— then you don’t have to get involved in order to make sure that you are dominating — or getting a big weight in the conversation, which is quite different.

Yeah. I was thinking with the feeling of obligation to participate, people might feel that, “I really need my views to be counted. I really need the views of people and people like me to be included.” I guess, because you’re not doing the head-counting thing because it’s not a ...

Is there any downside to these kinds of direct democratic systems? Maybe ordinary people feel like they have to be participating in this public platform or other platforms, when they would rather be taking care of their children or doing their job or playing sports or pursuing other hobbies? Is ...

Yeah. We’ll stick up a link to that. And I might get some more links from you, because I think this is a super interesting system, and I’d love someone in the audience to try to carry it forward and try to apply it outside of Taiwan.

Wow. Yeah. There’s a lot more that I’d love to have said about this. I’m guessing this is in Mandarin, right?

Yeah. Interesting. It sounds like the thing that you are really incentivized to do is to make a statement that people in other clusters will agree with. So people who usually disagree with you, you’ve got to say something that you believe that they also believe. This incentivizes you to ...

Maybe this isn’t the rabbit hole that we should go down, but I keep envisaging ways that this could be gamed in a sense. Or at least that it could be non-representative, and people might feel, “Most people agree with me, but our views aren’t getting enough weight,” and feeling ...

To put that in layman’s terms, if you have 100 different statements that people can agree or disagree with, then it’ll be rare for two people to agree on every single one of them. So people end up far apart in space in that sense. But then if you’re like, ...

Wow. OK. So if me and someone else, we both basically have exactly the same opinions, the same feelings towards all of the statements that people put forward — then we’re effectively counted as one person. That then raises the alternative concern that 99% of people might all feel the ...

I see.

So do you do some weighted voting?

A concern that I would have with Polis, or potentially all platforms of this kind, is that the people who participate in them could well be non-representative of the population of Taiwan as a whole. In particular, you imagine people who are working full-time, people who have children, for example ...

Yeah, that makes sense. I guess it is the case that we get stuck and talk the most about the stuff that is the most intractable, because the other things have been solved and we’ve moved on.

It reminds me of this case I heard: if you ask people are they in general pro–gun rights or anti–gun rights, then they tend to split pretty 50/50. But the more specific you are, like, “Should someone be able to buy this gun under this condition and carry it in ...

It seems like a really key phenomenon here is that the more specific the question, the more agreement you have. And it’s super interesting that that’s the case. The US, in this exact moment in time, seems particularly bad for people talking at a very abstract level about values — ...

What’s the most difficult case that you’ve dealt with on the platform?

OK, yeah. That’s super interesting.

The key difference there is that because you are not attacking or directly responding to a specific individual, it doesn’t become personal. You don’t get as strong negative feelings coming up.

I see. Do you think that in general, this is a useful way that people could have better conversations? Not just on this platform, but in general: rather than responding, instead you just state more clearly your view and then try to get people to agree with that?

So if someone’s saying something that I disagree with, how can I contribute my rebuttal, or my alternative opinion?

Why do you think that sort of closure to response is really valuable in that case, even though in general, we want people to be able to provide their opinions? That’s the key goal.

Right, yeah. I was going to say, one part of your vibe is that people should be able to share their opinions, they should be able to make the tools that they want, society should be very open, we want people to contribute their ideas. But then I’ve heard on ...

I see. It sounds like, to some extent, this is a tool for having extremely productive and clear conversations across groups of people who don’t agree. It’s through that very good communication, and being very clear about what you are saying and aren’t saying, and not amplifying the views of ...

Has this worked consistently when it’s been applied, or are there some times when people just disagree too much and you don’t reach a reasonable consensus?

Interesting.

OK, so you’ve got the clusters, people understand different views, and there’s some things that people agree on. Then you’re saying you generally see a process whereby people converge — they begin to agree more and more, because you’ve started to highlight the goals and the feelings that they share. ...

I see. OK, I’m just trying to picture this in my head. So you’ve got the different clusters and you’ve got a bunch of goals or feelings that most people share. But presumably there still isn’t a consensus on what the policy should be, because that’s why there’s a disagreement ...

So across the different clusters, you’ll be like, “What are the things that people do agree on, even if they disagree on other stuff?” And then you try to build from there to find a policy that would match with the things that have broad support?

OK, that makes sense for systematizing the disagreements and understanding them. And then what do you do? Because you talk a lot about building consensus. What do you do in order to try to get —

Ah, they post feelings, OK. They post statements that then people can express agreement or disagreement with. And you use that to basically see that there’ll be different clusters of opinion, where there’ll be a whole lot of people who basically say yes, no, yes, yes, yes — they all ...

Right, OK. So there’s a lot there. Let’s unpack that a little bit gradually. I want to come back to the quadratic voting, but let’s talk about this method first. I guess you’re saying you’ve got some difficult policy issue where there isn’t an existing consensus, say, “Should Uber be ...

And that’s the design of the place. And then you’re suggesting that instead of trying to improve discussion of politics on Facebook, you can design new spaces — create new places that are conducive to people getting along, people having constructive conversations, people reaching consensus. Potentially just designing your own ...

It’s a mode, but not a person. Yeah.

Yeah. Interesting. So there’s this issue of the public trust in the government, but also the government needing to trust individual citizens to be responsible with all of this information and to make good use of it. To what extent do you think the Taiwanese public is unusually trustworthy in ...

If it can be better, then you have the tools.

Yeah, it’s interesting. I suppose it’s cooperative, but it also does pose a challenge to the general public, to not merely complain about things being bad.

OK. Sounds like you agree.

But not merely outrage — also for people to come together and explain how things could be better and help to improve them, and thereby preserve, to some degree, people’s confidence and trust in government. This potentially has implications beyond just the government providing good services. It has implications for ...

And it seems like, to some extent, you’re providing an outlet for people to —

I think the main theory I have for why that is, is that the internet has made it a lot easier to see when government is messing up. In the past, it was easier to cover up or just not be aware and not be salient when people are being ...

We’ll come back and talk about the nuts-and-bolts implementation of a bunch of these different projects that you’ve been working on later in the interview. But first off, it seems like in general there’s something of a crisis of confidence or trust in government. These issues have always been with ...

Yeah. Do you know what format the podcast is going to take?

But also, if I’ve been reading the history right in preparing for this interview, it seems like the 2014 protests led to a real sea change in how the Taiwanese government interacts with the general public, how it collaborates with volunteers and citizens, and gets information from people. It’s stuff ...

It seems like a lot of people have asked you to do interviews over the last few years. I guess because Taiwan seems to have been a real pioneer in a whole lot of these areas that you’ve been working on — it’s been well recognized for doing a really ...

I think podcasts like this are very important as well. [laughs] But what about over the last month or year? What kinds of active things have you had on the boil?

I hope to talk about technologies that you’re excited by and get a sense of your overall political philosophy. But first, what are you working on at the moment, and why do you think it’s important?

Thanks so much for coming on the podcast, Audrey.