They invited this team after three months of our Presidential Hackathon, into their own GovTech Hackathon for another three months to co-create solutions with them.

Right, exactly. We index each of them with a specific SDG target. When, for example, New Zealand learns of this, which is a new situation for them as well...They didn’t have water shortage. Now, because of climate change, some places are starting to.

They’re trilingual, each team. Even if they didn’t start out as trilingual, we make them trilingual. Then, like last year, one of the five winners eventually used machine learning to automate the detection of possible leaking points in the supply lines, the pipes, of water so that the people who ...

This one address the plastic waste at sea, and so on, and so forth. Basically, all of them, we coach them, the 20 teams, into trilingual teams. That is to say they have one person specializing in public policy, they have one person specializing in the domain, and the one ...

It’s done by AI, and they also use crowdsourcing to provide a dictionary to translate the prosecutor’s documents into laypeople’s language, because we’re now about to introduce a jury system. It’s very important that every knows what happens in the court process.

It’s also an interesting one. This used machine learning to automate the sentencing for drunk driving, because no judge or prosecutor feel a lot of personal accomplishment doing those very mechanical chore-like things.

There’s more than 2,000 nodes that upload to a distributed ledger so that people can be sure that nobody can mutate each other’s numbers. It’s a public accountability system. This is, I think, very promising. This one used machine learning to detect false debts and illicit financial flows well before ...

They will actually sponsor those WaterBoxes. Then the farmers also will sponsor those WaterBoxes, because they want to know, who is the pollutant in the upstream area, multiple upstreams? These are cheap IoT devices that will just upload all those numbers, just as AirBox before that for air pollution measurement.

Yes. They are forced to move off after 20 years, anyway, but within 20 years, if they ever become pollutants, they will get shut down. It’s interesting, because then, with this law, which wasn’t there, it’s now in the plant’s interest, of the business owner’s interest, to prove that they ...

Where one side may be a plant, one side may be plants, so industrial and agricultural use. Because we have a new law that just passed that says if any industry nearby the farmlands pollutes the water, then the central administration has the authority to disconnect their power and water ...

I talked a little bit last night. Basically, the Presidential Hackathon is a process in which that 20 teams are selected every year -- this year, by QV -- out of 100 or so teams. They’re all very interesting social innovations, like this one uses the so-called WaterBox to measure ...

They must only talk about those 10 things in that meeting. Then they provide accountability by saying, "This year, we are just going to send our AIT personnel to the Presidential Hackathon. Check mark."

Maybe convincing to admit in Taiwan the US should participate in the Presidential Hackathon, and so on. All these statements are then, after the two-month period, read aloud by, moderated by yours truly, but participated by the AIT people, our Ministry of Foreign Affairs, such as the Ministry of Health ...

This is about the regional stability. This is about the importance of rule of law, freedom, religious tolerance, how to share our story, and how we actually are even better than the US when it comes to human rights and addressing equality, and how it can specialize agencies.

Even if you get 500 friends to vote exactly the same as you do, it makes no difference, because it can still get delisted from the list. You generally have to convince people on the other side as well. The first one is about Indo-Pacific alliances, the second one is ...

Actually, the picture looks like that. Again, only five divisive statements and a lot more consensus statements. This is called group-informed consensus, or GIC, meaning that it have to convince supermajority in both groups.

Group A really doesn’t like the idea. Many people feel unsure, actually, but still, there is more people opposing than for this statement. If this is mainstream media, they all will just focus on this and amplify the message for divisiveness.

This is the automated clustering. This is, based on your vote, your avatar will move among clusters. You can see cluster B, most of them think -- 85 percent of them think -- that every time PRC close the international door for Taiwan, the US should try to open one ...

Everybody agree with that, it’s just how. Many people voted, and it looks like this. You can very easily see what are the wedges, because this is principal component analysis. People are naturally grouped like Netflix clusters. Their feeling toward the PRC is obviously the main thing that drives the ...

Sure. This is a Pol.is system. It’s very simple. We are using it to talk about diplomacy, of all things. The de facto American embassy in Taiwan have collaborated on this particular we got. For example, our previous question, which runs for two months, was how to promote Taiwan’s role ...

Rather than voting for people. Easily half, or more than half, of people feel that they have lost something. Having this kind of AI-based agenda setting, even before we put things on the QV ballot, I think, is crucial.

Even if the QV voice crisis goes into some other items, they will still have participated in the shaping of those items. I think it’s very important after a vote, everybody feel that they have won something.

Then we can just take the top consensus, package them into a framed statement that can be a QV ballot item. I think, after this kind of early-stage consensus-making -- or really, just consent-making -- people won’t feel that they lose if they don’t get the QV result that they ...

Mainstream media, indeed, social media tend to over-amplify the divisiveness. Actually, most people agree on most of things most of the time with most of their neighbors. What this system shows is that, around one agenda, there is actually far more consensus among community members.

Once people agree or disagree sufficiently, they will see that first, their social media friends are all over the place. People who hold different opinions are not necessarily enemies. Second, that there’s far more consensus that anyone imagined.

Basically, around each agenda, like, how should we integrate autonomous driving into our society? Each person can state a simple statement of a binary sentiment that other people can agree or disagree, up or downvote, but there’s no reply button, so there’s no trolls.

We actually use machine learning to, on each agenda, ask the community to refine it, so that people can agree on common visions. This is what we usually use. It’s a system called Pol.is. It’s called P-O-L-dot-I-S.

If we try voting for platforms, specifically platform items like to-dos, instead of people, then I can see it working much better. Like, to be a chapter leader, to be some member, or something gives you a voice to raise the agenda.

You can’t really set up an arbitrary, rational Habermas discourse veil of ignorance to get that effect away. We know that there are people with charisma, like Glenn, [laughs] and that it would be very hard to decouple the charisma with the platform.

Also, I would like to also introduce different mechanisms for agenda setting, because too often, when electing one person, some people elect for their platform, or some people elect for their charisma.

I don’t know. I don’t know. That was the question I posed in our previous call. If he thinks a token, or fake money, as he calls it, QV that is renewed every election cycle, that, I think, could work better. It doesn’t privilege the old guards. I think we’re ...

I really don’t think that accumulated QV is the best way to go.

That’s actually the research question that I posed to Glenn, in that if you allow accumulation of votes, but your base is also still expanding, doesn’t that create what Vitalik referred to as, "Privileging this generation at the cost of the next"?

In any case, I’m familiar with that. The main thing is just, Debian isn’t expanding as fast as to require a new governance mechanism, but RxC is still expanding really quickly. The stability of the constituent, if you will, is the main concern.

Sure. I know that, for example, most of the governance principles, I learned from the Debian Constitution, which is an open source way of having fearless leaders democratically elected. There was no QV back then, so we have to conversant voting, but it was pretty good as well.

How would that work? If you organize a local chapter, your local chapter counts as one electorate or something? [laughs]

It’s like, people go to conferences to vote? That could sell tickets. [laughs]

We have 20 minutes.

Code is just implementation of a new norm, as inspired by art. We don’t see much of that in traditional institutions. I think those two main connections is where RxC can really offer partnership and value.

Also, the now environment also aware, art part, is also very good, because usually, these advocacy institutions see art more as an instrument. What RxC is that art need to lead the aspiration or consciousness change, rather than the other way around.

I think RadicalxChange has a much resonance with the entrepreneurship scene. That is something that those traditional societies often lacks. Even Media Lab isn’t directly entrepreneurial. I think that’s the main connection.

Very much so, very much so.

Yeah, Joi Ito, too. He’s actually one of the co-founders of this movement. I was contacted by the IEEE, John Havens, but there is also the architect of GDPR here and so on. I’m happy to just make connections.

It’s not anti-GDP, but it’s post-GDP. I think all the folks here are part of this vision. I think we are, like Sandy’s here. I’m happy to make introductions if you see any of...

Last evening, also, there was a talk about that. Localism mean different things to different areas. We can actually more holistically chain those different areas together in a way that is more system-syncing. GDP totally lose that.

If they think about next quarter, they don’t. GDP, for all its uses, is actually not very good for long-scale thinking. Based on SDGs, CXI is also doing, like, 10 indicators that are much more holistic. You can combine multiple levels of abstraction, instead of just having localism, I think.

I think Sandy is heavily involved in that section. I am more involved in the post-GDP design, like expanding the planning horizon to be multigenerational. The various sectors, if they can think in a multigenerational way, their interests actually align.

Within it, there is also what the CXI people call data agency, which I think is more or less same as data dignity, maybe 80 percent overlapping. The data agency part, that combines digital identity with data policy and things like that, is also part of the CXI.

We’re all part of this MIT/IEEE thing called Council on Extended Intelligence, or globalcxi.org. We just published our first call for comments. Extended intelligence, which is assistive intelligence plus collective intelligence, is the main banner.

OK.