Taiwan’s Covid-19 Technology Dividend
Interviews with Taiwan Digital Minister Audrey Tang 唐鳳, and with Taiwan A.I. Labs founder Ethan Tu 杜奕瑾 .
Taiwan’s innovative application of technology in combating the coronavirus has started to yield a economic dividend of increased direct foreign investment in the tech sector.
Issues discussed include what parts of the Taiwan Model in combating Covid-19 can be replicated by other countries.
A.I. Labs has developed a mobile phone application for tracking the coronavirus, practicing social distancing, and isolating confirmed cases without requiring collection of personal IDs or centralized data storage. A.I. Labs is sharing the algorithm with developers wanting to build similar apps for their countries.
Taiwan’s open source ethos has spawned many new startups in the precision medicine sector.
To hear the full version of the Taiwan AI Labs, music composed by artificial intelligence link here: https://drive.google.com/file/d/16dU8JQPwtAe7AWgje5AVgJ4nZn9Dy7mV/view?usp=sharing
You’re listening to the Taiwan matters podcast in depth discussions on the issues that matter in Taiwan,
Ethan Tu (00:18):
Taiwan is doing a great job because we are transparent, open and all the best experts they trust us. They will pick our solutions.
Audrey Tang (00:33):
We’re now seeing Taiwan as really the natural choice if you’re working in the digital field, specifically, because we’re the only country in Asia now that allows for complete freedom of speech, of assembly, and other civic freedom.
I’m your host, Nicholas Gould. You just heard Ethan Tu. Founder of Taiwan, AI labs and Audrey Tang. Taiwan’s digital minister without portfolio, both hold legendary reputations in Taiwan’s digital space. Ethan Tu, while a student at Taiwan university founded PTT, the world’s largest Chinese language bulletin board system. Audrey Tang, was a software coder and silicone entrepreneur. While in his teens, both are leading proponents of Taiwan’s free data movement. Our topic is Taiwan’s technology dividend. The profit Taiwan stands to gain from its innovations in the fight against COVID-19. At this time in early July, Taiwan is in its third month without a domestic case of COVID-19, but the Corona virus continues to spread at alarming rates in the U S and other parts of the world. The scale of the pandemic is driving countries to look for better ways to fight the virus. One idea being tried around the world with varying degrees of success is the use of mobile phone apps to automate the process of contact tracing, social distancing, and isolating. The problem with many of these apps is overcoming people’s fear of losing their privacy and a general lack of trust in governments having too much personal data. Taiwan has a software solution and it is sharing it with the world. Taiwan AI Labs is offering an open source code, which leverages Bluetooth to collect decentralized data without needing users’ identities. It is just one example of how Taiwan’s open source free software ethos is helping the global anti-Corona virus effort. We’ll be looking at other such innovations and what it is about the so-called Taiwan model that fosters its technological creativity. We join digital minister, Tang at her office. Minister Tang, thank you very much for being on Taiwan Matters Podcast. Taiwan has gained alot of international praise, international recognition, for the success that it’s had fighting Corona virus. A large part of that success we could credit to good old fashioned contact and tracing. Very speedy response. Closing the borders, controlling international travel. But another part that played also a very significant role was Taiwan implementing big data technologies. The use of the internet. A lot of things that weren’t available in the SARS experience in 2003. And I’d like you to talk about how you think Taiwan was able to respond in that way with its big data, response.
Audrey Tang (03:44):
Sure. Another important factor that I would like to highlight before talking about the data is that you’re now wearing a medical mask, even though we’re more than one-and-a-half meters apart. And that contributed a lot to the success of Taiwan’s response as well. In fact mask distribution is one of the most important measures used to counter the coronavirus pandemic. Because in Taiwan wearing a mask signals to other people that you are taking care of yourself, right? You’re not putting your hand to your mouth and you’re washing your hands properly. And these social signals make sure that people have the right habits. So that way the coronavirus does not spread. Even though there may be individual community cases, we’ve never entered a case of broader community spread and therefore no lock downs. And how do we enable everybody who feels like wearing a medical mask to get a medical mask? That’s where big data comes in. We have a national health insurance system, which covers more than 99.9% of the citizens and also many residents. So to anyone with an NHI card can go to a nearby pharmacy and collect now 9 masks, if you’re a adult, every two weeks, or 10 if you are a child, but how would you know which pharmacy near you has masks in stock? Well, they publish in a distributed ledger. That is to say everybody who has more than 100 applications can see, or through voice assistance can hear, through voice assistance, and see through real time maps and so on which are the pharmacists near you that still have masks in stock? And also when you swipe your NHI card and get 9 masks, you can see after a couple of minutes they have depleted the stock by 9 or 10. And if it rather increases after a couple of minutes, you would call 1-9-2-2 and report something wrong. It’s going on real time distribution of data. Data that is jointly audited and accountable to the social sector, that has been at the root of the trust which has been built by the mask rationing system.
I have journalist friends who called me at the beginning of the pandemic, wanting to know how it is that Taiwan was doing so well. And I would explain to them about the use of data about, getting a QR code when you enter the country, the quarantine system and using GPS. Their response was well, that would never work here. Americans or Europeans would never accept that kind of intrusion.
Audrey Tang (06:14):
We do not use GPS, for the record. We use cell phone tower triangulation.
The difference being in the level of privacy?
Audrey Tang (06:22):
Exactly right. with GPS the phone knows exactly which room you are in because it’s a very precise geo-location. We used data collected by the cell phone towers. It’s already collected. This is not new data that we’re collecting. Your telecom already collects that data for the service. And we already use that. For example, when there is a heavy rainfall following a typhoon and you are in a risky area that can suffer from the landslides. for example. There is a automated SMS which goes out to people in that area that says you should probably evacuate. But it does not track the GPS location of your phone. We just have this, what we call a digital fence, a polygon in the map that we send the SMS to. So the same system has been repurposed for the quarantining purposes.
Yet there was another situation where they merged the NHI data with the customs immigration data. Do you see that as being more of an infringement on privacy?
Audrey Tang (07:27):
Well… according to the CDC (the center for disease control), it is the duty of patients to tell the medical officers and the clinicians of their recent travel history, if they have been to Wuhan and so on. But of course people may have imprecise memory, or maybe they do not want to divulge more private information, and so on. So, the integration is actually part of what the CDC already requires for the patient to tell the clinicians. That we have their travel history now, of course, is debatable whether an automated showing of travel history to the doctors, is a proper proportionate use of the private data. And, I would simply say that when the customs and CDC rolled out these measures, the approval rate of those measures was 91%. And now it’s at 94%. I’m not pretending there’s no 9 or 6% of people who are thinking that it’s not appropriate. And it’s good because we have not declared an emergency situation. So these 6% of people keep us honest because we have to make an account to the parliament, to the MPs. But I would say that this has won the popular support of most people,
On that point of popular support… do you think that the experience in Taiwan, what’s being called the Taiwan model, is it replicatable in other countries, for instance, the United States or Western Europe? In the sense that, is the population in Taiwan more trusting of their government in protection of their privacy.
Audrey Tang (09:01):
I would say that these ideas came from the civic sector. That is the root reason of why people adhere to them. Because the mask map is certainly not a government idea. It’s prototyped by a Tainan civic technologist, Wu Zhan Wei, Howard Wu. And. the linking of the data around the customs, and the NHI systems was not a government official’s idea. This is from a clinician who told the vice premier of the data idea. And the mask use, for example, as a social signal, is actually by popular experience after SARS. And so when the CDC tells people to wear a mask in, for example, public transport, they do that. But when the CDC says that you can take off their mask when you’re keeping the social distance, people wear to mask anyway. And so because of that, it is the civic sector that wins the legitimacy. Because, most of these were their ideas. So for example, around the mask, use of the mask is billed as something that protects you from your own hand, that’s a part of the Taiwan model which every other jurisdictions can very easily adopt.
There seems to be difficulty, in a lot of countries, rolling out a contact tracing phone app.
Audrey Tang (10:16):
Because it collects new data. Here in Taiwan, we do not collect new data.
In other words, the system that you’ve designed for contact tracing is the distributed information,
Audrey Tang (10:28):
Right. In essence, people are not used to the new collection of data. Of course, when you introduce new data collection applications, you need to make sure that it’s serving the best interest of the individual using it. Otherwise they would not opt in, of course that’s human nature. So that why we have not rolled out any Bluetooth or otherwise social spacing app level contact tracing, precisely because people did not have experience using that sort of application.
But, now there is… There will be shortly… Or the, the app is built. It just hasn’t been released.
Audrey Tang (11:01):
It’s built by the social sector. It’s by a not for profit, a social innovation organization called the AI Labs and they designed it so that it’s international. So they work for example, with their UK counterparts. But I do not think that they are rolling it out in Taiwan.
They’re not rolling it out in Taiwan, basically because the pandemic is more or less over now. Now it’s post pandemic, so there doesn’t seem to be a lot of need for it. But, I’m wondering if you think it will have success outside of Taiwan?
Audrey Tang (11:28):
It would need to have a incentive design much like if you wear a mask as the sign to protect other people, to respect other people, that’s a harder sell because then the person wearing it is essentially making a personal sacrifice for the community. But because in Taiwan we say you wear a mask to protect you from yourself, because you’re not touching your mouth. Let others know you are washing your hands. As a reminder. to distancing is much more easily. So the incentive design, the ideas we’re spreading, it’s as important as the underlying technology. If you have an incentive design that can get people to think, Oh, it’s in my best interest to do so… When the contact tracers come for me, when the medical officers asked me questions, maybe this app helped me to protect my privacy and my friends and family’s privacy by sharing with them only the essential information, instead of any other information I might have divulged during traditional interviews. Then that may be a better incentive alignment, then the traditional way of saying, no, you just have to install this app instead for the public, common good, that may not work as well.
We’ll rejoin digital minister, Tang shortly, but for more on the mobile phone social distancing app, I went to Taiwan AI Labs to speak with Ethan Tu. Is it officially released? Is it ready?
Ethan Tu (12:52):
So, it has already passed review. We already are on the Apple App store. It is ready for release right now, but we haven’t released it yet because… you know, in Taiwan… We are more focused on to support other countries to release it… Because, in Taiwan there are no confirmed cases. So for Taiwan it is not a priority.
We may not be seeing AI Lab’s, social distancing app in Taiwan, but dozens of institutes around the world have downloaded the algorithm and versions of the privacy design will be part of the more successful applications we’re starting to see in the UK and Germany, for example. Taiwan, AI Labs has developed other useful tools in the fight against COVID-19. Many of these fall under the category of precision medicine, such as a chest X Ray classifier, which can automatically scan for COVID-19. You should check out their website to see all the different projects. All the work is done under the principle of open-source, free-software. Before establishing AI Labs in Taiwan, Ethan Tu worked on artificial intelligence at Microsoft, and I asked him how his lab is different. Do you think there’s a trade off between an open system and…making money?
Ethan Tu (14:13):
Making money does not conflict with economic benefit. Actually, openness will promote the economic growth of new talent, and new innovations. That’s was also one of the major reasons we want to fund AI Labs in Taiwan because I think the best way to promote the software ecosystem is starting from the open-software institutes like AI Labs for future innovations. I believe after a couple of years we will see a lot of good innovations. Like right now we already have a lot of good start up companies leveraging the results of our medical research and creating their own solutions. This creativity is much faster then the private owned company, because if we do innovation in the open way, all the people they can contribute their base part. It’s just not like a private company. When you talk to me, I need to sign an NDA. No. When you want to work with Taiwan AI Labs you need to agree on openness, before you can collaborate with us. This is our main principle.
AI lab does more than healthcare. They also apply artificial intelligence to image recognition, natural language, and even music… Let’s listen. (music clip) This music was generated by feeding a computer, an image of a Corona virus. The rapid advances of artificial intelligence make many nervous about the loss of human agency in critical activities and decisions. Already, AI is replacing people and eliminating jobs. Ethan Tu is well aware of these concerns.
Ethan Tu (16:04):
If Taiwan government wants to leverage any AI solutions they should only do that with solutions that can be validated, is transparent. You know, AI will become a superpower and if the algorithm itself is not transparent it will become a black box. And that black box will control your life in the future, including your Medicare, including how you drive your car, including the decisions you make. You think you are making your own decision, however its actually the decision is picked out by the AI algorithm. So we are promoting openness of the algorithm and transparency of the platform.
Where do you think the next major contributions from Taiwan artificial intelligence will be?
Ethan Tu (16:49):
I believe the Medicare, specifically precision medicine is one part that Taiwan can contribute to the world. Taiwan is going a great job, and the reason it is doing a great job, is because our citizens are trained to be transparent, to be open. And all the best experts around the world, they can join us to do research. I think this creativity will become more and more impactful because the people around us, they trust our solutions. They will pick our solutions. In this way the innovation here, it can fit in most parts of the world.
Taiwan, AI labs, founder, Ethan Tu. AI labs is a not-for-profit organization, incubating AI startups, but the Taiwan government is prioritizing artificial intelligence as a strategic industry and spending on related infrastructure for more on this, let’s go back to our discussion with digital minister, Audrey Tang. The government has placed new technologies. I use that broad term new technologies, including artificial intelligence, and other applications, as a strategic industry. As the Digital Minister, how do you see that playing out? In other words, do you see Taiwan being able to leverage the success that it’s had in coronavirus, and technological applications to that end, economically going forward?
Audrey Tang (18:19):
Oh, definitely. For example, we have developed in our app for mask pre-ordering if you have plenty of masks or if you don’t go out very much, then maybe, you do not collect the rationed mask from the pharmacies and the convenience stores during the mask rationing. And so you have some quota that was not used. And to date, there is more than 5 million medical masks of uncollected quota, voluntarily donated by the individuals who are using the app for international humanitarian aid. And Taiwan has sent all of them to all the front line nurses and doctors in other countries. And because they understand the quality of the masks made in Taiwan is so good now the ministry of foreign affairs, they’re also working with many other economies to basically export our plant building ways. So given a parcel of land, electricity, water, and their PE supply we can make a industry for you that makes 24 hours a day around 2 million medical masks or RN95 your call. And, the entire design is localized, meaning that your local technicians will be able to maintain it without any intervention. And so that sort of export of course builds this new recognition of the brand of made in Taiwan. And now when you have produced 2 million masks per day in those automated plants, you would want maybe a mask map or some way of rationing those masks in your nearby pharmacies or post office or convenience stores. In which case you can talk to our IT vendors. And so all of these exports makes sure that people think of made in Taiwan, not just as products made in Taiwan, but also process made in Taiwan.
Some of the applications, a lot of them are using artificial intelligence, especially in the precision medicine field.
Audrey Tang (20:12):
When I say AI, I always mean assistive intelligence, meaning like a good assistant it needs to act in your best interest, meaning that it’s values need to align with yours. And when it makes decisions that differs from your values, you need to ask it to give a full account of why. So value, alignment, and accountability are what a person would expect of a human assistant. So you would expect the same of assistive intelligence that’s based on machine learning. And so I think including Ethan Tu the director of AI Labs align with these essentially social sector ideas of AI that we build a social norm and prove that this assistive intelligence work with the social norm instead of against the social norm. And this makes our assistive intelligence very easily exportable because you can just retrain it using your local norm. This is not the top down way of us imposing a social order on you. And so I think this makes Taiwan very naturally aligned with other liberal democracies.
Do you think that we will see more international direct investment in Taiwan’s tech sector as a result of these new applications?
Audrey Tang (21:26):
Well, definitely. I mean, as part of the Asia Dot Silicon Valley initiative already since Google acquired a large part of HTC and Microsoft and many other multinationals setting up their AI research and development labs in Taiwan, we’re now seeing Taiwan as really the natural choice if you’re working in the internet field, in the digital field, specifically, because according to the CIVICUS monitor, we’re the only country in Asia now that allows for complete freedom of speech, of assembly, and other civic freedoms. Previously there’s many journalists that are located in nearby jurisdictions, such as Hong Kong and so on, but now they are relocating en- mass to Taiwan, which is the headquarter of the Reporters Without Borders, for example, because they know they can think freely. And, the same applies to the researchers and scientists.
Yet one of the roadblocks or bottlenecks in that kind of direct investment has always been work visas and the openness of Taiwan to accept foreign workers.
Audrey Tang (22:32):
I think talent circulation is very important. So for many people, myself included who were born in Taiwan and worked overseas and go back.. It’s not just me personally. I also bring many international friends with me to Taiwan to experience in Taiwan or even after they stay for five or more years they can be also Taiwanese. Meaning that they get a citizenship of Taiwan, the Republic citizen, without giving up their original nationality. And, there’s also the gold card program, which I think the first one who got it was the founder of YouTube. And so all of these make sure that when…Ethan Tu actually is another good example because he was part of the Cortana team in Microsoft, and when he founded, AI Labs here, he also poached… Well, invited many of his colleagues back to Taiwan, even though they were not Taiwanese to begin with. And, so I think this kind of talent circulation is not a zero sum game. We want people who want to innovate to base their work in Taiwan, but also introduced Taiwan innovation to other people. And, so work visa or even volunteer visa, all sorts of different international NGOs, for example, we’re working actively so that they can base their headquarter in Taiwan.
You are a proponent of what you call digital democracy, I think you’ve made huge strides. Have you raised your expectations for where you want to take the country in its digital development, next?
Audrey Tang (24:03):
Definitely! We have achieved largely the broadband as human, right, which is Dr. Tsai Ying Wen’s promised for her first term anywhere in Taiwan. Now you have 10 megabits per second at just 16 euros per month, unlimited connection. Otherwise it’s my fault. And, when we are deploying 5G, we will start for example, in Taitung where it’s very mountainous. Maybe there are still spots where 4G signals are weaker than usual. And so we start with 5G in those parts, making sure that everybody benefits equivalently from the digital opportunities offered, especially in healthcare and education. And, we also make sure that people who want to work with say self-driving vehicles and so on, have the sandboxes to play with so that we can figure out the social norm around those vehicles. My personal office, the Social Innovation Lab served as a living lab for self-driving vehicles for quite a while, for a year and half. And, now I’m very happy to see, for example, on the dedicated bus lane, there’s going to be self-driving buses that have capacity of 9 and 35 respectively now on the road of Taipei. And before the end of the year, they will start carrying passengers when the Metro is closed-off after midnight.
Minister Tang, thank you very much for your time.
Audrey Tang (25:19):
And, thank you for listening. And, for more information checkout, Taiwan can help.