Thank you for the great testing report for the Mountain website. Some of our native-English-speaker friends said that the English version may not even be worth testing.

We did our best.

It’s good, because we are remaking it as a new website. All your better translations, all your feedback of searching for the snow mountain and finding nothing, and things like that, that will become our stack with the new designers. That’s very valuable input, and at the right time.

Great, and if you’d like us to come and talk about the results that we had…Obviously, I know you’re doing a different website, but if you wanted to do a review of the current one with the engineers, we’d love to have a session. We can use Chinese as well if that’s…

Unfortunately, after their delivery of the work and the premier’s political will to improve the experience, we’ve switched vendors. [laughs] We also have three people – previously PDIS interns – working on service design. They use very similar methods to test the Mandarin one.

Again, after seeing the two reviews, we decided to change vendors, but now they’re included in the new design process for a new website. We’ll probably leave the old one alone, and we may, actually, if you click “English,” just go to the new website. That’s one of the possibilities.

Oh. Everyone should be like, “Here is a better one.”

The old one is marginally useful for a Chinese speaker, but not useful at all for English speakers. Thank you for including that in writing and improving it. I’ll just share it with the new vendor if it’s OK with you.

Yeah, sure.

They’ll be in contact with you if they need advice on how to make it more native-language-sounding.

We’d love to do it. We’d love to test for free. We’d love to help.

Thank you!

OK?

Uh-huh.

I guess let’s get started. I think we’ll go back to that in a second – or maybe 20 minutes or so.

No problem.

I guess you probably already know Darice, but Darice is…

Yes.

…a member of the Alliance for a Globally Oriented Taiwan. She brings a lot of skills to the table – translation, media work, social media, journalism. I’m not…No, no, I appreciate…

You set up the meeting and figured everything out.

I’m a good bureaucrat.

Yep, he is. He’s very independent.

I’m a terrible bureaucrat. If it were me, I would just be like, “I don’t know what’s going on.”

At AGOT meetings we’ve been talking about what kind of things would be useful for foreigners in Taiwan who want to help to do. We’ve created different committees to talk about all these different things.

Obviously, not everybody can be in the room at the same time, so is there a way we could crowdsource this, some sort of suggestions?

Asynchronously?

Right.

Yeah. I’ve actually brought up Pol.is with AGOT a couple times. They went and researched it, but it seems a little technically difficult.

To set up one, or to…?

To set up, yeah.

OK. We’re fixing that. There will be actually a training workshop-ish thing with the Talent Circulation Alliance.

I think you’re going to it, right?

Yeah, I’m going to that.

We’re massively simplifying the setup, is what I’m saying.

It’ll be available. For example, if we’re AGOT, we’re not an NGO or anything. We’re just a group of people who are like, “Hey, we would like to help Taiwan.”

You can just start a conversation anytime.

Oh, OK, cool.

That’s the idea.

The question is will we need to be setting up our own server to do?

Not at all. You can just use our PDIS server. Our PDIS server used to be the development server. The ait-pol.is is the production server. I understand that not all social sector activities want to associate themselves with the AIT.

That was the main branding issue, which is why I register a new domain name called talkstotaiwan.tw. Now, you could just be tca.talkstotaiwan.tw, heot.talkstotaiwan.tw, psychologically distancing from the AIT brand, although it is the same machine, by the way.

Oh, it’s different branding.

It’s the same physical machine, but very…Also, when we build the domain name as talkto.ait.org.tw, someone, a blogger named TH Schee, made a very good observation that talkto.ait.org.tw seems like AIT is hard to approach, and people are trying to approach it.

Now, tca.talksto.tw means that it’s, you’re taking the initiative, and you’re a peer-to-peer relationship, and things like that. It is a different, but I think it’s a better branding.

Also, we had a couple questions about the Pol.is results.

You’re looking at the current one, which I’m going to promote massively on Monday in a Facebook post and a Twitter post. Retweets welcome. [laughs]

I’ll definitely retweet.

Usually we see two waves of participation. It’s just like filing income tax. The first week and the last week.

It’s kind of a tranquil period in between. [laughs] We’ll definitely push the last week’s participation through a printed column at “Apple Daily,” and also online social media campaigns. I think that people will vote much more than that.

If a person did not vote for more than seven votes, they don’t get grouped. There’s insufficient information. That also serves as a way to call people back to vote again.

Is that information available to the government when they start voting? If they’re like, “You need to vote seven times to be counted”?

It could be written on the description. It’s up to you whether you want to reveal it or not to your voters.

It’s a pretty interesting, those statistics, because sometimes it’s like I think it’s actually really easy to vote. How is it…?

Yes/no/yes/no/yes/no.

It seems like the average amount so far has been three or four votes.

Yeah, because also, people want to get their voices in. There’s kind of a competition of their attention. One is after voting three times, they think, “Ah, that’s beside the point. I just want to say this.”

Write their own thing.

After they write this, they may not actually go back to voting. There’s a different attention management thing going on.

There’s a lot of active experiment. Do we show the most divisive questions first? Would that increase the voting before commenting? Or do we show the most consensual one? Or do we just show a random one? Currently, it’s showing a random one.

You guys are going to put a more…

There will be a training course and everything.

For the AIT one, have you guys thought about doing ads in “Taipei Times” or anything like that?

You mean for the security cooperation?

Yeah.

It’s a interesting topic, for sure.

Because a lot of Americans or English-reading people, they do still read Taipei Times. That might be a…

Not necessarily Apple Daily.

Yeah, not…yeah. [laughs]

I’m interested, but also, we only have 12 days to go.

Oh, that’s true.

Maybe we can start with actually the talent circulation one, because first, it’s less de facto embassy-sounding. The binding power is not…Somebody from MOFA will talk about this in our next year’s Indo-Pacific consultation agenda, which means a lot for the academia and the organizers, but less so for the “Liberty Times” readers.

Maybe let’s start with the TCA collaboration.

That’s true. There’s going to be trainings. Oh, then one of my last questions about this one is, how do we know that the results are robust to the population that you’re trying to reach with this?

Robust in two senses. First is whether it allows for independent analysis. When you set up a conversation, you can just click “Open Data.” If you click that, anyone can export the entire history – of course, not linked to the original social media login, but otherwise, entirely open for independent analysis. We have toolkits to analyze that.

You don’t have trust. You don’t have to trust the report interface. You can run your own analysis…

Oh, OK, you can independently…

…by yourself.

If you look at my GitHub, there’s a repository called Polis Tally, T-A-L-L-Y, where people did run independent analysis. The first time the Polis is using the public sector in Taiwan is 2015, when the UberX case is run. Uber ran their own independent analysis based on this data, and drawing pretty different conclusions, actually, from the same base data.

That’s the value of robustness, because it could be inspected in multiple dimensions. That’s the first one. The second one is whether it’s been hyper security hijacked, bought, manipulated, and things like that. Fortunately, we’re running in the HiNet government cloud. They’re the same system that hosts our online tax filing, so cyber security, less of an issue. [laughs]

That’s always good.

Right. You can always add CAPTCHAs and whatever.

OK, cool. I think that was mostly it for Pol.is. We’ll just wait for you to go to the training and then ask you all the questions. [laughs]

I guess a question is, the training to me, it sounded more technical than anything. Is it also…?

It’s also moderator, also how to set up a conversation, as well.

I’m happy to bring back that information to you. If you guys want to invite Darice, and she has time, I think that’s great, too.

Also, if you have a better idea of a landing page, I can just point the domain to you.

I wish talkto.ait.org.tw would go to their website, because they have that website. I think it is on their website that…

After the four conversations, it will go back to that website. They have sequential conversations. talkto.ait.org.tw points to the current one. talksto.tw doesn’t have to follow the same name. You get to the decide the norm, and I own that domain. I don’t own ait.org.tw.

It would be bad for me to own it.

Conflict of interest.

A little bit, yeah. My thinking is something like, you could have tca.talksto and maybe like – maybe that’s too much – current.tca.talksto.

No, it’s all yours. Just let me know which CNAME to point to. [laughs]

Let me think about it, but I do like how they keep track on their website of how things are going.

Or maybe talksto.tw is the list of things ongoing, and tca.talksto is the ongoing conversation, maybe. I don’t know. You let me know, and we’ll make it happen.

OK, great. Thank you. We’ll go ahead, and we’ll see how this goes on Monday. Then John will be coming, all hands guy. He’ll be helping with more of content sourcing or content ideas. It should be good.

Right, but most important thing is always the instructions you give, like how binding this is. Also, the initial seed questions, because they’re going to inform the comments. After you vote on three of them, you’re going to write something that sounds a little bit like those three. That’s the two most important one.

OK, right. We could do the training, and then we can see those conversations if we have our own, if there’s an AGOT. I guess we’d have to talk with David and other people about how will we see, seeing that information.

The question is, AIT has a very interesting system where things are being auto-translated, I believe, right?

You can come in either language. It doesn’t matter. The Canadian government developed that part, as required by their federal law. Then we adopted it and then also show the majority consensus bilingual as well. They have to be English-French all the time.

Nice, yeah. I wish we could do that in Taiwan.

I know. You mean English-French?

No.

Yeah, sure.

…of that posed.

On the Join platform, we also use the same landing page technology. If you go to pdis.tw, pdis.tw/hike, you will see a very similar landing page to the AIT F40.

Did you say hack or hike?

Hike, H-I-K-E. It’s about five topics, the second of which you just did a feasibility testing about, right? Again, it’s like five simultaneous Pol.is conversations. It has a landing page of reasonably clear, I hope, clarification of binding power.

It runs until end of this month, and it is coincidentally bilingual, but not explicitly bilingual in the Pol.is part. In the Join part, it’s more monolingual.

The Pol.is part?

Yeah, if you click on any of those pictures, then you accidentally, probably, see some English translations, maybe.

Oh, I see. Oh, yeah.

They may or may not make sense.

Activate third party, yeah. The technology is from the Canadian government?

That’s right. They contribute to the Pol.is platform.

Oh, wow. That’s really awesome.

Cool, OK.

In the Join, Join is actually three different systems rolled into one. What you are seeing is the consultation part, and we do use Pol.is here. The e-petition part, completely different. The participatory budget part, completely different, but those are three. It’s easier to remember anyone.

Are they automated in connection, or is it a manual process to move data between those?

No, it’s automated. Whenever any ministry announces a regulatory pre-announcement, like for 60 days, like in regulation.gov, it automagically appears from the gazette through machine-to-machine, to the Join platform.

The same goes to the KPIs filed every tenth, every month, to the administration’s GPM net which automagically appear on the participatory budget platform, while e-petition is user-generated.

Then just to clarify, the point of the Join initiative is to directly talk with the government about different policies?

It is explicitly about every single ministry having the transparent accountability, for lack of a better term, across all the policy cycle. From people’s ideas, to collaborative meetings, to useful regulations and implementations.

Then we spend some budget, you can track how those budget went, and maybe propose some ideas. That’s a life cycle.

OK, cool. It’s like the US Congress petition website, but in Chinese for Taiwan?

That’s part of it, but also, it’s also We the People. The e-petition part is directly modeled after We the People, actually. Except with more binding power.

Participatory budgeting, we learned mostly from Madrid, from Spain. I think all these came from different places. The e-petition, the latest renovation came from Beta Reykjavik, which is Icelandic technology. It’s a large international community. We adapt from the best.

That’s awesome. Cool. Are you guys going to consider having an English version of this, or no?

That’s a great question, isn’t it? I think, if you go to sandbox.org.tw, that’s one way we’re thinking about it. I will welcome your input.

Sandbox.org.tw?

Yeah, sandbox.org.tw. As you can see, this is intrinsically bilingual. Everything is bilingual. Just like, I don’t know, the Canadian prime minister’s Instagram, everything is written twice. This is useful, because firstly, it lets people learn some English. [laughs]

Also, most importantly, when there’s a translational error, at least you know it’s a translational error. If it’s a different website, you don’t even know it’s a translational error.

Yeah, that’s true.

It’s a possibility, but this is new. We’re still weighing this way versus an English portal into the same system, which makes machine translation easier to introduce, but also makes errors harder to spot.

For the hiking website, which is probably going to be our pilot case, you also may actually have maybe actually agenda-setting power in setting whether we develop it in this sandbox way first or a dedicated English site first. What’s going to be useful to you?

I see. For this one, the English part, it’s all machine-translated, or you have actual translators?

No, we have actual translator, but they may be initially machine-translated, in which case it would be clearly labeled as such.

Oh, where is it clearly labeled?

I think they’ve completely the hand-translation now, so you don’t see any automated translation. One good example is Pol.is. You see third party translation. That’s a clearly, something like that.

Then you know it’s machine-translated.

Right. For example, in, I don’t know, the mounting permit from the police agency, I think, now use Google Translate and explicitly mark it as such, like this is Google Translate.

OK, I see. That one was real confusing.

Super confusing.

I was like, “Woof.” I think, actually, it might be nice to have option of both if you’re doing a new project. I feel like, for a lot of foreigners, they don’t know Chinese, so it doesn’t make sense for them to see the Chinese.

For some people that were bilingual, it’s like, I would like to see the original information so that I can double-check if it’s accurate.

That’s one of the things. Sorry for interrupting, but that’s one of the things that I think needs to be changed on a bunch of different websites in Taiwan, where you click English or Chinese, and it takes you to the home page.

It’s aggravating, because they all use the same URL scheme. Why can’t you just replace the URL with the alternate URL in the alternate language? You already know what that scheme…

To be honest, in many cases where it’s non-machine-translated, it takes weeks for the English counterpart to happen. That’s the main administrative challenge, actually. A very good case is the continental affairs council, although it’s not its official name.

The Mainland Affairs Council at yweb.mac.gov.tw, not sponsored by Apple. yweb.mac.gov.tw is actually extremely fast in responding to news, clarifications, and everything. They have very professional editorship.

The whole website is RWD. It’s a good website. If you click English, then you take a time machine and go back to August or even July. I wish that there could be a URL correspondence, as you said, but then that will be broken for a month.

While they get the translation.

That’s right. The two obvious solutions is first, accept that there is machine translation during those three weeks, and clearly mark it as such, or we just put everything bilingual. If you don’t see the English, it means that it’s not yet translated.

I don’t have a design background native to native speakers. If you think we should do both, that will cost a lot more, but we would like to do both as well.

I feel like both would be nicer. Plus, I feel like, if you did have the set box, I feel like if you did have both, you would probably get a lot of volunteer translators just coming out with it before the machine even finishes.

That’s a very good thought, like, “This is missing. I will contribute.” That’s an interesting thought, and then you just have a portal that says, “Take the Chinese characters away, and you are left with an English website.”

You enter a bilingual one, basically, instead of switching to English, you say, “Turn off the ideographic characters.” I don’t know how that will work, but maybe we can do some mockups and do some real testing.

I think that’d be cool, yeah. Speed of information in different languages.

Putting them on the same page seems like the best way to protect against old information staying on the page.

Then, if even I don’t know any Chinese, I can still know that, “Oh, they have the information. I can’t read it right now, but it’s there.” Then I can maybe ask someone else if I really need it right now.

Whereas if I only have English, I won’t even know that this information is available. Otherwise, I have to click through and do a very different…

I think that’s the Canadian mindset. They want people to accidentally learn French, or the other way around, but yes. [laughs]

That’s true.

Also, with the Belt and Road Initiative, I think it does help, you said, cram all the delicious into your web page.

OK.

Does this mean that we’re moving into the next part of this, or did you want to stay on…What do you want to talk about next?

I think that’s pretty much it. Also, I just wanted to get maybe an idea of how much does the Taiwan government actually care about having correct, reasonable, relevant English on their websites? Obviously, this has been an issue for a while.

As a native speaker-editor, I feel like it’s very important, because I feel like the government models what the rest of society should be doing. When they take such a lackadaisical approach to their publications, I feel like that’s not a very good example.

I’m wondering, is the government aware of this? Do they care, or is there anything that we can do about it?

What does the government even mean? We’re just a bunch of people.

It’s true.

Everybody cares differently. If you talk with Joseph Hu, he cares a lot. His Twitter is perfect English. [laughs]

Yeah, his Twitter is wonderful. I love his Twitter, very on-point.

I learn new English words from reading his Twitter. England is doing pretty well also, both on Facebook and Twitter. Sometime, even I learn some Japanese or simplified Chinese.

I feel like that, too.

Obviously, that’s because English is the lingua franca – wrong terminology, but anyway – the lingua anglica of international diplomacy, of foreign affairs, so they care a lot. If you are, for example, the Council of Indigenous Affairs, you should care, because the Austronesian link, but you mainly care about the indigenous languages.

If there’s a competing interest for the budget or for the website, web presence, of course, you’re going to prioritize Amis, Atayal, or something. In this spectrum lies every other ministry.

That’s true.

I think the National Development Council cares a lot. The NDC traditionally is the interface to AmCham, to the European counterpart, and so on, who have all strongly voiced that not only we have a not-long-enough regulatory announcement period before…

In the previous government, it’s like 7 days or 14 days, and the English translation is not even correct. They have to hire another translator to translate. Then 7 days has passed, and what’s there for meaningful participation? [laughs]

The NDC cares a lot, and by nature of that, every ministry that have to interface with the NDC, namely the regulations that concerns foreign people, are now required to be correctly bilingual. That’s something that the commission general cares very much about.

Other than their interface with NDC on one hand, and their interface with MoFA on the other, if I’m a minister, and I have parts of my purview unrelated to these two entities, then the priority is necessarily not the top.

Oh, OK.

This sidelines gracefully into the AGOT platform idea of hiring English liaisons to work in different ministries and be rewarded with gold cards or…

Or possible residency.

Oh, yay. Good idea.

Yeah, you like it, too?

Yes.

We like it, too.

We had that original response, by the way.

With France?

Yeah, France has a fellow in science and technology and also culture, like working as a staff, semi-permanent staff, but then also, serving as a liaison. I don’t know how much the other way around. I’m pretty sure France has that model, with not even francophone, just France-friendly countries.

That’s cool.

It would be great to have something like that here. They could even be staff of the GSA and be funneled out to different organizations, training in the same location, and funneled out to different agencies, assigned to different agencies.

How’s AIT thinking about this, because they have to drive this?

Oh, we’re not sure. I don’t think we’ve met with AIT yet.

Like Ryan and friends. Have you talked to the AIT folks about this idea?

This idea of English liaisons, we have not, but I’m not sure…It seems to me that this is more of a Taiwanese government idea, where there would be a special visa, or a special…

Our idea would be sending fellows to the US, or to Japan, for that matter. It’s our talent that become a foreign government’s fellows.

A foreign government?

Ministerial fellows. It will be, like the French fellow is actually coordinated by the French foreign service. If it’s our MoFA, it will be sending our people out. What you are talking about is the other way around.

Yeah, like accepting fellows from people in different countries who are…

Not necessarily from the US. They could come from UK, Canada, or New Zealand, for that matter.

Anywhere.

I don’t think we have that in ANZTEC in the New Zealand agreement. I would have to check with Canada, but I don’t think we have that clause there, either. With AIT, it’s a lot more room, because we are talking about…What was the word in our Digital Dialogue about? Micro-FTAs.

Instead of like with Mexico and Canada, which is a large package, maybe we do some section, then another section.

That’d be great.

One of them may actually include something like that.

I think talent circulation is another buzzword. I think this falls under that purview of encouraging on both sides. America probably could stand to have people in the American government to come here, maybe, and do training like this. Then there’d be more exchange.

I also think that right now, there is a lot of foreigners in Taiwan who taught English, or have. Maybe they’re retired. Maybe they have an APRC, and they’d love to do something to help.

The AIT still have to vet them?

Yeah, sure. Who?

The AIT.

OK.

They have to provide us with a, like, “This is a list of US nationals that we think are not working for PRC.”

[laughs] OK.

No, seriously. This is serious national security matter.

Understood. Then I’ve noted and log kept. OK. That’s always something I can go back to AIT with.

Yeah, you can go back to them.

Did we want to talk about the website, or did you…?

Yeah, we can move on and do the website.

OK, I’m going to move on, then. I guess what we wanted to talk about was the progress we did with this website, with the mountain.me website. Obviously, you can see, we’re trying to put together an executive summary for the guy in charge, a list of issues or suggestions for UI/UX to the team, and a list of English issues to the team as well.

Also, we’re doing a report card. That’s what we’re calling it, but it’s basically just the same form for every website we want to review. Basically, talk about what are the different things that are weak or strong?

SEO, can we find stuff on Google or other websites to get there? Is the mobile experience OK.

I’ve read through them all.

Wow, OK. That’s all that stuff. Like I said, we’d love if you had any suggestions for how we could do it. Should we put issues in a task queue or something and share the task queue like a Jira, a ClickUp, or something with the team? Something, stuff, things like that, instead of issues, spreadsheets.

I like spreadsheets, obviously. I wrote one. [laughs] I think spreadsheets is the most scalable one, because you don’t need training to use a spreadsheet. I think the spreadsheet with three or four subsheets is definitely the way to go.

It also allows interesting combination with Typeform or Google Form, so that you can crowdsource this much better. It’s just writing more rows into it. If you design specialized software, you have to actually provide training, as we did with Pol.is.

Everybody knows how to use a spreadsheet. Because this is Google spreadsheet we’re talking about, you can even import the HTML from elsewhere. That enables system integration and things like that. I think spreadsheet is good.

Also, because if you require, like director general-level approval, they would have to print it out. If they download it as Excel, they know how to print it. If you use Power BI or whatever, they have no idea how to print it. Again, it’s an easy to visualize advantage.

I agree.

I also is very happy to see that the severe, like “must change now” mistranslations, because even machine translation cannot get this wrong. This must be a human translation, like controlling mountain climbing. It’s not forbidden to climb mountains.

There’s a word salad going on.

401, unauthorized exception, yeah. I’m just trying to think, are we in the right sort of zone, where we would be helping more than anything else? It sounds like we are, so great. Then also, we haven’t done this level of scrutiny for the NAER website.

Yeah.

No, it’s fine. I think the NAER is not obviously bilingual. While mountain nominally has an English button, that does the important part. It claims that it’s English, which is why we can say that, “Yeah, but…”

The Times website isn’t English at all. It doesn’t advertise itself as an English website. That’s, I think, the line we draw. We hire 20 interns to test usability every year. For example, last year’s theme was mobile.

If a website never has a website mobile, or it doesn’t advertise itself as RWD, we don’t even bother testing it. You know it will be horizontally scrolling, and they know about it, too, so what’s the point? If they have a button that says mobile, click into it, and then you start horizontally scrolling… [laughs]

You know, “This is not the way it was supposed to be.”

That’s right. I think the NAER will take a somewhat different structure. It will probably be your ideal English, like pure English or bilingual, interface into NAER. It would be like more of an interface reimagination.

It would be like the exercise we did on the tax filing system. Just ignore how bad the current filing system is. Let’s just ideate on a perfect one, but using the same API, the same data. It’s not like, “I wish this tax filing system can also help me to file passport applications.”

That wouldn’t work, because it’s extra jurisdiction for them. As long as it’s within the NAER purview, like other NAER websites, other NAER resources, it’s all fair game. Actually, if it’s within the Ministry of Education, it’s all fair game.

You can’t say, “I want to see land prices on the term website.” That’s out of scope.

That wouldn’t be appropriate, yeah. I think that’s good. Also, I was wondering if they were going to do anything about the quality of the translations and the notes itself out there.

Yeah, this was this morning’s conversation. We found a very important word of core competency being translated in both different Mandarin ways in both different fields. You’re bilingual, right?

Yeah.

For example, in the social welfare, it’s translated as 核心知能, “To know and to use capability.” In administration, it’s 核心能耐. Psychology, it’s 勝任能力. In traffic, it’s 適任力. In library science, it’s 素養能力.

Now, in education, now, of course, it was 核心素養. Only in information science, which is my field, it’s translated as 核心競爭力, which is a bad translation. In English, that’s “core competitiveness,” which is not the same with core competency.

Because IT people have, I don’t know, we were much louder, I guess. Somehow, the trendy magazines use the 競爭力 as the Mandarin translation of competency since 1994.

Oh, wow.

That informed two generations of students, thinking about core competency as about fighting with Europeans. It has immense cultural negative externality. What?

We finally fixed that. Then we really want to go back to the term NAER and say, “It was a mistranslation. That term doesn’t even belong to information science. I don’t know how they find way there.” [laughs]

That loop, I think, need to be closed. If you imagine a better way than the current forum board, which nobody really reads, I think we’re all ears and eyes.

I don’t know how the workflow is for adding things, too, because I don’t know.

It’s just giant switches. They have the Chinese. They have the English. They have, for example, which the units is, like academic, IT, or whatever, and it’ll explain.

They have a review board powered by professionals, but the NAER actually didn’t develop this system. They just maintain it from, I think, the NDC. At that time, whatever they were called. In any case, they are also very eager to keep the database, but make something, a radically different interface.

They are no longer in control of their current interface. They didn’t commission the current interface.

I don’t know if there’s a way to…I don’t remember. In NAER, was there a way to report a mistranslation?

No, there wasn’t. That was missing.

That would be a great feature.

I feel like reporting or adding suggestions, because a lot of the time, when you translate, for example, across languages, there’s different cultural connotations that you have to be aware of.

Whenever I do translation, and we have a vocab list, I’ll list all of those things, so that other translators who are doing the same document, the same project, they’ll know that, “We translated it this way, because of these considerations.” Then in the future, if things change, then we would consider, for example, other things.

Yeah, because they’re considered normative for the bilingual websites going forward, whether it’s bilingual or whether it’s a new website. What you’re doing now is actually going to have a rippling effect. Every organization in their canonical translation will refer back to term NAER.

[whispers] Oh, no.

The sooner we make it actually, close the feedback loop…Because if we don’t do that, then we may see 競爭力 everywhere.

Everywhere. It’s like, “It’s the only one. That’s not what it means.”

It’s not what it means. That’s right.

That’ll be good. If they ever need someone who’s native English to come in and be like, “These are the better translations,” or, “These are the considerations under review,” it’s here. Yeah, we did go through that as well.

I guess I don’t have as many outside interface, but quality of the database, I’m very, very concerned about it.

They were cutting edge in ‘93.

Yeah.

The society changed.

It’s true, yeah. It’s like, these things need maintenance.

There’s things you can do with volunteers. You could have a volunteer day. You could have people who are in different walks of life or different professions to come in and do a hackathon here, maybe, to say, “Let’s all get in a room, and then…”

I talked to them before becoming the digital minister. I was just a random advocate. At that point, they actually agreed to export their data, for example, to Wiktionary, which is a good place. People can then discuss and collaborate on it.

The problem is that they can’t really figure out to import back, because Wiktionary, everybody can change. Who will do the quality reviews? What about vandalism? What about people injecting political agenda? What about province of China?

Important issues.

That’s something we can help them figure out. It doesn’t have to be Wiktionary, or we can partner with Wiktionary, but with a review process in-between. One way or another, it needs to be collective intelligence.

Another thing I was thinking, actually, is they could probably just proceed. That’s a massive translators website. They have their own CMT databases, which are pretty good. It would be more updated than their current ‘93 version.

Then it would be easier if you got people to come in and be like, “We can work off of that and make it even better.”

That’s right. No, I totally agree. It’s just how to integrate it into the official flow. The best idea is that the official flow, review board members, are themselves are also community participants. Then that makes a lot of sense, because they carry the context back.

The second best, of course, is that they have an import function. “If the social sectors think it’s OK, we’re OK.” The worst case is, as I said, fear, uncertainty, and doubt. “How come we made so many mistakes? We refuse to accept that.”

We really need to shoot for the best case. Maybe it lands to the second best, but let it not fall into the fear, uncertainty, and doubt, which creates a divide between the people. The senior researchers, they’re all very fluent bilingually in ‘93. Society changed. [laughs]

It needs to update. OK, cool.

I guess, yeah, if you have any other websites you’d like to throw at us, that would be great. I know a new one, like an Airbnb clone, just came out. I don’t know…

I know. I think Pol.is itself would be actually what I would like you to improve the user experience.

OK. We could do that next time, Pol.is.

Yeah, because you are going to use that. [laughs]

We should get all of our suggestions in.

It’s important, because we are in charge of sending user experience feedbacks as concrete pull requests back to Pol.is Foundation. We made a lot of changes that the Pol.is Foundation did just merge straight in.

Just full disclosure, I’m flying to New York. If the premiere approves, I’m going to join the board of the Pol.is Foundation. When that happens, there will be a lot more international collaboration to make the user experience not only good for Taiwanese people, that are bilingual, or monolingual in one or some sense.

Also, make it so customizable so that even people in trilingual cases, in indigenous cases, and things like that, can flexibly enable these without incurring burden among monolingual consultations. That’s going to be a very important direction, if I successfully get to join the Pol.is Foundation. It’s called Math and Democracy Foundation, but yeah.

Cool. Fingers crossed for you.

I think it will go well. All right.

All right, cool.

All right, thank you.

Thank you so much.

Thank you so much for having us.

OK. I’ll share your spreadsheets with the…