 Okay, yeah. Just quick introduction. My name is Daisuke. I'm an AI entrepreneur in Tokyo. I'm running the Hakkazon community for the Python machine learning with 8,000 members. This time we are working on a book about how AI can beat COVID-19. Three members from University of Tokyo. They are medical entrepreneurs, researchers, so we will be publishing the book in February, and then this interview will be the special contents in this book. Thank you very much. Okay, sounds good to me. Yeah, okay. And then I think I'm allowed to record this interview according to Zack-san. Yes, of course. We will have the shared recording on the Skype side, and just for your benefit, once you start a local recording, I will also do a local recording, so it's higher quality for you afterwards, and I'll send it to you also on Skype. Okay, thank you very much. I'm going to start recording. Okay, thank you very much. Okay. This is now recorded. Let me just double check. Yes, this is recorded. Okay, I plan to do this interview for 40-45 minutes. Is it good for you? Yes, of course, and usually we will publish the video if you're okay with it to our YouTube channel, but if you think a transcript is preferred, we can also do a transcript instead of the video to publish. Wow, fantastic. Thank you very much. Okay, let's start. So our book title is How Can AI Beat Coronavirus? And then thank you very much for your time. So this time, as a preparation, I talked with as many Taiwanese as possible, and then these voices from them reflect a strong point of your strategy. An example of my Taiwanese friends, background engineer, female entrepreneur, ex-Google manufacturer, Gov Zero members. Okay, so I prepared nine questions. First question is, as of October 2020, many Japanese companies are suffering from bad economy after pandemic. My strategy is preparing as a pessimist and act as an optimist. So what is your advice or encouragement for Japanese business people and the citizens? What mentality should we have like a smart person like you? Okay, there is a saying, I often quote from Leona Cohen, and I quote, there is a crack in everything, and that's how the light gets in. The idea is that on a common issue like COVID, it both of course separates people because of social distancing and so on, but it also connects people like never before. Like previously before the coronavirus pandemic, we will probably not be having this conversation. And this book probably will not be published because that idea around AI before the coronavirus is definitely not specializing in the humanitarian part of AI. People think about transhumanism and people spend a lot of time on vanity projects. I say this without judgment, but by making people look like better than other people and so on. And nowadays these aesthetics are no longer the norm. We don't see people showing off Instagram or on social media anymore because we understand that there are much more pressing issues to be solved and we need to direct our energy to it. And so I think the main idea is here about the common good. If we have the common good in mind, then it makes it easier for us to find the common projects that we can all contribute on. But if we think more individualistically, like raising my status above other people and so on, that will be detrimental in the time of coronavirus. Okay. Well, this is a wonderful insight. Do you mean more, I guess, a company with social mission has an opportunity to go to business and then this non-social company should shift to the social direction in this pandemic? Well, in the pandemic, for example, we see that many companies are forced to adopt teleworking. But teleworking is not just a neutral technology. It also connects people more intimately. It allows people to spend more time with the people they care in their locality and also makes for a more diverse and inclusive workforce. If of course you have broadband as human rights, otherwise it excludes people also. So any new COVID strategies has this light side of people who previously were unable to accept a more socially flexible arrangement like telework are often forced to accept it. But even after the pandemic goes away, like in Taiwan, it's been five months with no local cases. Still, we use telework whenever appropriate because people have already have a good experience with teleworking as compared to maybe 10 or 15 years ago when teleworking was not very practical and the connection quality doesn't work at all. But nowadays, you can see and hear me just fine. I can hear you just fine. And this makes teleworking a modern reality. So it doesn't have to be social purpose, but we do have to evaluate the social or more pro-social consequences of the ways that we have changed in working together. And organizational changes are also a kind of technology, right? Teleworking is a social innovation. Okay, great. I check to several your keynote speech and then several times you are mentioning about being robust is important in the society and business. So in COVID-19, all the businesses need to be more robust. For example, some of the restaurant owners open in the kitchen in the 10th floor of the building and then using the sales channel only over each. Yes. This kind of small expense business is a more robust than big expense business. So what do you think about this kind of being more robust? Yeah, definitely. If you anticipated this need and already worked this way, of course, you're robust. But even if you do not anticipate and take a hit when the new coronavirus came, then if you can very quickly innovate and pivot to a way that works with the new situation, then you're maybe not as robust as others, but you're resilient. And I think resilience is a shared value too, especially around like island countries such as Japan and Taiwan, because we understand there are some like disasters like earthquake or typhoon that simply cannot be managed. Management doesn't work as a term against such natural disaster, but we can be resilient in the sense of not only minimizing the damage, but also goes around the post tsunami, the post earthquake situations as a whole society. So while robust is more about a single individual or organization, I think resilience is part of the culture, it's the entire society. These two work very closely together. Yes. Yeah, this is very valuable insight and advice. Thank you very much. My second question about myself, I and my wife got COVID-19 and were in the hospital for three weeks in April. That was a huge setback for us. But now, strangely, we feel somehow sank through for COVID, because I'm a self startup founder. COVID made me much stronger, bolder and smarter person. So for example, in biology, virus is somehow necessary for creatures, evolutions. So in my opinion, how can we leverage COVID as an opportunity to become person or stronger person or better society or stronger society? Yeah. First of all, congratulations on your recovery. And also, I think COVID shows that it is actually possible to live in a way that is sustainable in many places where there's a lot of air pollution, a lot of water pollution and so on. Because of COVID, people have seen a clear sky for the first time. People have seen the clean water for the first time, because those structural issues around the environment were simply neglected. And especially for younger people, maybe for their entire life, they have seen the water and the air and the environment as a kind of damaged vision. But with the COVID comes the opportunity to see the natural habitat again as lively as it is. And you can't unsee it, you can't go back. Once people understand that it is actually possible through teleworking and through actually Uber Eats, as you mentioned, it is possible to reduce our overall footprint on the environment, live much more sustainably. And I think that is also a much needed way for the society to feel better even after COVID and recover in a resilient and sustainable fashion. Yes, thank you very much. Next question is from my Taiwanese female startup founder. She said you're representing Taiwanese multiple freedom as an icon, which is great. Japan is a single ethnic country over thousands and thousands of years. This kind of pain, attention to the detailed culture created Toyota's very efficient car manufacturing system as a system, which was great. But sometimes this lack of diversity brought Japan from growth or innovations. So under pandemic and global business competition, diversity is very, very important. So how can Japan implement the essence of diversity into our single ethnic country? What is the right balance between a single way of thinking or diversity? And then how can we learn from Taiwan regarding diversity? Yeah, I think diversity is not just about ethnic though. In Taiwan, of course, our ethnic composition is fairly uniform, but our cultural composition is very diverse, right? Even for people who identify as ethnic Han, there are the Taiwanese Holok language, Taiwanese Hakka language with at least six different sub languages, and also of course Mandarin, but also spoken in different ways. So that's at least three different languages. And out to that, the National Languages Act recognizes the indigenous languages and also the sign language as the national language. So in our live streamed Central Epidemic Command Center press conferences, you actually see people sign a lot and learn about sign language vocabularies and so on. So first of all, I don't think it's only because I'm ethnically belonging to some group of people, it would prevent me from learning new cultures and new languages, just like maybe you work in AI. So your native language is probably Python, but it doesn't stop you from learning, say JavaScript or Rust or Go or some other languages. The more languages you learn, the more transcultural you become, you will be able to see the same problem from the objective oriented fashion, from a functional fashion, from a declarative fashion and so on. And that makes you a better programmer, definitely, than if you only know one language. And so I think this transcultural attitude is more important than ethnic diversity. Ethnic diversity and inclusion is of course very good, but without this transcultural idea of intersectionality, it is simply just one dimension of a multi-dimensional human being. And so I think intersectionality is more important than ethnic diversity. Okay, that's great. Thank you very much. My next question is according to my Taiwanese friends, you never showed anger or you are always calm. I mean, mindfulness is being calm mindset is very important under the pandemic, especially social media was chaos in this time. And which kind of three mindsets should we have to be the part like you? Yeah, first of all, I turn off all the notifications. I always work in the do not disturb mode. So I will not, for example, my phone will not go bing when we are having this conversation. And this is important to be distraction free. And the second thing is I limit my social media or indeed checking my inbox only once per half an hour. So I focus for 25 minutes, and then I check social media or email for five minutes. It's called a Pomodoro method that also helps a lot. And certainly, I think it's important to have sufficient amount of sleep every night. I sleep on average eight hours every night. If you have sufficient amount of sleep, then most of the short term memory you gather the previous day will be written into long term memory. But if you have a very short amount of sleep, then that's writing to the long term memory is not yet done. And so you will live the other day with a lot of thoughts from the previous day still in mind. And that interferes is kind of an internal distraction. And so having sufficient sleep also very important. Okay. Similarly, I used to work in Italy. So they are very optimistic about the way of thinking. And then recently, some of the celebrities committed suicide in Japan, and the suicide rate was like 15% increase last month in Japan. So people are like getting pressure from the pandemic about the economy. So well, similarly, but how can we maintain this optimistic mindset that in the pandemic? Well, I mean, it doesn't have to be optimistic. If you feel depressed, that is also a good time to reflect to do some long term thinking to make sure that you don't miss some threads that could be missed when you're much more active, right? So it's like the time and tide. If you feel pessimistic, it's good for long term strategizing. If you feel optimistic, it's good for actions. The trick is to make sure that the cycles are in a balanced fashion like in and young within the space of a day or at most a few days. If you spend half a year being optimistic, half a year being pessimistic, that's much more bipolar. But if you can limit that interval to at least a few days or a single day, like maybe do some like strategizing when you're going to sleep and list the actions you can take when you wake up, that's of course the best rhythm. So getting a kind of internal tempo going on, I think it's very important, a tempo, the speed of going more active and more reactive that will tend to balance one another. Okay, thank you very much. And then a little bit back into the previous question, but we are Japanese single ethnic country and paying attention to the details. This is really strengths. But at the same time, we overthink sometimes. So we have many overthinkers. Overthinkers has to become pessimists sometimes. Or how can we overcome this overthinking culture? Yeah, I think that's where the Lerner Cohen quote, I just quoted why it's so important. Because if you overthink, if you're a perfectionist, when you share the work you do, you know, other people may admire you, but there's really not much they can offer to support or help you because you already got it so perfect. And so a little bit of imperfection in open source, we say, and I quote, release early and release often, unquote, meaning that even if you only have an idea, even if it's very much imperfect, as long as you say so upfront, and instead of pretending like this is completed work, you can just say, oh, this is a stub, like in Wikipedia, this is just a stub that people can grow upon. That's an invitation for other people to co-create. And so I think the importance of release early release often really balances the perfectionism in overthinking. If you think a little and then publish, and then think a little and then publish, you can meet much more diverse set of people than if you think to 100% and then publish. That really makes sense. Thank you very much. My next question is in your interview with Andrew Reynolds, which happened in May 2020. You were mentioning about warm power versus sharp power. Also hashtag Taiwan can help. This is a very important concept in the 21st century. So the world should shift from competition to collaboration. So is warm power coming from Taiwanese culture, or since actually Taiwanese people are very, very warm to me? So how did you accelerate it with the power of the internet? Yeah, I think of course the Taiwanese people are very much like pro-social, and the pro-socialness is amplified through the internet by making sure that people work out loud. There's actually a book about it and working out loud meaning that just sharing what I'm working on like this interview is not just for the two of us or for reader of the book. Actually, we're going to publish either as a video or as a transcript everything we have said. And so basically dedicating it to the commons. And by commons, we mean that people can take it and like the rap, band, those monos in Japan, they maybe take whatever we have said and sample it into music. And that's fine too. And so the point here is that if you work out loud, you can meet people who take the kind of work that you do and in the direction that you did not anticipate. And that builds social relationships much more easily than if everybody have to sign a contract and pre-agree on whatever deliverable there are. Of course, there's a role of contracts in the financial sense, in the business sense, but a social contract in which that people agree to give more into the commons than we take from the commons is also very important. And that's the underlying idea of Taiwan can help in that we are much more willing to help than we take from the international community. That's great. Thank you very much. So in your perspective, what's our biggest impact of Taiwan can help? Which kind of things happened? Yeah, for example, there's a website called Taiwan Can Help That Us and that asks who can help Taiwan, Taiwan can help. And the great thing about this website is that it's not run by the government. There's zero government funding in it. It's just YouTubers, people who crowdfund and crowdsource the entire content. And this morning, I actually went to the massive online open course, the HaHao channel, which you can see in Taiwan Can Help That Us. If you scroll down a little bit, where they recorded from our vice president at the time, Dr. Chen Jianren, the crash course on COVID-19, because he literally is not just a vice president at the time, but actually wrote the textbook on epidemiology. And so this idea of people pulling in useful resources, sharing it with the international community, with the English dubbing and subtitle, and so on, signifies the idea that Taiwan Can Help is not just a government thing, but rather everybody in Taiwan can dedicate, for example, a few masks for the international humanitarian aid. You can also see that in Taiwan Can Help That Us. So it shows the diversity and the cross-sectoral participation. Okay, great. And then a little sub-question from that answer is I personally think, because I'm doing an AI project, this time in a pandemic, including infodemic, some people are looking at the science thinking, like thinking of a scientist, and got panicked from the false information or biased information in social media or a television or newspaper or something else in the towns. So science thinking is becoming more important now, especially for things like pandemic. What do you think about it, and how can we improve this science? Yeah, definitely. I think the scientific method is something that people can practice. It is not just an abstract idea that you just learn about in school and then throw away. It's something that you can just keep doing. The idea is simply to learn about a basic prior or basic facts, and that's what the government can help. For example, with the vice president recording crash courses about coronavirus or the central epidemic command center holding live press conferences every 2 p.m. during the pandemic and answer all the questions from the journalist, and that gives the basic observations. But now, based on observations, part of science thinking is you can form your own hypothesis, and through induction and through experiments, you can deduce new ideas and propose it for the community to evaluate, to replicate, basically. For example, there was a person named Slai Quan Yu who tried to use the rice cookers and to disinfect the musk, and you can actually see the rice cooker right there. Basically, Slai tried to put the medical musks in such traditional rice cookers, which unlike the newer IH cookers, they don't have a ventilation hole, and so that makes the heating model very predictable. It will very quickly heat up to 110 Celsius and then very quickly cool down. And because of that, in the few seconds where it reaches more than 100 Celsius, it kills off the virus but doesn't destroy the material of the medical musk. And so he basically outlined the procedure of doing so and then share it with the community. Now, because the medical musk are theoretically one-time use, of course, the Taiwan Food and Drug Administration is quite skeptical of this idea that you can somehow reuse it, but then the TFDA did their own experiments, and much to their credit, they invited Mr. Lai to the daily press conference to teach the evidence of the hypothesis while the Minister Chen Shizhong actually cooked a medical musk in a rice cooker in front of all the people watching the livestream. So this is just one simple example, but it shows that science is something that we all participate and we can all replicate. It's not that just for a handful of scientists, everybody can be a citizen scientist. And when it comes to reusing musk, these experiments are being replicated around the world. And nowadays, we see from peer-reviewed journals that this even works for N95 musks. That's a great story. And then similarly, my question is, this time people got confused between fact and opinions. And then main Japanese people believe that news is a fact, but news can be biased because of the sponsor, because of the own of the newspaper company. So as a scientist, scientists tend to think to believe in the fact supported by the evidence. So this kind of thinking is important, I think, but what do you think about it? Well, I mean, even if there's no bias in the people running the news organization, the source that they collect those ideas from may be biased. And even if they use the most rigorous methods like data science, still if you have the wrong expectation of the data, then you may collect the data, but then accumulate data bias without realizing it. So there really is no escaping of bias. And what people can do is like listing an ingredient in the food label to list exactly what are the sources. Like in the journalism, if they use a research paper, it helps, of course, to include the DOI number so people can fact check themselves. And if we're doing data science, it also helps to list the data sets that we actually draw our conclusions from. So our peers may evaluate independently. So I don't think any single individual can be free of bias. But what we can do is to list our sources so that other people may benefit from independent verification. That's the beauty of that scientific method, it grows with the community. Yeah, that's a great way to do it. Yeah, thank you very much. My next question is your collaboration with Seattle Company, Paul dot IS. Paul's police is very innovative. And then it is like a real time visualization of the democracy. And since we learn the machine learning hackathon community here in Tokyo, we are very interested in your insights. So what did you find by data analysis in people's opinion discussion? And then that will be lead us to the next generation politics system. And then we should do the same here in Japan, I think. Yeah, we use police from everything to, you know, do with the open mountaineering, that's the hiking to the opening of the oceans, that's like sea sports and so on. And to diplomatic conversations like with the de facto US Embassy, the AIT on the digital dialogues, and pretty much everything, right? There's even a website, polis.gov.tw that is our locally hosted instance. So it's very much ingrained into our institution already. And the best thing I found about polis is that it's AI in the 20th century. It's not AI in the 21st century, whether with deep learning and so on, because it's actually much harder to explain the results if you use 21st century AI. If it's just 20th century AI in polis, it's mostly principal component analysis, mostly K means clustering. These are ideas that even for a primary schooler, if you take some time to explain it through the concepts, they can verify it themselves. But if you used any of those cutting edge like transformer models of deep learning, explaining it is actually very, very difficult. And so I think it helps to build data competence, that is to say, as self confidence in producing and analyzing data, instead of just consuming the results of a data analysis, it's like media competence is the ability to produce media versus media literacy, which is just about viewing or reading media. It's important to start small, to start with well understood techniques like the 20th century AI, and then build confidence in all the participants' ability to not only replicate the results, but also do their independent analysis. And then maybe you can add some more like deep learning as a assistance, for example, smart categorization and so on. But these are nice to have. The core functionality, I think, must be very accountable, that's to say, easy to explain, and also value aligned, meaning that it serves the best interests of the person using it. Okay, thank you very much. The counter question about that comment is, I think the beauty of the power of the machine learning is finding counter-intuitive fact. So did you find any counter-intuitive fact in your own opinion? Yeah, definitely. Yeah, the main Polish insight, which is surprising to pretty much everyone participating, is that actually people agree mostly with each other's points, most of the time on most of the issues. And it's very counter-intuitive because people may be used to the more anti-social social media, which wastes people time just going through the ideologically charged controversial issues. But in each social topic, these controversial ideologies are maybe just a few points. If you just say, okay, we agree to disagree, but then there's much more that we agree. For example, when UberX came to Taiwan, everybody agreed on the insurance, taxation, the fairness, principal registration and so on. And while, of course, people may differ on what constitutes a sharing economy, some people think, you know, you're not even carpooling, that's not sharing economy. But some people think, oh, it's sharing economy if there is a platform that does matchmaking of the idle hours of cars and so on. If you'll get caught up in those ideologies, of course, the society goes nowhere. But counter-intuitively, Polish shows that we actually agree on those rough consensus on those common values after all. So why don't we just regulate those? And then we can talk about those abstract concepts that are controversial later. Okay, so do you think eventually optimization of the policies can be done based on the data science? Definitely. I think data science offers a very rare but important glimpse, which is much more holistic to the entire society. A lot of the difficulty that we have when talking about like climate change, this information crisis, COVID, whatever, is that because the problem space and solution space is so large, it's very difficult to hold them in one's mind. So one would oversimplify. And once people oversimplify, but in different ways, it creates ideological differences, even though their value may be the same. And so data science can give a holistic like dimensional reduction, right? It shows a interactive portrait of the actual problem space without oversimplifying two or only one or two dimensions. So do you have any other example of the fact you found in the police and then made it made it as a policy? Yeah, definitely. So for example, when we did the consultation with the AIT, the de facto US Embassy on promoting people to people ties in like across the world, there's a very controversial statement that says, and I quote, Taiwan should make English and additional working language even France have done the same, unquote. And it was very controversial. People don't like it, but there's also people who really like it. But the statement was the most consensus is that we need to prepare our education environment so that in due time, people will feel comfortable speaking native English. And so basically, it's the same statement, but set in the future, like for people who are in kindergarten, we need to prepare so that 10 years, 15 years down the line, they feel comfortable thinking English. And now everybody agrees with that. And so I think this is important to realize that if you set far enough to the future, if you are patient enough, something that looks like a very controversial topic may actually get everybody's support. And of course, we develop our bilingual nation strategy based on that insight. Do you think eventually data science can pick the weak voice in the society? If we decide to do so, of course. Okay, thank you. My next question is according to Gov. Zero Teen, your co-fact is the one of the most active repository in the GitHub. And in this platform, people can crowdsource the fact check the rumors. It can be done in 60 minutes as an average. And then this could stop infodemic under COVID-19. And then how did you come up with this idea? And then how can we expand this open source plus crowdsourcing idea for other use cases and eventually the other countries? Yeah, the co-facts team have already spread, for example, to the Jurong Gong University where we had a workshop. And if you remove the S and go just to co-fact.org, it actually goes to the Thai version in Thailand of the co-facts repository. And so it has already seen a lot of adoption around the world and especially in Thailand, which is very nice. But I cannot speak on behalf of the co-founder, Johnson Liang, who really goes by the nickname Mr. ORC. It's actually the main mastermind behind this chatbot and also the main contributor for the first version of the co-facts. Of course, now it's a very large team. So I would encourage you to maybe ask Johnson. I mean, here is his GitHub repository and his contact details. And maybe he can share the full story because I'm just a promoter of co-facts. But I'm not part of the team that started co-facts. Okay, I will definitely do so. Yeah, thank you very much. My next question is a little technical. You are known as a programming wizard and then my Taiwanese engineer who's actually working for the GitLab Taiwan and says you can write 200 lines of code and can build amazing performance applications. So what are the top three key points when writing effective codes? Well, first of all, it helps to write it with the community. If I write just the bare bones without design, graphics, favorite icons, and so on, but I do it in the open. And I say, I know this looks ugly, but send me pull requests, then everybody can contribute. And that's really the main key point is to work out loud instead of working in isolation. The second thing also helps that you identify the project's social good in the very beginning. So even before you write the first line of code, think about your read me, think about your tagline, think about your project name and hashtag, and how to communicate the social good of your project to any random stranger so that they will be motivated to contribute to the code base. And third, always instead of asking people to sign a lot of constant forms and so on before contributing, just give them the capability to contribute. That's radical trust. And it's always easy nowadays that you can revert their changes. If it turns out their changes doesn't make sense. But asking for permission is much more difficult than asking for forgiveness, right? If people can't just go and change and say, oh, sorry, forgive me. I got it wrong. I made a typo or whatever. It's no big deal. But if everybody has to ask your permission, then you become the bottleneck for the project. And so by relieving yourself of the bottleneck, you can delegate away even the delegation to other people. That's great. Thank you. And then your way of using radical trust is really interesting. What's the main difference between radical trust and trust? Yeah, radical trust means that we give the right to commit. For example, when I was working on nowadays called Raku, but back in the time, the perks implementation, I give commit bits just to say the right to push into the shared repository, not only to people working on Perl and Haskell, but also Guido van Roosen, the creator of Python. Why not? And even when someone, you know, flames writes some toxic language about Perl on any of the forums, we often just reply saying, hey, here is a commit bit. You can contribute and make it better. Even evolved so that when there was a committer who was just giving birth to one of their child, we immediately gave the child a commit bit too, even though they're just newborn. And so basically, it's like Wikipedia, right? Anyone with an IP address can change our code. And we can figure out the forgiveness later if they've made some bad changes. Okay, great. Similarly, lots of people in Japan, and maybe globally, wants to become an engineer, IT engineer, this time because of the maybe bad economy, IT engineer can be highly paid and then can work remotely regardless of the place. So do you have any message to encourage this kind of newbie or wannabe engineers? Yeah, I think in Taiwan, we don't call it software engineering. We call it program design. And it's the same thing. But it conveys a different idea. When you call it engineering, you think of someone who interacts mostly with machines. But if you call it a designer, that's someone who mostly interacts with other people. And so of course, a real programming work involves working with people and working with machines, of course. And whatever code you write must be readable by machines, but also readable by humans. And so there's two sides of this job. I'm not saying that there's only one side. But since you're talking about software engineers, I would like to encourage them to not forget about the design part, computational thinking, scientific thinking is great. But also design thinking, learn about how to explore the various different values that people have, how to define a common value out of the diversity, how to develop according to various different prototypes concurrently, and how to deliver the final work and also allow other people to deliver their version of their work. And that's design thinking. Okay, thank you very much. So how can typical, because for example, one of the typical backend or machine engineer do not learn designing, then how can they learn designing without being a designer? Yeah, I mean, design thinking is not about being a professional designer. This is not working through this iterative process to how to understand your fellow human beings and how to define those challenges and redefine problems and create innovative solutions. And if you can learn programming by going to, I don't know, stack exchange or whatever forum that people are having a such a conversation, I'm sure that you can also browse, for example, the interaction design foundations description about design thinking and find the references and join the communities. I mean, design thinking always starts with a idea. And this idea is often just part of your life, right? You can see something that you want to see done better. And then it reaches you. And then you can reach to other people asking whether there are some other issues that you also want to solve in that vein. And that is the basis of interaction design. And that will then teach a different lens of problem solving. So if you're interested, I just pasted the interaction design.org beginners guide, you don't have to follow all the lessons. But that outline shows what's the learning path about design thinking. And you can start from a user experience perspective, from a user interface perspective, from an perspective, a user researcher perspective, there's many, many different entry points into the learning path of design thinking. Yeah, okay. Thank you very much. And I'm also a big fan of design thinking. But my biggest learning today is what we are saying is design thinking, except UI UX design exists. In case of machine learning, in case of backend, everything can be designed. Yeah, that's definitely the case. Mechanism design, market design, service design, everything can apply the design thinking to. Okay, thank you very much. My very last question is everyone places your achievement mask supply application and its data visualization. So technically speaking, public API and the open data are essential to build it. So in your ICFP keynote, you said someone from Korea contacted you, but they couldn't make it, they couldn't build the same application in Korea. So we do not have a pharmacy open data API either. So how can other countries, for example, government or big company proceed this open data movements? For example, pass waiting big company to provide more public API for allocating the budget to support civic tech hackathon. Yeah, first of all, I was referring to early February, where we did API for Taiwanese pharmacies. And at the time, of course, the Korean people do not yet have the open API. But what's amazing is that Fijian Kyung from Thailand City, part of the Gov Zero Slack channel, shared the API that he had with the people in Korea. And so for example, the Seoul city quickly, well, not that quickly, but eventually around March, set up such a API and followed by other cities as well. And so I think it's just a few weeks time, I think by mid-March, they do have a mask availability map going in their mask rationing system. And I also met the developers over video conference. Some of them are just like 15 years old. So it's excellent. And so what happens, I think at that time, is that the Korean government were seeing that there is a need, but they do not have an example. And what the civic technologies can show is that once you use the API, then instead of coding anything from scratch, the map that Fijian Kyung made in Thailand City can immediately work also as a map for the Seoul city, even though Fijian Kyung doesn't speak Korean, but he speaks JavaScript. That's the important thing, right? And so I think that enabled a real collaboration. In Japan, there is a very similar story about the code for Japan. People who work with the Tokyo Metropolitan government to work on the stop COVID dashboard. And again, because they work out in the open, the government also helped on the internationalization and localization, and I personally also have translating it. So as long as it's open, it's on GitHub, I think there's no boundaries. It's not just people in Korea or people in Taiwan or people in Japan. Everybody who speaks JavaScript can co-design the API and the government basically accepts whatever work is there. And then basically it's like a reverse procurement where the government only have to guarantee the stability of the API, but doesn't have to do anything else. Okay. Did you get any request from other countries regarding the mask application? Yeah, definitely, definitely. And many people are seeing it not as something that they can just copy, but rather a model like the Taiwan model where they can engage the civil society, the social sector, who do not previously understand that they can also make government digital services. For example, on the stop COVID dashboard in Tokyo, I think all what the Tokyo people do in the metropolitan government is basically giving them a domain name that is a government. But the actual link is still to GitHub. And the GitHub is the main place where people do the poor requests and so on. And that means that it's made in the world. It's just given a Tokyo metropolitan domain name. And then of course it's forked into many other cities as well. Okay, great. Thank you very much. That's all for my questions. And then thank you very much for answering. Yeah, sure. Would you prefer we publish the video or would you prefer if we publish the transcript? Yeah, could you publish the video? Okay, of course. Transcription. Yeah. Okay. Okay. That's awesome. So I'll just publish the video on YouTube then. Okay. I'll central the link. Okay. And then my last request is, may I take a picture? Of course, of course, you can take several pictures. Just for a moment. Okay. One, two, three, smile please. One, two, three, smile. Okay. Thank you very much. That's it. Thank you very much. Excellent. And then we will publish a book. And then we will promote the book because this time publisher is an IPO company, public company called Shoeisha, so they have a selling power. Okay. Okay. Take care. Live long and prosperous. Thank you very much. Bye-bye.