Thanks for joining, talk to you again soon.

Full transcripts and an extensive collection of links to learn more are available on our site and put together by Katy Moore.

Audio mastering and technical editing by Ben Cordell.

Alright, the 80,000 Hours Podcast is produced and edited by Keiran Harris.

Just a reminder that we’re looking to hire advisors to counsel people one on one about how to have more impact with their work and a new Job Board lead to take it to the next level.

If you’d like to hear more from Audrey I can recommend her interview on the podcast Conversations with Tyler, from back in October 2020. That’s actually what inspired me to invite her on the show. There’s a link to that in the blog post associated with the episode.

Yeah, I look forward to hearing the show. We’ll let listeners know when it launches. My guest today has been Audrey Tang. Thanks so much for coming on The 80,000 Hours Podcast, Audrey.

Oh, thanks so much.

Fingers crossed we’ll have a technical solution to all of these moral quandaries soon. Well, I super appreciate all of the work that you’re doing and I’m really excited to learn more about some of these experiments, and our coordination between people and aggregating preferences and learning and beliefs and ...

They’re also really healthy, but it seems like the complexity of the signaling system within mussels isn’t that much more complicated than what you get in plants. So there’s a slight, somewhat equivalence argument that if you’re willing to eat plants, then maybe eating mussels isn’t so much worse than ...

Yeah. I’m largely vegan, but I also eat mussels. I haven’t really thought that much about oysters, because I don’t super like them, but yeah, I’m a kindred spirit in this, it seems like.

And I guess the reason to eat oysters, but not other meat, is that oysters probably aren’t conscious? Or if they are, very minimally so?

All right. We’ve gone well over the 90 minutes that you originally agreed to speak with us. You’ve been super generous, but we should wrap up and let you get back to your work. One final question I have is that I heard on one of your previous podcasts that ...

I see. So is that typical? Basically, that things just get constantly forked when people want to change them, rather than necessarily having to have some political process that decides what is in the official or the most popular version?

OK. So that’s one way of handling disagreement. I imagine that there might end up being one map that more people use than others, that’s kind of the default one. Or not really?

No? OK. Yeah. Explain that.

With the g0v projects, how does the internal politics work? So say you’ve got something like the mask mapping project. Presumably not just anyone can just go on there and change the code. There’s presumably some administrator, and then other people might have to get approval in order to change ...

It sounds like in Taiwan there’s a lot of citizens and volunteers and programmers who are effectively contributing to provide these quasi-government services because they’re really motivated to do so. If you took those people and instead hired them to be actual public servants and provide services like these as ...

Yeah. Yeah. That makes sense.

I guess I’m trying to figure out if there’s some kind of limiting principle to how far g0v can go. Are there any cases where people tried to do a shadow government service or website, and it turned out to be harder than they thought, and it was actually better ...

Can you think of any cases where you’ve had to give up on a project because it just wasn’t going to work? Or maybe it was working or it was up, but it was net harmful, and so you decided to scrap it?

There’d be all of these objections that people would raise, plus just general anxiety that people would feel if they’re not used to this system. Have you had to overcome those barriers? Or do ideas for open data or open design sometimes get shut down because of these worries?

So just imagining a hypothetical example in the US or the UK. Let’s say the people didn’t like the website and the forms available with the Department of Motor Vehicles. So you’re getting licenses, booking appointments for tests, and things like that. I imagine if you started creating an alternative ...

Yeah. How far can you take this? Is this true across many or most Taiwanese government services? Or is it only a minority where that’s feasible?

A sort of soft fork. This would be very hard, I would imagine, to do in the UK or the US because people wouldn’t have access. They could make, for example, an FAQ website about the DMV on that domain. But they wouldn’t have access to the APIs, they wouldn’t ...

— .g0v, and then builds kind of an alternative to the DMV — just themselves or with some friends. And the hope there would be that they’ll effectively provide a better service that then would just get wrapped into the normal government website once it’s been demonstrated that people prefer ...

So just to explain g0v to the audience. Basically, the idea here is to build kind of a shadow set of government websites and services. So you can imagine if people don’t like the Department of Motor Vehicles, dmv.gov, someone registers dmv.gov –

Before 2014, was the Taiwanese government perhaps unusually closed, or unusually backwards? To some extent, has this been a real swing of the pendulum in the opposite direction?

Right. To some extent, you are the representative in the government for this community, or this platform. Do you have to do any politicking in order to maintain support for it? Because people presumably have some criticisms. They might think sometimes it’s a bad way to spend money or a ...

Just so I can understand better how this community works. It sounds like there’s one part of it, which is citizens wanting to present ideas and concerns to their government. And there’s another aspect of it, which is people who want to volunteer their time to civically minded projects — ...

I imagine that a lot of people would have reservations about delegating so much control over what they spend their time on and how things get implemented to an open community that people can easily join. I guess all of us end up delegating a lot of decision-making power to ...

Right, right. So if the community strongly upvotes some proposal, what’s the way that you are most likely to be able to help it get implemented? Is it coordinating with people in government, inasmuch as that’s necessary? Or providing technical advice?

Interesting. So basically, you allow the community to decide how you’re going to allocate your time to some degree. Does that create any problems where people might vote for something that you don’t personally feel that motivated by?

People must bring you tons of ideas, and you would have to decide to work on just a tiny minority of those.

Then you also have the reward that potentially, if things go well, their idea might then get scaled up to a much larger scale. So the prize in terms of social impact is larger than it would be if that were not available. Is that the majority of what you’re ...

Then you’re curating their ideas, and getting other people to basically vote on how good they think they are, in some decentralized way. And by broadcasting their ideas, you’re also potentially attracting programmers or people who want to work on those projects specifically. By publishing your conversations, they could figure ...

It sounds like you’re providing a few things. I suppose one is that you’re meeting with people, and presumably giving them advice on the strengths and weaknesses of the things that they’re thinking of working on. As a minister in the government, talking with you maybe motivates people and provides ...

Talking about how you get things done in a previous interview, you said that you don’t do much of the actual programming or implementation or deployment of projects anymore. Instead, you just design the spaces for people who care about a topic to make things happen. Can you elaborate on ...

The approach of humor over rumor, just for listeners, is rather than try to get Facebook to change its policies in order to combat disinformation, instead you have a large team of volunteers who are constantly alerted to disinformation that is circulating — ideally is newly circulating. And they come ...

You seem to have accomplished a lot over the last decade, and I get the sense that the way that you work on a day-to-day basis or month-to-month basis is a little bit different than other people. So I’m curious to learn more about it. What’s a project you’ve contributed ...

I see.

But I guess that’s kind of the world as it is now. You’re potentially right that if we had extremely good translation AI systems that can translate between all of these different, smaller languages as well as we can translate between Spanish and English today — which is really quite ...

Yeah. In the world in which we live at the moment, it seems like there’s far more people flowing out of these smaller Indigenous languages than are flowing into them. So on net, there’s a lot of language disappearance. And I get one reason is that those languages tend to ...

And so just through voluntary choice, voluntary association, these languages gradually become less useful to speak, and then more people peel off and stop speaking them. Then gradually they disappear, excepting kind of books by linguists. Would you want to do anything about that, or is that just acceptable?

In listening to some of your previous interviews, you were saying that an example of your conservative or conservationist instinct is that you would like to see Taiwan’s numerous smaller Indigenous languages preserved and ensure that people can continue to speak them, even though Mandarin is probably quite dominant. A ...

Yeah.

Yeah.

Right. But if they did it in the real world, then we could no longer use taxation as much to help poor people, say. That would be a very common objection that you might get from many people.

So in the case where people can move community completely, an immediate objection that some people would have would be, say, what if all of the billionaires went and formed their own community? And they set a minimal tax rate — 1%, 0% — because they didn’t need —