OK, welcome to the podcast.

Thank you.

I should say I might be running, for the Bangkok governor. I think we can learn a lot from Taiwan and Taipei. The thing I heard you doing some…

The lab.

The city lab.

Yes, the C-Lab and Social Innovation Lab.

Which will be a very interesting engagement.

The last time we met, we actually talk about this a little bit.

Yeah, yeah, yeah.

If you have time… you haven’t been to the place?

No, where is it?

It’s a great place, I’ve been there once.

If you have time, actually…

We have time, we have time.

OK, yeah.

…we can just go there, have a conversation, and I can show you around and have a tour.

It’s easy to describe this in the abstract fashion, like this is a sandbox, people can test self-driving vehicles. People can come and adapt them to the everyday needs of the people, but the most important thing is that all the innovations that’s featured there is open innovation. In a sense that it welcome other people to adapt and build based on it.

It has three layers.

The basement is an incubator, so we have 20 or so teams every year, and they either must be a start-up as in founded in the past five years, or they must be one of the sustainable goal producers, in achieving one of the sustainable goals. Either they can enter the incubator, and they can just show around their ideas like self-driving vehicles, or using machine learning to repair pipes, or to use telepresence to let people see the air quality…

Ah, the AirBox.

…and water quality. They recently have won, I think, the top prize for the civil IoT competition by doing this, but for water quality. People can see the flood, the public construction, the typhoon, and so on, and link what people’s ideas of what works as a water management to the government’s view, because they don’t always have the same view.

Using this kind of visualization we can get people around the world to contribute to both the climate science part, but also more importantly so that people can buy a small box like this one, and start measuring water quality by themselves.

I saw someone got back to Thailand, because the MP Klaikong visited Taiwan.

Yeah, MP Klaikong Vaidhyakarn.

Yeah, yeah.

…and learned of this, if he bought physical prototypes. That was the air part, but now we’re also working on the water part.

Can I buy it too? Where can I buy it?

Of course, you can buy it online.

Oh, OK.

We can also introduce you to the last person that won the competition. His name is Xiao Long, Zack has his contact. I think it is very important, especially on a city scale, that we let people see what people care, the priority, is determined jointly with the citizens.

For the Presidential Hackathon, we use a novel voting method, where everybody have 99 points, and they can vote for any of the project, but one vote is one point. If they want to vote four vote, that is 16 point, because it’s quadratic.

If they want everything to be voted into one, that’s 9 votes costing 81 points, but everybody has 99, so you cannot vote 10 votes, and it’s by design. Because then people with 18 left, they won’t want to squander those votes, so they will find something else that they learn, they have synergy.

We will get into a place where people usually vote for four or five project after learning about their synergy. When the top 20 gets announced, everybody feel they have won.

If you run a popular vote with direct voting like everybody votes just to the one project, most people will feel they have lost when you announce the budget, like participatory budget or something like that. If you use quadratic voting, then everybody feel they have won. It’s a different…

What have you learn?

Hm?

What they have to understand it? The water, after the trial and error, they do…

They have to see the brief report and the idea, the sketch of the idea. The top 20 means that they get the coaching help from the President’s office, from all the different ministries, to help realizing their ideas in three months. It’s a commitment of resources. It’s a little bit like participatory budget.

The top five performing teams, they get the President’s Trophy, and the trophy is a projector. If you turn it on, it shows the President giving the trophy to the team. The President basically said those five team, whatever they did in the last three months, I commit myself to make it into national policy within 12 months, so it’s a presidential promise.

As a mayor, or as a governor, I think this is a really good what we call a binding mechanism, because if you promise too much, your staff cannot deliver, but if you promise like five teams every year, that’s about the right amount.

That’s good.

Will it appeal towards the people with technology? If they have other group that can maybe more inclusive?

The important thing here, as with any participatory budget, is that they have to solve a real social problem. As you can see all the 20 correspond to one sustainable goal target, because we limit it to the 169 sustainable targets.

It’s not tilted toward technology. Technology is the 17th, and you see only one that is technology. Everything else talks about health, about agriculture, about sustainable cities, about life underwater, and talk about river, and this is the WaterBox by the way. Basically, because it’s quadratic voting, it doesn’t privilege people with a lot of followers on Facebook, because they cannot over-concentrate their vote.

It privileges people who build coalitions between sectors, between different political ideologies, because if you vote 7 and 7, that’s still below 99. But if you vote 9, you don’t have more than 4 to leave for the others. It rewards people who can form efficient coalitions.

When you said hundred, hundred what?

99 votes.

Oh, 99 vote.

When I vote this one, I can only vote 9 vote, which cost 81 points, because it’s quadratic.

But they count 9 here? They count how many points?

This is if you vote 9, it costs 81 points.

It cost you 81, but…

Yes, because it’s squared, right?

Oh, OK, but they have only 9 here.

Yeah, they count only 9.

Oh, only 9 so your value decreasing.

Yeah, so each marginal cost same as marginal return. It makes people want to vote exactly as they understand the landscape, not strategically.

How about the profile of the voters, are they more like a younger generation who votes, is it including elderly?

It’s everybody. In Taiwan we have 10 million people on the platform that this vote takes place. Taiwan has 23 million people, almost one-half of population. We found the most active is the 15-year-old, and 65.

Really?

I think it’s reasonable, because 15-year-old don’t have other way to participate, they cannot vote. Online, is their only way. People who are 65, they’re retired, and they care more about next generation, about sustainability. They will find more time to learn about this.

They have to load the app, and then they…

It’s just a website.

Oh, the website.

Yeah, but you can also do it through ATM. You can tell the cash machine to display choices, and you can just press plus and minus on it. People did that in Portugal. It’s easy, it’s a very intuitive design, but the net effect is that everybody feel they have won.

If it’s one vote or many votes to one, many people feel if their supported team doesn’t win, they lost. This way it makes people vote like five, seven different teams and have the balance between them. Any of which they won, they feel they have won. It’s a different design.

Who is winning the most vote? What topic is that?

These are the top 10, and this one use drone and machine vision to look at sea pollution. Instead of waiting for them to hit the shore and clean it up, it can predict where it will go, so we can trap that at the sea, and clean that before…

The plastic.

…yeah, the plastic, before they hit the shore.

This one is the WaterBox, of course. This is a very cheap device that people, because we have a new law this year that says if anyone pollutes the water on the agric land, with a industrial plant on agric land zone, they will get their water and electricity cut by the Minister of Economy directly.

They want to prove, of course, they didn’t pollute, it’s upstream, so everybody have the incentive to install such WaterBoxes.

It’s currently being, there’s a partnership with the New Zealand government, they just participated in the workshop. New Zealand accelerator helped this team to find its business model, and then they also are looking into piloting it in New Zealand as well, because New Zealand has a similar issue with dairy farms and water, and so on. This is what people really like.

You can trace back the source of the polluter?

Yes, it’s easier than air.

Ah, stream, yeah.

Yeah, stream is very simple.

One direction.

Yeah, stream is very predictable. Very predictable.

Even pollution for Bangkok come for other places, you know. Cambodia, burnings, yeah.

Exactly. This is interesting, because this is by the courts, the judicial branch, because it’s presidential. Not just the administration, the judicial also joined. So the courts, they proposed an AI that analyze drunk driving cases, and show exactly why such drunk driving is sentenced like this.

Because it solve a problem for the court. A lot of judge get public negative feedback saying that they didn’t fine or they didn’t put to prison the people who drive drunk, drive under influence, but actually there is a factor, a lot of factor like whether it’s the first time, whether they cause damage, and so on. It’s hard to explain.

They made an AI help assistance, so on their website, not only they translate the legal term to plain text, very easily, automatically, and anyone can get any drunk driving case and it analyze and show any citizen like this are the 15 factor, and why they only get three months, and that gets five months, and that get one year, and that’s because of those factors.

It might find a bad judge, too?

Yes.

That might be scary for the judge, maybe.

Yes, so next step for them is that whenever the judge is giving out sentence, they will show the recommendation. If the judge deviate from that range, the judge has to say why. That is for the courts.

Maybe we have an AI judge in the future.

It’s an AI assistant. Yeah. The AI is only showing the trend, and the judge can still say I’m going against the trend because of this factor.

That’s very helpful. I think many judge maybe have different opinion.

Yes, and also the drunk driving case for them is very troublesome, because it’s a large number of them, and it’s very mechanical. They don’t exercise a lot of creativity in passing out those judgments, so the more they can be assistively done, the more time they can concentrate on more cases that really require like investigation. Yeah.

It can be expand to other cases too?

Yes, yes. They did that actually for many cases, but the drunk driving is the first one to complete, so they chose it for Presidential Hackathon.

But they have a system that showed a lot of different cases. It’s called [Non-English speech] . Let’s see their website. Right. Here. They’re really good at SEO as well. Not only drunk driving, but also like stealing, fraud, and hurting people, and things like that.

It’s all very easy. You can just click on one, and choose which, like this is stealing, whether it’s for illegal means, whether it is for personal benefit, or just for survival. Are they mentally retarded? Where are they, and so on.

The AI can analyze any case to automatically fill in those factors and show that the damage should be this, and how many months should be done. It really saves a lot of time for the court people.

Yeah, that’s interesting.

Yeah, but…Not be dependent on the judge, the word, it could be…

The verdict, if the judge exercise their judiciary discretion, they need to say why, and that’s the only difference.

Oh, you have to have reasoning.

It’s not arbitrary. There’s many cases, but the Presidential Hackathon accelerates the development of those ideas.

How far this project go? It’s finished, complete?

Yeah. This one, the judge one, will roll out I think early next year. I think they’re done beta testing, so it would just roll out. Actually, they invited me to beta test the system, I think a couple weeks ago, right? I was in the Shihlin district mock court, and I was this judge.

Oh, really?

I was sitting right here.

Oh, but you’re online.

Yeah, using telepresence, and this is 360, so people can really feel how is it like to be a judge, to be in court. This is a mock trial with a lot of ninth-grade students, but the important thing here is that it showed everybody how a real judge would have been using this system for discussion. It really put everybody on the judge’s seat, so to speak. It’s very convincing. That is…

How you feel being a judge?

Ah, really nice.

Really nice.

No pressure?

No pressure.

Really?

There are three judges, as usual here, so I have a more senior judge to correct my mistakes if I made any, but it seems I did well.

He agree on my sentencing.

So you have AI assistance? Beside you.

Yeah, as a system, yes. Everybody can see it on the projector. That’s pretty mature. WaterBox, as I said, just roll out its first prototype visualization last week. They’re still waiting for the feedback.

This the group that’s using your space, that you mentioned? They are the one who are using your space in the City Lab?

Many of them, many of them, yes. The AirBox was using this lab for experiment, and they eventually became, one part of it became the WaterBox module.

For example, there was a team that use the same AI to go through what we call CSR, corporate sustainable report, so corporate or social responsibility – that was the old term – but nowadays they’re called sustainability report, so sustainability strategy.

They wrote the same AI, but go through all the CSR report and to find what are the trends. If you are a social entrepreneur trying to solve a problem with social enterprise, it can automatically pair you with the large company that has the CSR programs.

They even sell the service to the large companies, saying our AI can help you write your CSR report. You can just give me some numbers, and how you would like your shareholder to feel, and I can write a report for you. Then you edit it.

It’s a lot like the sentencing helper and the sentencing analyzer.

So can we commercialize this too?

Yes. It can be commercialized. It’s already commercialized. If we go to the incubator, I can show you the team.

Wow, interesting.

That’s the basic idea. That may be one of the most high-profile way that people can see the incubation ideas, because we tour around Taiwan, and we also invite the winners to share their project to people around Taiwan. People around Taiwan can say we have a use case here. Maybe they’re on a remote island, and they want to have a telemedicine with the large hospitals and the doctors.

They can say this is a real town hall, I just go there, but the social innovation lab has 12 ministries, section chief or higher, looking at me facilitating this work. The local people just meet where they already meet. I go to them, and then I connect back to Taipei, and the people here, section chief or higher, brainstorm on the problem that people here see.

For example, they say the local nurse is not authorized to give medical treatment when supervised by a remote doctor. It can only by a nearby doctor, but that’s a law change. Using the Presidential Hackathon, we just escalated that into the minister level, so this year the law has changed, so the local nurse can operate based on a remote doctor.

But that crossing many ministries?

Yes, 12 ministries.

You need to have staff from each ministry joining?

Yes, so my office is one delegate from each ministry.

Yeah, so you’re piercing through all the ministry.

That’s right.

That’s nice.

You do away with silos, you know.

It is horizontal, horizontal leadership.

Excellent.

That’s all I have.

Then how many city lab? You have one major one?

There’s a major one in Taipei, but there is a larger one in Linkou, that’s in New Taipei City.

In the suburb.

Yeah, in the suburb.

We went there today.

There’s one in Taoyuan as well, and actually two in Taichung, and one in Kaohsiung, and we also have points in Taitung, and Hualien. Like all around Taiwan.

Who operate these centers, your ministry?

Usually municipal government.

Ah, the local government.

They contract out to what we call an intermediary, someone who incubates social entrepreneurs. They’re a partner relationship, like Taichung contracts out to 覺心營 which is one of the intermediaries running start-ups with an impact.

Do they have maker space, something like that?

They do, they do.

Physical machines too?

Yes.

I think that’s important.

Also the green screen, also very important.

Ah, OK, for communication.

For telecommunication, because like in the main lab, the large lab, there’s a very popular program of indigenous language, because all around Taiwan, people want to learn about the indigenous natures, but they don’t have teachers in every precinct.

They just use a virtual learning tool. The green screen is so you can have the Jade Mountain, or have any river, or have something that’s immersed with the teacher, and the teacher can then teach about the indigenous culture.

You do remote teaching?

Yes, yes. Almost, I think, a hundred or so schools connect to the social innovation lab to learn about a language circle.

Include health, education.

Yes, and culture. That’s how it currently functions. Also, kitchen, very important. We always have good food.

Oh, really, so that’s a…?

Yeah, because if the 12 ministries hear about a new idea, they will resist, but if you give them really good food…

Yeah, I heard about that.

You have enough sugar in your brain, you make the idea efficient.

You become more open.

You should provide good food.

We have a really top chef in the social innovation lab.

I should remember that.

So that’s what we have.

I think we can start that in Bangkok, doing…I mean we find a open government, the open platform.

Yeah, and you have a law already, the social enterprise law. It makes for very easy bridging.

Then we’ll collaborate with the MPs, the ones who supported the act.

After our election I will go to Seoul, I think, to South Korea, and they already have a MOU of their social innovation lab, and our one. The Taichung one I think. I will go there to be a faculty.

Oh, really?

The program says that we share our mentors. We can go there and try to help them incubating.

So hopefully…

Maybe you can set something up like that?

We should. I should invite you next time, if I happen to be able to do it. I think that’s important. Now we have tried something, but maybe in a very small scale in Bangkok, city lab, they’re testing some simple things.

That’s great.

Where is the lab though? Is it far away from here?

It’s about 15 minutes from here, but it’s literally in the middle of Taipei. It’s the most central part of Taipei, near the central park, the Daan Park.

You go there in Tuesday, on Tuesday?

I go there every Wednesday, and also virtually every other Tuesday or so, and also sometime during Friday for the collaboration meeting, so almost one-third of my workdays.

From what time to what time?

From 10:00 AM to around 7:00 PM.

This one, the Taipei one, it’s also run by the local…?

No, no, no, it’s the national. It’s by the cabinet directly.

The national. Oh, OK.

It’s walking distance, like 10 minutes’ walk, from where all the ministers live. It’s very easy for us just to go there and have a chat.

And other minister visiting the center?

Yeah, the larger area is also the Minister of Culture, it is also putting a lot of effort on what they call the culture lab. It’s not just social innovation lab, it’s part of the larger culture lab, contemporary culture lab.

The Minister of Science and Technology, the Minister of Economy, have also visited the space. When they visited they didn’t just take a tour, the sit down and have a conversation like this one.

Yeah, spend time. For the town hall meeting that you mentioned, so this one, they do it locally and then you…?

I join them, I join them.

Remotely?

We used to ask them to come to the cabinet offices set-up in the various areas, but that makes people uncomfortable. We found they are much more comfortable if they just go to where they normally have town halls, and we join them with the live streaming equipment.

For a town hall meeting, how can you prevent it’s being dominated by a certain group of people? Meaning some group, they have voice, NGO, sometimes they control the meeting.

The planning starts well before-hand. We make sure that first it’s radically transparent. Everybody in Taipei, Kaohsiung, Taichung, and so on, are also participants. It’s very difficult to take over all the six different places. That’s the first thing.

The second thing is that we made sure that anything they propose, they have to correspond to one of the sustainable goals, so they must be talking about public benefit, not just private benefit.

In this kind of setting, because people generally understand no matter which faction they are on, they’re in a minority. This is a real collaborative setting. They will not be very loud, because first of all, it doesn’t matter, Taipei can just dial the volume down.

You have control.

That’s right. Most of the ministries are in Taipei anyway, so it doesn’t matter.

Not confrontation.

Right. You cannot punch people over the projector. It’s also much safer for the public service, because they feel if they innovate, they get a credit. People can immediately see who proposed the solution. If they upset the local people, or local people become upset, only I am at risk, nobody else is there. [laughs]

I think it’s also good to have the physical meeting too, combining with technology. You have face to face, so I think that makes this good.

Yes.

I think maybe even better than everyone online.

Very much so, very much so.

I think at the town hall…

You cannot just comment online, you have to go to one of our six meeting places to join.

That’s important. That’s a key thing.

Yes, we have a very detailed rundown and so on. If your team want to learn about it, we’re ready to publish and share with you.

Maybe a good thing, yeah. I think I would like a visit sometime.

Maybe let’s just go there.

Yeah.

Like now.

OK, cool. Wow.

Let’s go.