 Hello, I'm Audrey Tang, Taoist Digital Minister, and really happy to be here virtually to share with you some thoughts around digital social innovation, that is to say public participation for public benefit through digital means. When I think of social innovation, I often think of Dr. Tsai Ing-wen, our president, in her inauguration speech. She said, Before we think of democracy as a clash between two opposing values, but now democracy must become a conversation between many diverse values. Indeed, whenever there is emergent technology, there will be people thinking more about economic growth, versus people who think more about environmental sustainability. There will be people who think about the automation and the time it could save, versus people who think about inequality and social justice protections. Now in my office, that is the social innovation lab in Taiwan, I actually meet with people with different positions and find collective values that we can agree on. That is to say, anyone can just talk to me every Wednesday from 10am to 10pm, and I tour around Taiwan every other Tuesday or so to talk with the local people about things that they care and about the positions that they hold. But everything that we talk about is radically transparent, that is to say, it's published online. And so people with this kind of asynchronous conversation can just gather around an emerging technology such as self-driving vehicles, and collectively find the parameters. Now for example, here are the actual self-driving vehicles that's roaming around in my office. These are self-driving tricycles, because they are slow, and because they are open source, open hardware, and open data, anyone can just tinker around to fit these tricycles more into the social norms of how people expect to interact with these technologies. The greatest thing about this kind of effective partnership is that we don't have to jump into loss whenever there's technologies that we don't quite understand or we don't quite have a first-hand experience. Rather, we can adopt a sandbox approach of having one year of the civil society and the private sector forming the norms around such technologies, and then based on those norms introduce policies, and based on the policies, we write code, we write them into the parameters in artificial intelligence, and it's only after the norms, the policy and the code that we turn them into law. One of the prominent places where we employ machine learning in social innovation is called presidential hackathon. Every year, Dr. Tsai Ing-wen, our president, blesses five teams out of a hundred or so teams from the entire society. Each team that made into the top 20, we make them into tri-lingual teams that combines domain expertise, technical knowledge, as well as the public service people. That is to say, the top five teams every year take three months to deliver a prototype. For example, last year, one of the five winners is the water savior because they save water using, and they use machine learning to ensure that in Taiwan, the Taiwan Water Corporation don't have to employ the people but spend most of their time listening to the water pipes that are not leaking. Rather, these excellent skilled masters can just wake up and look into a chatbot who informs them about the most potential leaking places. In this way, we increase water use efficiency so that the leaking point, they used to take two months to discover, now can actually be discovered in a couple days. And they actually went to Wellington to New Zealand for three more months after the presidential hackathon in order to help the New Zealand people to adapt to a water management efficiency and resources. The best social innovation have this shape, what we call data collaboratives. That is to say, the open data from the government is not the only source of open data. Rather, the social sector can also set up their own independent, for example, water pollution management or air pollution measurement or other devices through innovations such as distributed ledgers, other known as blockchain, but distributed ledgers. Everybody can make sure that people cannot change each other's numbers and we can share with everyone, including junior high school students, on the National Supercomputing Centre desktop 20 in the world and make sure that everyone in K-12 actually have access to in-place computing to test their models of air pollution, of water pollution, of all the sorts of different environmental data. And everyone can participate, even primary school teachers actually use those airboxes to educate the data stewardship for children. If you have maintained a data measurement device, you have to care about all the different obligations as data operators. That is something that is very hard to teach, but it's very easy to learn once you are a data operator. And through this way, we make sure that data literacy and numeracy is designed into our curriculum starting this August that emphasizes creative and critical thinking skills when it comes to information and communication technology. And so through this way, we enhance the reliable data and have everybody participate in the data collaborative works. Now the five winning teams from the Presidential Hackathon, they each get a trophy from the President. There's no prize money because many of them, actually all of them has public servants. However, what a trophy is, is actually a projector. If you turn it on, it projects the President herself giving you that trophy. So it's very interesting because the Director General, if they don't like your idea, you can just turn on the projector and summon the President and the Director General will make sure that all the budget, all the regulations, and all the personnel that you need will be fulfilled because what a trophy represents is a presidential promise. The top five teams after three months of prototyping gets the award of having adopted within a year into the national policy, and we make sure that we maintain your project basically as part of public service. Now in this year, the top 10 teams were still on the finals round, but the top 10 teams has four teams that are very much like water saviour in that they combine artificial intelligence with collective intelligence. For example, the one that also works on water pollution management is basically a water box that can very easily be placed into any river and so on and continuously uploads its measurements into just like air box, a distributed ledger, or a shared data center so that people can collectively find out what actually is the root cause of any water pollution. Or for example, the promoting the rule of law and ensure equal access team in front of the judicial branch use machine learning to make sure that everybody can understand the prosecutions. And for instances that are very mechanical like drug driving, they actually use machine learning to explain why each evidence and discovery will end up in what sentences and thereby save a lot of time for the people in the judicial branch in the court from those mundane processing of the very mechanical cases. And there are also two more teams. One uses machine learning to figure out illicit financial flows. And there's another team that use machine learning in particular computer vision to look at the sea and find out the plastic waste, where they come from, where they're flowing to, and how to categorize them so that we can manage them right on the sea instead of just in the beach. And so all these are very creative uses of machine learning in collaboration, not against collective intelligence. And we collect those innovations, just index them. There's more than 400 teams now and they establish like companies, cooperatives, associations, foundations, and so on. So that you can discover each other using the common index of sustainable development goals, the 17 colors in each locality, in each county. And so that people can form natural equal systems that partners with, for example, a large enterprises sustainability strategy. And so as I mentioned, I personally go around tours in Taiwan to discover such teams. And every time I tour, I actually use telepresence so that the 12 different ministries and five different municipalities and cities are jointly looking at a local situation and we can figure out creative ways to solve them in a way that are responsive and inclusive. Instead of asking people to come to our website, we actually go to people with technology and make sure that their voice is properly amplified. But once their voices are stated, how do we actually find the consensus? We use an open source system called POTUS to do exactly that. It is a AI-powered conversation and all the avatars you see here represent people feeling differently around the same fact. So we ask again people to contribute open data, the facts from all the different sectors. And then we ask people's feelings around this. In fact, you may feel anxious and she may feel happy and it's all okay. And we allocate explicitly three weeks or more for the feelings to resonate with each other. And then we brainstorm on face-to-face meetings to find ideas. The best ideas are the one that take care of the most people's feelings. Now, after those ideas are ratified by the government, we make sure that all the controversial issues are still tabled for further discussion. But we can legalize and ratify the rough consensus. And so the interface is very simple. Anyone can press agree or disagree. But there is no reply button and so there is no way for Troll to grow. If there is a reply button, of course, people with the most time wins. But however, with no reply button, we almost always end up with a shape like this. You can see there's only five divisive statements. And if you look at popular media or indeed some social media, you'll think that's all there it is. But actually, most people agree on most of their neighbors with most of things, most of their time. And so we can just call them a rough consensus and use them to do the performance indicators to measure the thing that should be measured as dictated by the rough consensus. But however, those divisive statements then become further topics that the entire society can think together on. And through this way, we deliver innovations from those common values that are to the benefit of everybody. Not just one sector or the other sector, but all the different sectors throughout the common value that is the sustainable development goals. As the digital minister, my own index is out of 1718 to build reliable data. 1717 to build effective partnership. And 1716, as I mentioned with New Zealand, that we offer such co-creative innovations. In the open, we actually make sure that we always co-develop the solutions with our like-minded countries and friends. Finally, I would like to read you a prayer or a poem that is my job description. Two and a half years ago, when I become the digital minister, the HR people asked me to write something about this position because I'm the first digital minister. Now I'll share it with you. When we see the internet of things, let's make it an internet of beings. When we see virtual reality, let's make it a shared reality. When we see machine learning, let's make it collaborative learning. When we see user experience, let's make it about human experience. And whenever we hear that a singularity is near, let us always remember the plurality is here.