Episode 1 of Our Human Future with AI, the sixth series of the Global Campus of Human Rights podcast To the Righthouse, hosted by Thérèse Murphy. The episode’s full title is “‘The People are the Superintelligence’: How AI affects political deliberation and democracy.” The transcript follows the one published by the Global Campus of Human Rights (PDF); links were added during editing.
Hello and welcome, everyone. This is a podcast series by the Global Campus of Human Rights. Our topic is artificial intelligence, or AI for short. More specifically, this podcast is about our human future with AI. I’m your host. My name is Thérèse Murphy. In this episode, I’m going to ask: how does AI relate to democracy? More broadly, I’m asking: how does AI relate to society, to people? I’m delighted that today my guest is Audrey Tang. Audrey has a fascinating bio. Her bio includes a background in civic tech, being Taiwan’s cyber ambassador, and also its first digital minister. Currently, Audrey is working on a range of projects, including a fellowship at Oxford University, where she’s developing new ways to think about how we align AI to society. Audrey, great you could join us. You’re very welcome.
Thank you, Thérèse, and I want to say first how glad I am to be talking with the Global Campus here, because the course you built — A Rights-Based Approach to AI — is exactly the frame I would choose myself.
Great to know, Audrey. So, let’s see how much of that we can explore in today’s podcast, to help our listeners understand more about what it means to take a rights-based approach — some would say an ethics-based approach — to AI, certainly a values-based approach to AI. Let’s get started with the questions, Audrey. I know that you believe digital technology can empower people, bridge divides, and renew democracy. Those are powerful claims. They’re very big claims, and I imagine that our listeners will be very interested in them, probably intrigued by them. At the same time, I think at least some listeners will be a bit sceptical. So, Audrey, tell us why you believe technology can renew democracy and empower people. Over to you, Audrey.
Well, when people imagine technology meeting politics, usually we picture the worst of it. Immediately people think about the echo chamber, the outrage that keeps us doom-scrolling, the platform that profits when we are at our angriest. And of course they would be right that this exists in the world. But I have come to believe that this high PPM, polarisation per minute, toxic environment is a choice in design — it is not an unavoidable feature of the digital age. So, we can build the opposite.
And one very particular, personal place where I learned that: in 2015, the ride-hailing company Uber arrived in Taiwan, and the country split — drivers furious, passengers divided, the government under tremendous pressure to simply ban ride-sharing. But instead of staging a debate where people shouted at each other, we opened a conversation on an open-source tool called Polis. And Polis has an unusual rule: there was no reply button, and you cannot quote someone in order to dunk, to attack them; you can only write your own short statement, or agree and disagree with what others wrote. About 900 people cast some 30,000 votes, with nearly 2,000 more joining the live sessions.
And the system is not about picking a winner — we drew a map together, and we found the sentences that the people in completely different camps — drivers, passengers, engineers — all agree on. We call them the “uncommon ground” — that rarely-discovered common ground — and starting with passenger safety and surge pricing without undercutting meters, and civil servants copied those points of agreement straight into regulation.
So, here is the claim, more concretely: I think the right to take part in public life is already protected pretty much everywhere in law — indeed, UDHR states it in Article 21, and the binding Covenant also says it in Article 25. So, what was missing was never the abstract right; it was the infrastructure to exercise the right to participate continuously, not just once every four years through voting. So, broad listening is that infrastructure, and to me that is not techno-solutionism — that is just a human right, a collective right finally given a way to be practised between elections.
That’s fascinating, Audrey. What you say resonates very strongly with me — that whenever the conversation, or even our own thoughts, turn to technology, we typically catastrophise. You know, as you say, we imagine the worst. But you’ve told us, and given us a concrete example — the arrival of Uber in Taiwan — that we can design in, we can build the opposite. We can build a future using technology in ways that align with people’s interests. And I love the way you talk about finding the uncommon ground, so finding common ground, which can be elusive. But again, you gave us a concrete example about how, using technology, common ground that feels uncommon, seems uncommon, was actually found. And that, looking at it from a rights perspective, the rights are there — we have the text. What we need to do is build the structures. And I know, Audrey, that’s what you’re doing, and you’ve done it in Taiwan, and you’re doing it, or have done it, in a range of places around the world. So that’s what I’d like to ask you about next.
Because given that you’re working in these different places, and also different spaces, around the world, and that many of your ideas you first developed, or they had their origins in Taiwan, where you’re from — in Taiwan you were part of the open-source movement, then, as we heard already, you were the country’s first digital minister, you’re a cyber ambassador for Taiwan — so here is my question. And I’m interested in this, of course, from a human-rights perspective, you know, about how ideas move around the world. So, the question, Audrey, is: you started to develop your thinking, your ideas, in one place, in one country, Taiwan, and today you and your ideas move around the world. So how is that possible? How do ideas that emerge in one place, that have their origin in one place, move around the world and have value in other places? Tell us about that, Audrey.
Indeed, the open-source and free-software communities are inherently global. We never lock anything behind a wall called intellectual properties — that was the entire point. So, when we built a civic technology in Taiwan, of course we also published the code openly. Polis was published from Seattle, using a licence called copyleft, which means any time we make a change to it, we also must publish our modifications, and then anyone anywhere can take it and fork it — to fork means to make adjustments and take it in a different direction, but maintaining the possibility of merging back with the upstream. And so, we are not exporting a finished product, but rather handing over seeds, and letting each community decide what to grow.
And I want to be careful here, because for a human rights audience this matters: the method was not born in this comfortable, high-trust democracy. Sometimes people think of Taiwan and they are like, “Oh, you’re in East Asia, of course you all agree with each other” — of course not. In 2014, the approval rating of our democratic leadership was at 9%, so in a country of 24 million people, anything the president said, 20 million people were against him. And so it was not high trust at all — it was forged in a low-trust, freshly democratic society that was, and still is, heavily targeted by our friendly neighbour in terms of information operations and cyber-attack attempts. So, we built this system precisely because we were fractured, and that is why it travels: the condition it was designed for — deep division, very high polarisation per minute, this high PPM condition — is what many people around the world are now experiencing. Fragile democracy is no longer an exception.
So, the examples are adaptations, not transplants. For example, the state of California launched “Engaged California”, which it describes openly as inspired by Taiwan’s digital-democracy work, used the same family of consensus-mapping methods, run by the state with Carnegie California and the Berggruen Institute, with the Project Liberty Institute, to ask people: how is AI impacting your work, your belonging, your care and dignity — and the social licence to operate for frontier AI labs can be co-created with the people, not just for the people. In Bowling Green, Kentucky, a process called “What Could BG Be?” used the same bridging-based listening policy for community planning and reached close to 8,000 residents — almost 10% of the entire population — with over a million votes.
And the exchange also runs both ways. In Japan, Takahiro Anno, now a member of the upper house and the founder of the party called Team Mirai — the Future Party — took these ideas and built something new and faster for the Japanese context, and we are now learning from Japan.
So, the deeper reason that these ideas spread is not technical — it is the desire to be heard, to help shape the place you live, to hold the steering wheel together. That is one of the most human desires there is. So, we are not just giving people that desire — the desire is already there; we are just finally giving it somewhere to go.
That’s inspirational, Audrey. So, it’s fascinating to know that — and I love the verb you use — that the ideas were forged, the vision was forged, in a place where democracy was fragile, where it had been freshly created. And any form of democracy, fragile or not, needs to constantly be created and recreated. But so, the ideas were forged in this new democracy, in this freshly created democracy, and then, as you say, they get picked up around the world. They’re not transplants, they’re not being exported — you call them seeds. I love this description. And that in places around the world, people want these seeds. They want the opportunity to come together and to plant the seeds and to see what can grow.
If I could jump from that to a question about your own biography, Audrey — imagining you as a seed growing is what put it into my mind — in the introduction, I touched on some what many people would call highlights in your career today: civic hacker, technologist, Taiwan cyber ambassador, previously its digital minister. I know you won a number of awards — you hold what’s often called the alternative Nobel Prize, that’s a Right Livelihood Laureate. Congratulations on that! I know that you’re working with a range of actors and organisations. What I’d like to ask you is to talk to us about the journey that you’ve been on in your life so far. And what I’d really like to know about is: what are you most proud of? Over to you, Audrey.
Well, I want to be a good enough ancestor that leaves the future generation a wider canvas than the one that I was born into. My journey really begins with a single lucky decision. Because when I was in conventional school, when I was 14 — not out of rebellion, I was just very curious — the questions I wanted to chase were moving faster than any of my classrooms could. And with the blessing of the head of my school, Tu Hui-Ping, she basically said, okay, you can learn instead from the internet. So, I became homeschooled and found the free software, then the open-source community — a culture built on mutual respect and radical transparency. And I realised that this code of conduct was never just about the software kind of code; it was about a code for people to communicate, and also to govern ourselves in a different way.
So, I think the system that allowed not just 14-year-olds but also pretty much everyone — people who do not otherwise have voting rights — to educate themselves, and to also make impactful decisions on issues concerning them, that is what I am proudest of. By 2022, according to the ICCS study, 14-year-olds in Taiwan are now top of the world when it comes to civic knowledge. Instead of just literacy, where we consume information, it is the competency, the civic muscle, so that young people in Taiwan now know that they can shape the country before they turn 18. And that is what I think the lasting legacy of this kind of open-government work is: to empower the intergenerational bond — the 17-year-old, the 70-year-old, finally see each other as complementary, as the best friends in pushing for a more fair, more inclusive, more equal future, because these two groups care the most about the future, not just the next quarter.
As to your question, I think it is not a single law, it is not a single app — it is that we change the relationship between generations. We moved from a government that trusts only itself to a government that trusts its people, from building things for the different generations to building things with the different generations. And the Right Livelihood Award, for me, belongs to everyone in Taiwan who helped prove that. And the work I am now focused on is to carry it into the Age of AI — what I call Civic AI. With my co-author Caroline Green of the Oxford Institute for Ethics in AI, we ground AI in the ethics of care pioneered by Joan Tronto. So, the idea, very simply, is that we need to align, to steer AI systems not through a one-time safety patch, but rather an ongoing relational health — something we tend, like a garden.
Thank you, Audrey. And I’m intrigued by the theme that’s running through our conversation today. You just referred to caring for people, for our democracies, for our societies — caring for them a little bit like a garden. And earlier in the conversation, you referred to seeds. So, this is a very provocative and insightful way of thinking, I’m finding. Also profoundly important to hear you talk about your own education and the transformative power of access to education, and in particular building what you describe as this civic knowledge, the civic muscle. And that it’s not simply about the 14-year-old, the 17-year-old, but also the 70, 74, 77, etc. year-old, and how building that civic muscle means that they can work together. And by changing the relationships, we empower people. Very, very powerful to listen to that, and have the gardening, the growing, the tending, the caring vision in our head.
So, I guess I feel an obligation now, Audrey, to ask — you know, in the midst of this inspiration — to ask: isn’t there something that’s worrying you? What would that — what is that, Audrey? Like, what is worrying you? And then, because I sense you are somebody who, you know, you believe in building infrastructures, building new ways to think about things — so even if something is worrying you at the moment, is there also something that’s really exciting you at the moment? Maybe it’s your work at Oxford University with your co-author around care, maybe it’s something else. Over to you.
Well, what worries me the most is the pull toward hypercentralisation. When times are uncertain, there is always a reflex to consolidate, to rally around the flag, to imagine that all the world’s information should be funnelled into one giant cloud, one supermodel, one place that sees everything and defends us against the various different risks. And so many people approach AI nowadays with a mainframe mentality — mainframe is the cloud during the 1980s when I was born — and that concentrates risk, because it builds a single point of failure and a single point of control.
And for years, some people in the tech industry have told us data is the new oil — something inert, something to be extracted, piped to a central refinery, something that you can sell as plastic or something. So, I want us to abandon that picture altogether. Data is not drilled like oil, not sold like plastic, data is soil. Data is something that is based on relational aliveness. Soil is alive, and it is relational: different soils grow different crops because they hold different histories, different climates, different patterns of care. If you pull data out of the place and the community that produced it, much of its meaning simply drains away.
And the human-rights question follows immediately: who owns the soil? Who tends the soil? And my answer is that data must be shaped by the community — not pulling humans into the loop of AI, but rather AI tending the relational health of communities, AI in the loop of humanity, where the community sets terms, not the platform.
So, what excites me is that people are working on the new trajectory. The superintelligence that everyone is racing to build in some distant data centre is not something that I am betting on. The superintelligence is truly us — it is the collective wisdom of people when they are given the tools to hear one another. And we are watching civic technologists and communities all over the world now taking open tools and adapting to their own challenge. Right now, we are living through a renaissance, like the personal-computing revolution, with AI. People are finally realising we already have the blueprint to rebuild civic trust. In offline mode, with no internet connection, you can run already frontier models on your own laptop. And so I am not excited because the danger is small — I am excited because the alternative to this hyper-concentration is real; it is already being built in a hundred places at once.
It’s a powerful vision that it’s already being built — you say in a hundred places, probably even more, I imagine you just chose that number — at once. And again, I heard you there continuing to paint this picture around the idea of caring for the garden, caring for the earth. Data is not like oil, like drilling for oil — data is soil. That’s such a powerful image, Audrey.
To close our conversation today, Audrey, what I’d like is to ask you: what message you’d like to leave our listeners with. You’ve provided many different inspirational messages during the conversation, but with what message in particular would you leave our listeners as they seek to contribute to what you’ve described for us today — building a future where we have a human-centred future with AI? What would you say to our listeners?
It’s a very simple message: do not wait for governments and big tech to decide AI for us. AI is not a force of nature — it is a tool, an artefact made by each of us, and it has to be guided by each of us. So, the shape of a human-centred AI future depends on whether we show up to shape it.
And I don’t just mean engineers — artists, civil servants, educators, historians, shopkeepers are not guests we just generously invite in as customers; we are already doing the work. Historians already keep civic memory, shopkeepers already know their local supply and their local relationships better than any dashboard. So, our job is to make that knowledge legible, and not just to hand out some universal basic income.
So, I think the actionable message is this: learn about personal intelligence, or PI. You can go to civic.ai to learn about my own personal intelligence setup — I call it Kami, or knowledge artefact management intelligence. It is very easy to set up. And democracy, like any technology, any social technology, is a muscle — and like any muscle, it needs daily exercise, not just one workout every four years. So, using Kami, you and your friends can build and break and rebuild those tools; you can debug the system that you will one day govern. And the book Plurality, freely available at plurality.net, is in the public domain for exactly this reason. So, once you have some experience working with Kami systems, read the book Plurality — take it, fork it, teach it; consider it a gift.
Thank you for these gifts, Audrey, and for the message that we need to build and to retain and develop our civic muscles. And to do that, like other sorts of muscles in the human body, a specific muscle needs to be exercised. And thank you so much for sharing a range of resources with our listeners that will help them to get started or to refine the exercise they’re doing with their civic muscles. As you said, to shape our human-centred AI future, we need to show up — each and every one of us needs to show up, with civic knowledge, for which we need this civic muscle. Audrey, it’s been inspirational talking with you today. Thank you so much for sharing your time with the Global Campus of Human Rights.
Thank you for this. And may the Global Campus of Human Rights live long and … prosper.

