I started the recording, so Skype will send us a mutual recording afterwards.
Excellent. Thank you. I’m going to do a backup recording on my phone as well just to make sure that…
…here too for afterwards. Good stuff. Thank you for taking the time out to talk to me. Really appreciate it. I’m really excited to speak about this.
What’s been going on at the moment in Taiwan? Because you say there’s big stuff going on in the Parliament at the moment.
Yeah. At the moment, the Parliament has been not having the all-of-Parliament meetings for quite a while because the KMT were protesting the ractopamine import of the American pork, the US pork issue. That’s the current issue at hand.
Otherwise, it’s fine. We just had a Pride parade [laughs], a really large one, so life is normal.
That’s fantastic. I’ve read about the Pride parade, and I thought, “Wow.” We’re here in the UK. We’re just going under another lockdown. We’ve got a lot to learn from you guys.
Can you start at the beginning and tell me the story of when COVID hit Taiwan? I was talking to a couple of people about this, but it would be great to hear this from you. You were there. You were involved in this.
Last December, when Dr. Lee Wenliang, the PRC whistleblower, posted on their social media that, and I quote, “There were seven new SARS cases in the Huanan market.”
It gets noticed, and the young doctor with a name nomorepipe reposted that December, on the 31st, on the early morning, to the PTT, which is the Taiwanese equivalent of Reddit.
It immediately caught the attention of our medical offices so they not only sent an email to WHO the very same day, when Dr. Li Wenliang was still being questioned by his institution.
The very next day, January the 1st, we started health inspections for all flights coming in from Wuhan to Taiwan. We think that could lead for saving the Taiwanese people because we set up this, interrupt them, and come uncensored even before we have the first locally-transmitted case.
Afterwards, it’s mostly about the fast iteration, the trust from the CECC to the people when the citizens report any new situations, like the young boy calling in, saying that he doesn’t want to wear the pink mask to school. The very next day, on the livestream conference, everybody wore a pink medical mask, and so on.
This kind of 24-hour response cycles really make sure that people understood the science. On early February, we understood already that if three-quarter of people wear the mask and wash their hands, then the R value will be under 1, and it will not be a community spread, and we never would need a lockdown.
Because Taiwan in 2003 did have a lockdown of the Hoping Hospital when SARS 1.0 came.
Everybody was very traumatic back then, and so this time, we don’t want any lockdowns. Everybody just had access to the mask through the National Health Insurance Card, which covers not only citizens but also residents. By early March, we did get the three-quarter rule implemented, and people wore the mask, washed their hands. They understood that the masks are there to protect oneself from one’s own unwashed hands.
By early April, I think we get 90 percent of people wearing mask and washing hands. At that time, then the R value is definitely under 1. We never had a lockdown.
On the infodemic front, there’s the cute spokesdog, Shiba Inu named Zongchai, that translates the science into funny meme pictures that of course also help the people to build the mutual support instead of just criticizing people who work in the nightlife or people who return home and so on.
We did this without a lot of division or name-calling or things like that. By May, life has largely returned to normal, and we start planning the triple-stimulus voucher to encourage people to go outdoor more. It’s been more than 200 days with no local cases. Just a very brief summary.
That’s absolutely amazing. I know you mentioned memes to educate people and that sort of thing. You’re quite susceptible as a country to disinformation, aren’t you – not to falling for it, but you get a lot of disinformation attempts from China, for example. Is that how you partly how you manage that with that kind of a wider campaign?
Yeah. It’s the triple-two principle. Whenever there’s a trending rumor, we make sure there’s humor over rumor, meaning that within two hours, we roll out two pictures, each with less than 200 characters, that goes even more viral than the disinformation. The idea, very simply put, is that if you read both the conspiracy theory or disinformation and the funny clarification on the same day, when one goes to sleep and wake up, the long-term association tend to be on the joy and humor part instead of on the anger and outrage part.
If people go to sleep only having seen the outrage part, they forms a ideological association that makes it very difficult to clarify without reinforcing existing stereotypes, so the response time is of essence.
How do you detect where the disinformation is coming from? Is it an automated machine learning-type system?
Yeah, it’s just a collective intelligence with some assistive intelligence. Much like how do we know there’s SARS cases from Wuhan Seafood Market? Well, it’s because people notice and flag it, and upvote it on the PTT, which is Reddit-like.
Just like, how do we detect spam? It’s because people flag incoming email as spam and dedicate their fingerprint to the Spamhaus, which is the international organization designed to put a reputation score to the email senders.
In a similar way, there’s a lot of people in Taiwan working on technology such as the Cofacts, which is a way for end-to-end encrypted channels such as LINE chat rooms – if you see something and you suspect it as disinformation, you can forward it to the bot in order for it to be crowd-checked.
That then fuels also the Dr. Message, which is a antivirus company called Trend Micro. It’s a commercial offering called , which is another offering, and so on. These basically get a lot of people into the habit of flagging incoming disinformation as spam.
The professional fact-checkers, such as the Taiwan FactCheck Center, MyGoPen – they’re all part of the International Fact-Checking Center, can act on it early, that is to say, when it’s still being end-to-end group-tested, instead of going to the truly viral social media platforms.
Our response, which is always on the social media itself, sometimes even precedes the virality of the disinformation when it comes to the public social media such as Instagram, Twitter, or Facebook.
That’s very interesting. What about the older generation? You say it’s a real collective effort, and I imagine there is kind of an age limit within that effort. Is there an older generation who are…
The elderly are very active on the LINE platform as well – I would argue even more active [laughs] than the younger people who are much more likely to use Instagram.
The elderly also want to help in correcting the disinformation. It’s just that for them, you need to make sure that those lifelong education materials are in their language, and also in a way that is understandable by them. They are a wise bunch of people, and they use LINE a lot.
We also make sure that, for example, on “LINE Today,” which is the news feed on the LINE system, the LINE platform gives a prominent section to the fact-checking, the real-time fact-checking.
I think they gave it the second-most column, so that it’s top news. Then entertainment and gossiping, and then fact-checking. It’s very likely that the elderly people will also see the fact-checking before they scroll down. Maybe that’s because we’re not finding enough. If we were finding enough, we would overtake entertainment and gossip and to be the top section.
Wow, brilliant. Taking it back to your response to SARS in 2003, I spoke to someone yesterday who was telling me how there was quite a lot of legislation made between 2003 and now.
That foundation has been key for you for things like privacy protections and things like that. Can you tell me a bit more…
It’s a societal inoculation. Because in SARS, the municipal government was saying a very [laughs] different thing from the central government.
We see that, actually, around the world when people get exposed to SARS 2.0 for the first time – that actually much of the source of not disinformation, but it was not intentional, but rather misinformation, because people get misinformed. That’s a real problem.
Basically, what we did is that the central epidemic command center, or the CECC, has the power to interpret the science in a way that all the municipal governments as well as other ministries are subject to the CECC interpretations. There is a very strong like just five people maybe every day, every 2:00 PM, just answering all the questions from all the journalists and also the toll-free number 1922. That’s the single source of truth.
Of course, like any science, epidemiology meet amid new evidences, so it’s not like it’s top-down and they’re never wrong. It’s just like when they’re wrong, they admit it very quickly, [laughs] and in a very accountable and competent fashion, and correct it on the very next day’s livestream conference. That builds clarity.
There’s a lot of confusion around the rest of the globe, isn’t there, in terms of what’s true and what’s not, and they’ve got so much disinformation. That seems to actually be the key to the problem. You guys have got a lot of data, haven’t you, and you’re using this data obviously to try and improve your…
But we don’t collect new data.
You don’t collect any new data at all?
We do not collect data that we were not already collecting before the pandemic.
Really? This is all put in place because of the legislation that was brought in over the years as well?
Exactly. Also, another reason is that we never declare a state of emergency. According to our constitution, which is like a German-ish constitution, if the parliament doesn’t approve anything, then we can’t do anything. It’s a pre-approval style of constitution.
Whereas in many jurisdictions they declare a state of emergency, where the administration gets to do pretty much anything, and then the legislation post-approves it in a post hoc fashion, in Taiwan we never do that.
Because of that, everything we do must be subject to the parliamentary interpelletion. Because of that, we don’t really want to collect new data, because any new data collection would go through this very lengthy process of parliamentary debate, so basically we reuse whatever we have.
What kind of data are you collecting that you’re using for things like contact tracing, for example? How do you collect the patient data and that sort of thing?
For example, for the digital quarantine, which is the most important piece, we make sure that when you return to Taiwan, you either go to a quarantine hotel where you’re physically barred from leaving for two weeks.
Or if you live in a flat with your own bathroom, then you can do a home quarantine, in which case we very clearly tell you that your phone, or if you don’t have a phone, then we give you a phone for two weeks, will be put into the digital fence.
The digital fence is that very simple idea that the telecom already know through triangulation roughly where the phone is, because they use it for roaming service. It’s the telecom providers automatically sending SMS if the people go outside of the quarantine zone.
Because it’s not GPS, Bluetooth, or WiFi, the resolution is very coarse. Even in the most urban areas, it may be 15-meter radius. Because of that, we don’t know which room you are in, but we do know which district you are in.
Because of that, the telecoms do not need to hunt that data to third party processors. They process it entirely within their own data center, and the SMS is sent in a format that very much resembles the advanced earthquake warnings and flood evacuation warnings or a forest fire or whatever that people are already accustomed to.
They understand that when we send the SMS to the people in quarantine, and also to the health officers, that SMS has no way to, for example, track their email [laughs] or read their WhatsApp, because it’s not an app.
Because of that, even if they use a flip phone, a feature phone, that still works. That is the way to use the telecom signals transdata in a way that’s constitutionally protected.
Also, after the 14-day quarantine, of course, the log is rotated. We cannot keep the data, and the telecoms never hand it to the third party anyway, so the third party cannot sell it to advertisers or something.
The telco data and the cellphone network that you’re using for actually pushing out these messages, is it just a regular cellphone network, it’s not up for special emergency?
Right. It’s just triangulation.
That’s amazing. If people were to break the quarantine, what would happen to them?
If they stay in the quarantine, then we pay them about 30 Euros a day as a stipend for collaboration with the quarantine. If they break the quarantine, then they pay us back up to 1,000 time that, so they can fund 1,000 people [laughs] for a day.
The point here is that the fine is so heavy that people just wouldn’t break the quarantine. That is why quarantining is the first and foremost thing, because quarantine is physical, just like this mask is like a physical vaccine. The boundary is very clear unlike, say, a Bluetooth Bangor or something.
It, of course, may do some contact tracing. Because it’s new technology invented after the pandemic, had we rolled it out, then the privacy budget, the privacy implications, the service security implications will be much less trusted because it’s new technology. Everybody understand this physical quarantine idea.
That makes a lot of sense. Do you have a contact tracing app at all?
No, no, no.
When someone comes into the country, I was told that they were needed to download an app when they come in.
No, no, no, no. They would either file a paper form, or they can scan a QR code and file an online form, but it’s not an app.
Is it that they will be tracked once they’ve filled in the form, so their cellphone data will be tracked?
Yeah. They just say, “Oh, this is my cellphone number.”
And it’s done over the network, but on randomized times of…
That’s right.
That’s fantastic, how interesting. Those people, I was told if they had their phone off, that they would have a health inspector turn up to their house or the police…Is it that far?
If their phone runs out of battery, then a health officer knocks their door I think within minutes.
Fantastic. People in general, from the sounds of it, this obviously involves a lot of collaboration. Does it involve a certain type of a collective personality, a form of altruism rather than…
No, I don’t think so. It’s entirely rational self-interest. This is akin to how we build the mask use. We know that wearing a mask alone doesn’t work, you have to wash your hands, too.
We say, “You wear a mask to protect yourself from your own unwashed hands.” This links hand sanitation and mask use together. This is appealing to one’s own self-interest.
If we had say, “You wear a mask to protect the elderly, if you wear a mask to respect each other,” that’s a collectivist argument, and that actually wouldn’t work.
(laughter)
It wouldn’t work. [laughs] We know as a fact it wouldn’t work, so we always design to appeal to their rational self-interest.
The same thing goes on for a quarantine. They abide the quarantining, because we’ve sent them a lot of food, this huge carrying package, a video conference with people who are professional counselors and so on, and a lot of gratitude for their collaboration and so on, elevate social status.
After the quarantine there’s no shaming possible, we all have gone through it. Because of this, it makes far more sense in a self-interest way to abide the 14-day quarantine. Otherwise, it’s a huge amount of fine and social shaming, too.
With the quarantine, is it a two-week quarantine for people who have come into contact with someone, and then the isolation is the same amount of time?
Yeah, it’s the same amount of time, and it’s the same arrangement, except when it comes to isolation. Then, it’s not the medical officers or health officers knocking your door, it will be a police officer knocking your door, but otherwise it’s the same. [laughs]
You’re saying that you pay people €30 to stay at home. What about people who needed to go out to work? Would they be earning a lot more than that normally? I mean…
Yeah, yeah, €30 a day for their stipend, and of course, if they can telework, they of course would telework. We spend a lot of time making sure that those quarantine places in the hotels have good bandwidth so that they can do videoconferencing and so on.
Even in the most remote places, we arrange to have at least one of the telecoms, [laughs] the CHT to arrange at least one of those telecoms will have good 4G network. For people who cannot telework in their profession, of course, there’s a act, a law that says that of course their employers cannot force them for it and things like that. It’s all very clearly spelled in the law.
That was all in place before this happened?
Do you think that if this would happen on a larger scale – which it sounds unlikely – that you would still be able to manage this system? For example, that’s a lot of people in hotels, in the isolation hotels. How would you manage that if it was such large-scale?
We’ll just work with more hotels. It’s not like we have a lot of inbound tourism, anyway, [laughs] so there’s a lot of capacity in the hotels. Those hotels are happy to also be quarantine hotel, because it guarantee them a steady influx of people, and it’s very fairly priced.
For example, the average quarantine hotel is about €130 per day, but the person who stay in it only have to pay one-third of that, which is just €44 a day. Then, the other two-thirds it stays subsidized.
And they get the €30 as well from the government to be able to almost pay for that…?
What about the border side of things? Is there a kind of technology at the borders? You shut your borders very early on, didn’t you?
Yeah. In the borders there’s two things going on. One is, of course, the mandatory health inspection which, as you mentioned, started as early as January.
Then after, for example, when the US and Europe begin to show the wave, we also retroactively tested the people who came in for the previous seven days, and found actually quite a few positive cases. Some of them are also retroactive.
With your testing system, is that a very active part of the system, because a lot of people outside of Taiwan are placing a lot of emphasis on the testing side of things. Do you think that’s important also?
Because we never had community spread, so that’s not a semantically relevant question. [laughs] If you don’t have a community spread, and you have confidence that your R-value is under 1, then basically it’s just quarantine, isolation, and contact tracing.
Of course, you test all the people the contact tracing identified. If people need to travel overseas, of course, they can pay for their own testing. Otherwise, there’s no need to just randomly test people.
That makes sense. What about the data side of things? You’ve talked about the cellphone data. What other data are you collecting and how you’re using it?
For example, we already have in the form of a IC card people’s national ID, and whether they are eligible for the national healthcare, which 99.99 percent of not just citizens but also residents have coverage.
That IC card formed the basis of the mask rationing system, which we designed to initially add 6,000 pharmacies places and resembles the chronic prescription, like many elderly go to the pharmacist every other week or every month for refilling their chronic prescriptions. It’s the same experience.
It just so happens that last year we equipped around 90 percent of our pharmacies who are connected to the national health service with very high speed fiber optic connection. That means that even if you have a long queue, you just swipe the NHI card, each taking just a few seconds, and then you can just hand out those medical masks quite quickly.
Afterwards, we also enlisted those convenience store, because it turned out if people work very long hours, once they go off work, all the pharmacies have closed anyway, [laughs] so it doesn’t matter how quick the fiber optic is, you still can’t get a mask.
The convenience stores are 24 hours a day, and so we implement the system of mobile preorder by March, and are still on the national health insurance app. Then, also the convenience store preorder where you can take your IC card, but still the national health insurance are on early April, and by late April everyone know one of those three venues to get masks.
Even for people who have plenty of mask stored, and therefore doesn’t need the ration mask, we even design a function in the app out of popular demand to dedicate one’s own quota to the international humanitarian aid. [laughs] Many people dedicated their masks.
The latest count as of yesterday is more than 700,000 citizens and almost seven million medical masks dedicated this way, and they can choose to show their name. [laughs] You can actually see those names who donate their mask to international friends. All of this is built on the national health insurance system.
Is there a name of the app that you’re using?
It’s called 健保快易通, but really it’s just the national health insurance app.
You had those cards in place for quite a long time, didn’t you? They’ve only started being used.
The card gets tested in the Pescadores Islands, the Penghu Island, in 2003, which is when SARS 1.0 came, and it showed everyone that the IC card really helps keeping track of the epidemic and the clinics and so on for that island.
Right after SARS 1.0, the entirety of the Taiwan main island and many other islands, all adopted the IC card. It’s been around since at least 2004.
It sounds like from listening to this and someone else in the world, where we’re not doing arguably very well [laughs] – I don’t think you can argue with that, we’re not doing very well. It seems like the biggest part of this is because you had a foundation in place.
You didn’t have to introduce any legislation. You were prepared because of SARS. Is there anything that’s been added retrospectively? Is there anything that the rest of the world could do, any advice that you could give us what we should be doing?
Yeah. First of all, appeal to rational self-interest. We have found that our messages that appeals to altruism doesn’t work.
(laughter)
It’s true. It’s sad but true. [laughs]
It’s human nature.
(laughter)
Again, wear a mask to protect yourself from your own hands. That works really well. Queues, spokesdogs really works. The trust from the government to the citizens so that anyone can call a hotline, and essentially ask a scientist or a proxy of a scientist and through a chat box and so on. That really works.
A very clear explanation of all the measures and the scientific “why,” not just the administrative “how” really works, because that enable people to innovate.
There was a case, and I think it’s a very illustrative case, where a nightlife worker, a intimate drinking bar worker, was diagnosed with COVID-19. Initially, at the first day, she didn’t tell anything to the contact tracers, says, “I’ll just remain home. I didn’t meet friends.”
The second day, of course, she said, “OK. I work in a intimate bar,” and because of professional requirement of anonymity, she didn’t want to work with the contact tracer. For many jurisdictions, at that point, there would be a very strong pressure for the people who work in nightlife to be fined or even put to prison for not collaborating with contact tracing.
That will lead to stigma, and that leads to business going underground, as in the US Prohibition era, and that will actually just make the pandemic even more unpredictable. The CECC had a lot of experts, and some of them had extensive experience working with HIV-positive communities.
Just like in the pharmacies distributions network, we always trust the pharmacy because the neighborhood trusts them, and so we adjust our system to work with them. In the quarantine system, we trust the people in the quarantine to not break the quarantine.
In this case, we also trusted the nightlife businesses, and we said, “OK. We don’t need the real name, because we understand you cannot give the real name of your patrons, but you need to make a real contact system in the case of a transmission.”
If people in the past two weeks could be effectively contacted, then no real name was necessary, and we also explained how the physical distance requirement worked in scientific principles. They developed very creative processes, because we didn’t fine anyone or put anyone in jail when that case appeared early April.
By early May they invented code names, single-use emails, prepaid SIM cards, scratch pads, so they didn’t need to send any data to the central government. They didn’t keep it in their business. Even hat with plastic shield that maintains the distancing, so you can see each other and drink, but probably no kissing. [laughs]
All these things are social innovations arguably, and then, the municipal government actually allowed those drinking bars and nightclubs to then open. When even the nightlife workers can join the fight, then the prevention efforts would garner more social trustworthiness. That’s just one anecdote, but it illustrated the entire point.
That’s fantastic, but it sounds like as a country, you guys are very digitally innovative. Other places in the rest of the world might not have the scale, they might not have that drive.
I don’t think so. I think it’s the collective intelligence, because I didn’t invent the mask availability map. We certainly didn’t invent the use of traditional rice cookers to cook the mask, [laughs] to cure the virus but not destroy the fabric.
All these are citizen inventions, and they call 1922 knowing that their new idea will be amplified to the entire society on next day’s livestream conference. I don’t think it’s because Taiwanese people are more innovative than other people. It’s because in other jurisdictions, their central command center, if there is one, probably did not crowdsource the inventions from their citizens.
What about your relationship with WHO as well, because you don’t have a relationship with WHO? How does that…
We do have some limited scientific access, but we did not have ministerial access. That’s true. In Taiwan, of course, these two are almost the same. When the top epidemiologist want to talk to the vice president, he looks into the mirror. It’s the same person.
(laughter)
In other countries, it’s not the same thing. Even if we have the access to their epidemiologists, it doesn’t mean that we can translate our policies into actionable policies for their ministry. Of course, more ministerial access would really help.
Going forward, what are your plans to maintain the fact that Taiwan has done very well? You have very little deaths. You’ve limited the number of cases. Is there a forward plan or…?
Sure. Life has to return to normal. What we are now doing is, first, we’re planning to distribute the vaccines, so the logistics of the vaccines, each having very different temperature requirements and things like that.
That probably has to build upon the national health insurance system too because the same person need to go for two shots for some sort of vaccines. They can’t confuse it with another vaccine. Otherwise, it doesn’t work. That’s a complex logistics system. We’re blessed with really good pharmacists and clinicians, and the national health insurance system. We’re working on that.
Also, we’re working to make sure that the quarantining process could be a lot more mobile. For some cases, for example, for short business visits, they just can’t do 14 days. If the digital quarantine can move with the person, and they don’t take public transport, theoretically, we can change the system so that the bubble moves with that person in quarantine.
It’s being implemented for the invitation to the Czech Republic Senate, the head of senate visit to our legislature proclaiming the 我是臺灣人, I am a Taiwanese, and so for the Czech case and for the secretary of health in the US, from the Azar case, and so on. For many of those visits, we make sure that there’s traveling bubble possible. We’re working to refine that system too.
How close are you to having a vaccine now?
Everybody says it’s next year. They never say which quarter next year. [laughs] I think that by Q1 we will have a pretty good idea of how the distribution mechanism would work. The best case is by the end of Q1 we have a vaccination. Our domestic vaccination research companies are saying something around Q2 to Q3.
We have a fallback. If it turns out that those hypertechnologically advanced, low-temperature choices somehow has a logistic issue, then we just delay the vaccination by a quarter. Still, we have a plan B.
In terms of the data that you have, because the data was all planned to be collected, obviously, before this happened, is there any additional data that you wish you did have, or that would be useful to have, or that you are looking at collecting?
Yeah. If this has happened again, I think we will make sure that the convenience stores, like Watsons and so on, are treated as pharmacists in the first place because it turns out they do have pretty good Internet connections. Their cybersecurity perimeters are pretty good. In the case of Watsons, Cosmed, and others shops like that, they do have working pharmacists in place.
Where we started only with the NHI-contracted community pharmacists, if I had a choice to revisit the early February decision, then I would also enlist all the places that has pharmacists to the distribution mechanism. Also, convenience stores could be brought on board much earlier.
There’s not a huge amount of stuff that’s not happened as a result of the data not being there, is there? Everything was already planned.
Fantastic.
We run yearly drills for this particular respiratory disease, for SARS 1.0.
You’re very well prepared. Excellent. We’ve covered everything that I had. Was there anything else to add at all? Do you think there’s anything we should know, anything we can learn from that we haven’t spoken about?
I think it’s a great opportunity for the world to look at this situation in terms of data and learn that, from Taiwan’s perspective, there’s no need to either go to this authoritarian intelligence lockdown thing or to succumb to surveillance capitalism in order to get all the data.
Instead of the two false dilemma, what we do is we trust the social sector with the innovation capabilities. It’s like the Pygmalion effect. If the government think the social sector is trustworthy and innovative, they become trustworthy and innovative.
When the social sector holds the data, as in the case of the intimate drinking bar, then it scales much better. People don’t have to make the false choice between freedom and human right on one side and public health on the other.
That makes absolute sense. Thank you, Audrey. I really appreciate you talking to me today. That was fantastic. Lots of information to digest. I’ll write the article up. I’ll send the link when it’s up.
Awesome.
Thanks again. Good luck with everything.
Live long and prosper.
And you. Take care. [laughs]