can check.

Exactly. I see the reason for, say, a referendum to be anonymous, but for things like MPs or a small, deliberative setting, I can very easily argue that it is the process of someone’s position changing that actually chronicles that person’s ability to have a conversation rather than just forcing...

Like everything was dark room. The vote was anonymous. You don’t even know how the debate went.

Yeah, and I think that for elected members, like public officials, anonymity is not a good thing because they have to show what they did to the people.

No, it’s not. Right. It’s good to have private discussions but when it’s binding, it needs to happen in the light of the public, of course.

The idea behind POP (http://poplatform.org/) is to have such a platform where people can discuss, and not just that but to hack democracy through it, by trying to get elected officials, or getting people to be elected officials for that platform, who are bound, by their word, to actually vote the way that the platform, the gathering, voted.

One interesting thing about Geza Tessenyi. He’s the person behind the idea. Is that he’s very interest on that, I find quite interesting, is that the person who’s chosen as a representative, who’s elected, can debate all they want and fight really hard against the ideas.

As long as they vote, they have a freedom of speech but they don’t have freedom of choosing what they vote for.

I am aware of this school of thought.

Yeah. It’s nice. For that, we need a public discussion. We need a platform. We had one called the République Numérique.

I know.

Are you aware of what...?

Yeah, and it’s not open sourced. I tried to get them open sourcing it but they didn’t, so we ended up writing our own.

Wait. How did that happen?

The vTaiwan system was open sourced, under creative common zero, it is public domain from the start. We really liked republique-numerique.fr for its ability to accept discussion to every single line of the text. We didn’t have that. We had a discussion board and we manually tested...

You couldn’t integrate that kind of thing?

That was before I was Digital Minister. I went to France. I was invited to La Nuit des idées. I think that was last January (2016). They were just discussing the République Numérique bill online at that point.

I saw this system, and it’s a very nice system, and I checked with the developer, I think, also last year, when we went to the same Nesta Event in the UK, in the parliament, where we talk about our respective projects. Of course, the Icelandic project about...They have this system where the pro and con are voted or ranked differently so that you never reply to each other.

Instead, you choose the best pro argument and the best con argument in the petition and then, the debate centers around the best, cream-of-the-crop, pro and con argument. It’s two-agenda setting, pro and con, this idea and the same person may propose pro and con.

It’s not a yes or no vote. It’s a vote of ideas. They were bound, so that they must discuss the best, say, three pro and three con ideas from the platform. I think that’s pretty good. This is open source. Our vTaiwan system, including Pol.is is open source but the Republique Numerique system from France, not so much.

Why?

Because they have a business model, I think, around selling this to local cities, customizing it and something. I don’t really know. It is an up-front cost to open source something. You need to clean the code base or something. I don’t really know. Scrub the code base.

We did adapt the best ideas from the Icelandic and the French system into the next version of vTaiwan and join platforms here in Taiwan, so that in a few months, the Taiwan petition system will also have this pro and con interface.

The vTaiwan system now, has this line-by-line, section-by-section, discussion built in, we basically rewrote from scratch because we thought that was a better user experience.

I think it was the people from my lab were quite involved in that thing because, for quite weird things because they were against quite a lot of the proposals because it put into all that, for example, the scientists.

Our work is in mathematics and such. We have the freedom to share it however we want, to put it on arxiv and such. After, I think it was, six months...

The open access clause.

Yeah. The people were really against it, for very good reasoning, which is that it went against our current practices, which is that it was open access from the start.

Exactly.

Putting it into République Numérique bill, you couldn’t really put...like from the start it’s open access, so it restricted our freedoms by actually putting them in written law...

Or, at least, discouraged that your current practice. Yeah, I saw that argument online. There’s one part of it, mandated online discussion of regulations before parliamentary debate, that was in the Assembly version but was removed from the Senate version.

Which is that all the regulations must be proposed online for 60 days or so of discussion, between the proposing, and Assembly actually debating it so that Assembly can get input. I think it was, ultimately, removed, but we enacted it last year, so we got there first.

That was before I was the Digital Minister. I was just rallying for it. [laughs] Now, everything must be deliberated online for, most of the time, 60 days.

What do you mean everything?

All the regulations changed. That is to say, things that usually the parliament here, they look at the regulations change that each ministry proposes. They have a right to put it into a hearing or a parliamentary debate. I think it requires one-third or a majority -- I don’t know -- to convert a regulation into a process that’s like a law process.

They usually do it without a public debate period. What we are doing now is for all the regulation and all the laws that affects trade, intellectual property, or things like that...

But not all the laws?

Not all the laws, because the laws already went through this MP debate period. The thing is that for regulations, the MPs don’t have to, and indeed, they almost never debate it. Then the regulation takes effect after 7 days or 14 days. What we are now doing...

I’m not sure about the respective domains of what is law and regulation...

A law is something that requires parliamentary authorization. A regulation is something that has a law already, and then it authorizes the ministry to make some rules based on this law.

OK, I see. The French terms are different.

All right, sorry.

No, I didn’t know what it means...

It’s a variation on the implementation of a given law.

That’s right. For the implementation regulations, the ministries did not actually have to ask for the public debate previously. If they announced the regulation change, usually it used to take effect after 7 or 14 days.

Within those 7 or 14 days, of course the MPs can say, "But we need to have a wider discussion," but they almost never say that. This makes it very easy for the ministry to push through changes that may or may not be a good idea.

Now we’re lengthening it to 60 days and mandating that for anything that’s 60 days review period, it must be first posted on the online forum, and then the MPs can make an informed decision of whether to convert it into a due process based on the input on the online forum. I think it’s working very well so far.

How many people among the citizens...Does it have a far reach?

Mm-hmm.

Are people actually participating?

Yeah, sure. We had some numbers.

You are quite good at getting the people to do the thing by making it easy only taking a few seconds.

Exactly.

How many of those regulations give rise to public debate per week?

At the moment, we have 114 ongoing regulation debates.

About two per day.

About two per day, but over 60 days, more or less. There is 127 that’s finished, so yeah, around two per day. It’s a lot of comments really.

This looks like Chinese to me.

At least you can see there is 13,000 replies.

Yeah, I’m looking at a long tail.

It is a long tail. The first page is around a thousand or so. There is a plateau around hundreds or so.

Maximum, you’ve got like 10,000, 15,000 comments?

Yeah.

Is that comments or commenters?

Comments, but also commenters. The comment is actually very time-intensive. Most people just vote they like the idea or not, which is a higher number, I think. Taken together, I don’t know, hundreds of thousands.

You still have a potential problem with that, which is democracy where the ones who are heard are the ones who shout, the ones who are interested.

Sure, but they do not binding in any way.

No, they’re not binding, which is why it’s OK. If you just had a platform with binding regulations, where people could just down vote, the most active users would have more political power...

Yeah, which is why we don’t do that.

The power is still in the MPs, but the people should inform them. We were wondering, for POP, between letting the whole platform choose which is a possibility, but then it won’t be representative.

It’s representative of the members of the party, which is still OK in a way. The other way around is to take a random sample among that, which should give the same but also maybe half and half, with half going to the global population.

Which also has the added impact of maybe pushing some people to join the movement because they can express their voice and have an impact. I think if we could use similar stuff as pol.is and like that, we could get a nice public debate platform if people join it. But then the question is how do we get to the binding part, and how do we make it fair?

If you start with basic "Ask me anything" setting, where you bind by having the most consensus and the most controversial opinions a guaranteed response from the official, but it’s not binding in the sense that they will vote according to this, then this is a software that guarantees a response, authentic, long form, maybe a video live stream response.

Then you can delay the problem. You don’t have to solve the representative problem. Still, it is a way to sway a politician’s will because they will hear questions representing the things that they’ve never considered. I think it’s softer bridge than jumping to binding votes. That’s my personal experience.

Do you think that people would be ready? That is the main question that’s been bothering me for more than a year. It’s not that long, but that’s when I started working on those things. [laughs] I couldn’t have been working that longer. Do you think that people would be ready to accept the legitimacy or a random subset?

For juries, people have already accepted that.

Yes, more or less, although we are going back on that.

You are?

Every country, countries are slowly restricting the usage of juries. Australia got rid of it. So did South Africa. France I think restricted their usage to only certain types of cases.

Of course, the non-polarizing ones. It’s funny, because we don’t have a jury system. We are considering to implement it this year. [laughs]

No, we’re not going all the way of the juries doing the decision. There is a judge. The jury’s binding power to the judge is limited. How limited exactly is now being deliberated in the national deliberation. I think it’s at least a direction worth doing.

What I’m saying is that Taiwan’s very much into experimentation. If that form doesn’t work, at least we document how it fails for the rest of the world to see.

Why is it that you’re so ready to experiment and move forward? I’m jealous. I’m also jealous of your transit system.

[laughs] I think one of the reason is that we had press freedom only since the ’90s. The freedom of press was...The ban was lifted in ’89, and then the first crop of media was in the ’90s.

In the ’90s, it’s already the era of facts, the era of international real-time news. The World Wide Web came like four years afterwards. The new media, there’s no long tradition of traditional media. The media came when the digital revolution is happening.

It’s the same bunch of people doing political work, doing media work, and the same bunch of people who experiment with Internet. There is no traditional values of five generations of labor unions or something that tells the civil society or the media people how to behave.

You don’t have inertia.

We don’t have inertia, that’s right. We have more inertia than Estonia, of course. They were founded after Internet, but we don’t have that much inertia.

Whereas we’ve been there for a while.

Yeah, the oldest. [laughs]

No, no, no.

Where the revolution was invented?

No.

No? [laughs]

Where the bourgeois revolution was invented.

Oh, OK. [laughs]

I was quite impressed to see that Colombia is much older and has been stable. It’s the same republic as was founded, whereas France at some point during one hundred years once we were...a bit more than a hundred years we had 13 regimes, major regime changes.

We really can’t decide. [laughs] Although I don’t if you’ve seen, but quite a lot of people are calling for a new republic.

Yeah.

It would be our sixth.

Every time I went to Paris, there’s people who pitched this idea, the sixth republic.

Maybe we need a change.

Maybe. I wouldn’t interfere, but I’m keeping close attention.

Maybe that change can be more inclusive by using some of the technologies and some of the ideas you love in here.

These things, for me, it’s constitutional, meaning not necessarily requiring constitutional change. It changes constitution between people’s relationship and the government system. That’s for sure.

Everybody now, at least in Taiwan, understand that the representational system isn’t that representational anyway. We still need full-time lawmakers, because they are good at it. Then how to balance these two is the question.

One way I was considering was you have a public debate. You have lawmakers using a system like this. I don’t know how the Taiwan one works...

It’s pretty similar.

Public debate, lawmakers amend the law and put it in a legal form, but the ultimate choice is actually made by a group of random citizens. The random citizens are informed by the debate, by how the public chose on each amendment and such.

Then the lawmakers say, "Well, we changed the law. We didn’t implement everything because..." For example, that’s what happened for quite a few amendments to the République Numérique bill, is that they went against European treaties. They don’t have the political power to change this.

So lawmakers change it, and then eventually, the citizens choose. We can actually even go a bit further, which is if the citizens choose, but the margin is too small, then we want the public to be sure.

If you win by 51 percent, the following day you can lose by 1 percent. The population changes its opinion quite frequently. Not by much, but it oscillates. You want to have a strong mandate.

Today, for politicians to get elected, they can’t have a strong mandate because the whole political system is based around them winning by the closest margin possible.

By gerrymandering?

Not gerrymandering, which is not applicable to French president, because we don’t have an electoral college.

Right. You don’t have an electoral college. That’s great.

Yeah, we do have some more or less gerrymandering with local elections and stuff like that, but it’s really not on the same scale. We do have...OK, I’m losing the words for that. [laughs]

It’s good. I understand the general point.

We have this bill in the parliament, called Citizen Participation And Constitutional Reform Procedural Law.

Short.

Yeah, very short. The idea is it’s for civil society to participate in constitutional deliberation.

It would bring up all sides of an issue, with general constituents in which the members of this parliament are acting more as proxies, as you said. They need to be bound as lawmakers to translate this consensus into legal form.

They need to do a good enough job so that it will eventually pass a very high barrier of referendum, which will then enact a new constitution, essentially making a new republic. The procedure was pretty nicely designed by constitutional scholars and so on.

The thing with this whole system is that it’s not about technology. It’s how many of the MPs really want a new republic, which is why it’s been sitting in the parliament for quite a bit now.

We have a problem, we were discussing it with my boss yesterday, who told me he saw one of the major politicians, whether on TV talking about sortition and attacking it, treating the other with such disdain.

There is major reason for that is that sortition is basically the death of the political class, because once you give power to the people -- ultimate power. Maybe you have lawmakers, and maybe they are elected. That is not absurd. In the end, the real power lies with the people, and that means the death of the professional political class.

Yes, but what I’m saying in this constitutional reform procedural law is that it says that for each MP’s district constituents. It randomly draws samples, one man and one woman. The overall sampling need to reflect the overall demographics in age and in other criteria of the national population.

It actually mandates a sortition method that will say these 146 people will...

You say that it has to represent?

Yeah.

You mean that you check?

Sure, of course. I think it’s modeled after the British Columbia in Canada, where they had a very similar process before. The idea is that the MP still represents the district, but in order to form this kind of pattern of assembly, a random sample of a man and a woman also from the district acts as a proxy, as a representative of the district’s persons.

These people first have a consensus, and then the MPs of their districts of course can also participate, but in the sense that they are merely enacting the legal translation of their consensus.

That’s interesting.

I can send you the link to the draft. I think it’s pretty well-written.

It’s just the political mandate of passing it...

Is it in English? [laughs]

I don’t know, but I think machine translation is good enough that you’ll get the basic idea. It’s proposed to the parliament exactly one year ago. We haven’t seen a political window.

I had two more questions. Or three, maybe. [laughs] First thing is, here you had a bottom-up system. People called for such a platform where you had a movement that...

That demands this platform, really.

Yeah, that demanded. You mentioned in one of your interviews gov as a consumer. The government has a need because it has needs. It has to answer to the people. Part of your job is creating the tools for the government to answer to the call of the people.

That’s right.

Do you think that you could actually have a top-down initiative because we are creating POP , and there might be a will of the people, but there isn’t an organized thing as big as the movement you had here.

Do you think it can succeed? Listen to the people, but the masses aren’t really organized yet and propose a tool?

I think it only happens when the legitimacy of the whole governing system is in crisis.

Should I wish for a fringe candidate to be elected in France, then? [laughs]

I don’t know.

We have a few who are very nice. Thankfully, one of them didn’t get enough signatures, but he really wanted to send nuclear bombs...

We had one who within six months of being president would have started World War III.

In any case, I’ll tell you the idea of Mayor Ko Wen-je in the participatory budget when he became mayor late 2014. He is an independent. He belongs to no party, just like our Premier and the previous Premier.

Mayor of...?

Taipei City. Famously, he said at a point, that the whole city council is his position because we have a direct vote to the mayor, and then we have votes to the city council. This is unlike the Paris system, where the mayor is simply the majority in the city council or the district council.

Because he is independent, but most of his city council is partisan, it basically means that the entire city council doesn’t have to listen or have to agree with what the mayor says. By doing participatory budget, and by doing all sorts of direct connection between the city public servants and the public, the citizens, essentially.

He’s short-cutting the relationship between an independent mayor and the citizens, and it did cause backlash in the city council, but they couldn’t really overturn the entire participatory budget that year, and it now becomes a kind of balance between the parties and...

When you talk about the participatory budgets, what exactly do you mean by that? We have a participatory budget in Paris. Is that the same thing?

Right. It’s like this. Very, very, very similar, because we also had a citywide campaign and also every district sets its own participation methods and there’s a visualization. Conveniently, the website address is http://budget.taipei/.

How much of the budget is actually...What is the percentage of the city budgets?

Let’s see what they are doing.

We could start saying for the first year it’s going to be 12.5 million Euros.

OK. 12.5 million for the first year. I don’t know what Taipei City’s budget is.

Percentage-wise, it’s 0.3 percent. It’s not fair because a lot of this is maintenance budget, but the city doesn’t have a discretion anyway, so out of the investment budget I think it’s a higher amount. Let me get you to the numbers.

In any case, it’s accompanied by this diagram of how much went to the education, how much went to social welfare, and so on, and whether these things are being removed.

This is one project?

Yeah, this is one project, and not one PB project, one city project.

Yeah, yeah, I’m talking about the circles.

Yes.

Each circle is a project?

It’s a group of project, and then you can drill down, and then look basically the entire report that, say, they sent to the city council. For people who are interested in only the transport and so on, you can see the details of the entire city budget, basically.

The idea is that during the PB, the first three weeks or so, every single item had a discussion forum next to it so that people can ask questions, and after three weeks, other public servants of other cities areas went on and replied to every single one of them.

Basically, it’s a Q&A platform for the existing budget so people can be more informed when they are proposing ideas.

The people could propose ideas...

That are not already done, that are really innovative and they know how to fit it into the existing budget structure.

I am quite in favor of that kind of thing. I think the Paris budget is a bit higher in proportion.

Yes. It’s pretty high.

Yeah, and it’s a lot of money.

It is.

It might have been something like 20 percent of the investment budget (correction : investment is 20% of the total budget, and participatory budget is 5% of that, so around 100 million €). I’ll have to check it. Is something burning?

No. Maybe the 3D printer is not working. You said you had another question?

Yeah, just to add. What do you think of the potential impact of the debate platforms which are not public but which are private, via the effects of Facebook or whatever, should the state try to regulate?

No. There’s always the private sphere, where people talk about non-policy issues, so it’s silly to regulate people’s private sphere. Of course, people use Facebook now also as a public discussion forum, and it’s very hard to draw the line.

People get information...we saw people talking about fake news at Saturday’s meeting and major tech companies control the information in a way. Which can completely change a public debate.

Of course, which is why they are also now self-regulating because they don’t want to lose legitimacy altogether. They need to put some balancing, you know, fact-checking, next to the news and so on. We welcome these efforts, but we want to say that we regulate these efforts, it would be silly.

Of course, by fact checking they need actual factual information, in short, one URL per fact form and so on. This is something that the government can’t help preparing for things that we originate, but we’re not special in any way.

Any affected party, anyone who says they’re being defamed and so on, should be empowered to create this kind of instant clarifications. Ultimately it’s a matter of self-regulation, and the thing that we have, and that Facebook doesn’t, is that we have the potential to be a recursive public.

In a sense that our platform’s rule is our platform’s constitutional principles and everything, it could be affected by people’s input and our algorithms of visualization. Everything is there for contributors to change it, and I think in the long run it may be more effective than Facebook, but we don’t know fore sure, so it is worthwhile to try.

One thing is, I’m afraid of such technologies like Facebook or Google, and I think at some point there could the role of the states to be in this position. At some point when the private power becomes too big it becomes hard for the public to keep it in check.

While they’re doing that is just to create an open source and viable alternative, like what you’re doing, except your thing is just for public deliberation. If someone were to create a state-sponsored search engine akin to Google that might be an idea to eventually give the ultimate control to the stakeholders because at some point it becomes public good.

Google is a public goods because it affects society so much that it is grave if people can’t control what’s happening, and if it’s completely obscure. Problem is, the value of Google partially relies on that fact that it deals with our secrets, so I don’t know.

I don’t know. Personally I use DuckDuckGo.

In any case, I don’t think it’s just algorithm, though, it’s the user-contributed content that allows Google to see what’s important and what’s not important, the data also, not just the algorithm.

Given the examples in algorithm, none of the data that Google gains by is originated by itself, they rely especially on user input. It wouldn’t be possible to tune the algorithm as they did without user input.

I mostly meant to the problem with Google Search for the public is that people could change their website to...

To game it.

Yeah. It’s harder to game when it’s secret, the problem is it’s easier to...

I mean, they have incorporated so much machine learning component, such that even if they published the algorithm, but not the actual model that’s being trained, then it’s hard to game anyway.

It’s like alternate form of intelligence now, and we can’t very easily get a human understandable explanation of why it thinks this way.

We can get a printout but it would take 500 years to read through it. [laughs]

I think that’s our time.

That’s our time.

Thanks a lot.

It was very enjoyable.

Very interesting. Thanks for receiving me, and giving me the opportunity to spend some more time to talk with you.

Have a good trip back.