Skip to Main Content
institute
of desiGn
Search

Bonus: AI and Democracy with Terry Williams-Willcock and Chelsea Pratt

August 5, 2025

36:06

S3: Bonus

00:00
00:00

In this episode from ID’s Shapeshift conference, we hear from two New Zealand design leaders on how intentionally designed AI can address challenges to democracy. Terry Williams-Willcock, Chief Customer Officer at RUSH Digital, and Chelsea Pratt, Design Lead & Strategist at Thoughtfull, discuss how they’re making civic participation more accessible with AI.

Transcript

Intro

Welcome to ID events. A series on the With Intent Podcast from the Institute of Design at Illinois Tech.

This past May, design and tech leaders gathered at ID’s Shapeshift conference to reimagine how we approach AI—shifting the conversation from what technology can do to what we think it should do.

In this episode, we’ll hear from two New Zealand design leaders on ways that AI, when combined with intentional design, can address challenges to democracy. Terry Williams-Willcock, Chief Customer Officer at RUSH Digital, and Chelsea Pratt, Design Lead & Strategist at Thoughtfull, discuss their approaches to making civic participation more accessible.

Here are Terry and Chelsea on using design and AI to strengthen democracy.

Terry Williams-Willcock 00:45

Kia Ora, that’s Hello in Māori. Yeah, thank you. It’s an absolute honor and I just want to, I suppose, express the same gratitude you guys reaching out to us early on in this was pretty special. We were all super excited to be part of this. So yeah, thank you both of you, Albert and Anijo for that. I do want to open this with a Māori welcome. It’s not a prayer, it’s just a way to really set the scene for this presentation. Really, it’s called a mihi. It’s actually from the New Zealand Designers Institute as yeah, as a way to open up talks and events like this. So without further ado, I will do a thank you. That’s super nice. Thank you. I’ll just explain what that means because there is deep meaning behind all of these beautiful words. So thank you for scrolling back up because I know it in Māori, but I actually don’t know it off by heart in English, which is an ironic situation. So it talks to pulling together the rising essence, pull together the collective belief so we can ascend our crafts, strengthen our relationships, to bring prosperity to our worlds, to lift our people, let the shelter of our home rise, remove the murk and let us find in clarity, hand me the taonga, which is a treasure, a beautifully carved taonga, which is a treasure in our country and we extend the breadth of life to all of you. Without further ado, I’ll hand over to Chelsea to talk about what we’ve been up to.

Chelsea Pratt 02:45

Kia Ora, I just want to firstly extend my gratitude as well. It’s amazing to have come this far for such an event and grateful to be here with you all today. So, okay, I want to start with a simple concept. AI won’t save democracy, but design might. At the Auckland Salon back in September last year, we were tasked with the question of would AI supercharge or cripple democracy, essentially how people engage in society in New Zealand? And it’s a provocative question, it really is. And the truth is, it could do either or it could do both, because the outcome really doesn’t depend just on the technology, it depends on us as designers, on the choices we make, the values that we build into the systems that we create. And it really depends on the questions that we ask ourselves when we’re actually coming to design. What is this for? What problem is it solving and who is it serving?

Chelsea Pratt 04:04

So we are here today as two award-winning design agencies from New Zealand Rush, alongside with Thoughtful Co-founded by Jeff Sco as Anijo mentioned and brought together by a shared obsession with solving human problems at the messy intersection of technology and design. And while we may be perceived as somewhat competitors in a small market back home, we believe that the most important problems are solved together because after all, you may be able to move fast alone, but you can move a lot further together. And so while our work coexists with technology, it starts in the real world working alongside real people with real needs in real communities. And we are lucky enough to do that with some pretty amazing people and in a pretty special corner of the world. So this photo was taken across the road from where I live. And although we are yes, an island nation, somewhat small and somewhat far away from here, New Zealand has always been a nation of firsts. We were the first country in the world to give women the vote, pioneers in innovation, the first jet boat to go up our winding rivers, the first to create bungee jumping.

Chelsea Pratt 05:24

I mean, who decided it was a great idea to jump off high things with a stretchy cord attached to your feet? There definitely is no shortage of courage in New Zealand.

Chelsea Pratt 05:36

The plastic syringe was invented by a kiwi which has gone on to change the safety and accessibility to modern medicine for millions. And last but not least, the anti-nuclear stance that New Zealand took as a country led by collective values. So we are a country that is small enough to experiment, but big enough to matter, and most importantly, a country that’s well trusted for boldness when it comes to doing the right thing in certain scenarios. And that is critical in the topic that I’m talking about today because trust really is the currency of democracy. But right now that trust is under threat. When I was planning for this talk, this meme came across my feed and I thought it was particularly relevant. As designers, we are inherently taught to be curious, to be excited to look ahead and think about everything with possibility and what could be and what could we create.

Chelsea Pratt 06:37

But when it actually comes to design and curiosity today, it’s becoming quite a tricky topic. We live in a rapidly changing world and it’s sometimes simply asking a question is becoming a lot more overwhelming or trying to learn about a complex topic. When you really want to go online, ask a question and just understand something more deeply. As designers, that’s what we are taught to do. However, go online today and you do risk being labeled, judged or simply misled and AI. Well, AI doesn’t create this problem, but it certainly accelerates it. And in many instances it’s spreading a lot faster than the truth. And we’ve heard a lot about the truth today, but when trust breaks down, engagement breaks down and people stop voting, they stop thinking that their voice matters. In New Zealand, youth voter enrollment is dropping in the UK their overall engagement with voting is the lowest it’s been since 2001.

Chelsea Pratt 07:45

And globally trust in institutions is at an all time low. But here’s the most worrying part. People aren’t disengaging because they don’t care. They’re disengaging because they don’t know who or what to trust. And for minority groups of citizens, that challenge is exacerbated. It’s even worse. More and more people are starting to rely on voice assistants and AI driven technology for different reasons. That could be to do with impairment, visual impairment, physical impairment. It could be to do with language barriers or simply the convenience that it is bringing to our world because it is becoming easier and easier to access information through technology. But if the information feeding those tools is incorrect, biased or incomplete, then that’s not just an inconvenience that is a democratic barrier.

Chelsea Pratt 08:46

I want you to imagine this. It’s coming up to an election. It’s time to start to think about it and what you are going to vote for, what do you care about and what matters to you and what impact do some of the policies and parties out there have on you and your community, but now take away your vision so you can’t see. And so you make your way around the world without your sight. You rely on those around you, you rely on technology and you rely on the trust that you have in those things. But how do you actually know the truth behind all those things? When it comes to your friends or family, you’re asking questions to understand what policies may mean for you. If you can’t visually see it yourself or you’re relying on voice activated things, transcribing it back to you, but how do you actually trust what’s behind that? As technology evolves and changes so rapidly, so then when you turn up to the voting booth on the day, again, it’s not designed for you. How do you find your way into the voting booth? At the voting booth when you’re actually placing your vote, someone is obviously with you. And so again, you have to trust that person and inherently trust what is actually being put down and what does that mean for you.

Chelsea Pratt 10:07

So as you can see, the ability to participate in society is becoming more and more dependent on the quality of our information. And so we quickly realized that actually the question that we were posed, the issue starts way back before actually getting to the voting booth. It’s on how do you actually understand what the policies are? What impact did they have on you as an individual and those closest to you? So we asked ourselves, what would it take to design a way for people to safely, confidently and curiously explore complex issues? We wanted to not just talk about the issues but actually create something that would truly make a difference. And there has been a lot of talk today around the rapid evolvement of technology, making it easier and easier for dreamers to become builders and builders to become dreamers. But with that comes a lot of responsibility.

Chelsea Pratt 11:07

And so when we were tasked with our salon in Auckland last September, we decided to turn it more into a design sprint. We wanted to get voices at the table from all sorts of different industries that had all sorts of different backgrounds to help us solve this problem by taking away our biases and bringing in a whole lot of different expertise. So we had policymakers, we had government heads, we had tech leads from the likes of Google, we had educators, we had students, and of course we had designers as well. But the intent with this was to actually dig deeper into the question and actually pose the question to a broader group of people and actually dig into it. As designers, again, as I mentioned, we are inherently curious. We wanted to be able to create a space within those confines that we could be curious.

Chelsea Pratt 12:00

And with that, we started to look specifically at minority groups. So we tasked each group in the room with a different citizen group. So we looked at first time youth voters and obviously a lot of challenges that come up in that space when it comes to youth trying to figure out what they believe in, what’s important to them, what’s happening as the future is rapidly changing, what do they believe is different from maybe what their parents believe and how do they start to understand what their take is on policy. We also looked at elderly. Obviously with the rapidly changing technology, a lot of them aren’t engaging or don’t want to engage with the new technology, but with information moving that way, how do they stay in the loop? We looked at citizens who were in remote locations, new citizens who had just come to the country and maybe language was more of a barrier.

Chelsea Pratt 12:54

And then we also looked at physically and mentally impaired citizens. And this was where it ended up that the focus sat with the visually impaired was the area that we decided to focus because so many themes started to come out and it really felt like that was the area that the tools were missing the most. So in the classic design process as we do as designers, we went out and we spoke to people who were blind, who had visual impairment and really asked them what were the biggest challenges when it came to understanding policies. When it came to voting and actually really getting to the forefront of the issue. We then started to look at actually how we could bring that into a tool that was really usable. And what we learned quite quickly, which is often the case when you focus in on the minority group, is that what we were starting to build and develop was actually a lot more usable for a much broader audience.

Chelsea Pratt 13:57

And so a key theme started to arise when it came to what the tools of AI could be doing. Because here’s the thing, AI can be a wedge or a bridge, it can drive us further apart or it can help us to meet in the middle. And it all comes down to the questions that we ask and how we ask them as designers. So what we designed helped people who feel like they’re not part of the conversation to actually chime in a solution designed not just to inform citizens, but to empower them, to help them navigate complexity with confidence, to ask better questions and to stay curious safely. So now I’m going to put on a wee video, which actually we’ve put together that sums up the process that we’ve done some of the research that we conducted and also brings up the concept that we’ve developed. And then I’ll pass over to Terry to talk to the concept.

Video Voice Over 15:07

If you believe that democracies basically depend on trust, the arrival of AI powered misinformation, fake videos, fake messaging and so forth, especially targeting at you could really put democracies at risk.

Terry Williams-Willcock 15:21

Certainly important that we think about the consequences of our design and how we do influence behaviors. We clearly see it right now that is kind of under threat. The algorithms are pushing us apart, our opinions are becoming a bit more divisive. Democracy really does. It leans on the ability for us to want to have a conversation and not hate our opposition, but kind of maybe just disagree and disagrees. All right? Democracy only works when everyone can participate fully. So in our research with disabled voters, we consistently heard that accessing broader political information across broader channels was a major check

Video Voice Over 16:02

For youth. Think that the perspective of deaf or other differently abled people is not given enough voice in the country to shape the direction of the country.

Video Voice Over 16:18

You are spot on. There’s screeds and screeds and screeds of information and it’s all in English and it’s splash English. And sometimes you don’t understand the content, which means you’ve got to ask someone what it means and then that takes forever. The Māori hui to Wellington, I have no clue what that’s about, but I don’t know why the hui and what the issue is with national. I do not understand. You’ve got to pick up bits of information and it all becomes too hard.

Video Voice Over 16:48

How do you combat this amount of information, overwhelm, have it delivered to you in a way that makes sense and is relevant for you as a person, but do it in an unbiased way.

Terry Williams-Willcock 16:58

It’s a harder problem to solve a lot of the time when you’re dealing with someone who can’t ingest information in the same way that everyone else and actually they disengage from the democratic process. What should we do? How can we narrow this down? Talking about common ground and finding common ground and

Chelsea Pratt 17:15

The idea of like, let’s get someone else to chime in here

Terry Williams-Willcock 17:19

When it first opens, I’m imagining it talking,

Chelsea Pratt 17:23

It’s an important one too. It’s almost like a can you choose the voice? Voice is quite polarizing if you find it an annoying voice or kind of the type of voice that maybe you think is biased.

Video Voice Over 17:33

So in this AI driven future when prompting and all of the complexities of which model to use and how do I use this feature and how do I use this platform, we’re simplifying that for you. We’re orchestrating everything in the background and we’re doing it very simply.

Video Voice Over 17:50

So just pick the Treaty of Waitangi.

Video Voice Over 17:53

So it’s, we’re going to talk about something pretty fundamental to Aotearoa, the Treaty of Waitangi. It’s the bedrock of our country, but figuring out what it actually means. It’s been a bit of a wrangle, hasn’t it? The main issue is that there are two versions, English and Te Reo Māori, and they don’t quite say the same thing.

Video Voice Over 18:12

Kia Ora, I like the concept and I like the conversational tone. I understand the limitations about our limbs. It would be good to have a Kiwi voice on there. And it’s good that you can fact check it while you’re listening to it. So you getting, it’s not slowing you down.

Video Voice Over 18:26

What does a truly equitable partnership look like in practice? It’s a conversation we definitely need to keep having.

Anijo Mathew 18:34

What design brings to the table is a quintessential sense of optimism that in the most horrible of times there is a preferred state that exists and we can get to that state.

Chelsea Pratt 18:51

And I’ll pass over to Terry.

Terry Williams-Willcock 18:53

Thank you. Amazing. So we’re a bunch of optimistic kiwis, so it was nice you talk about designers being optimistic. Let me move over to the product. So I’ve got the excited duty to take you through what we’ve created. It is very much at its infancy. Am I going the wrong way? Let’s go the right way. We’ll get that. Yes, we’re in. So yeah, I can take you through Chime and like I said, it’s very much in its infancy. So we’re on a bit of a journey as you do as designers, but we’ve created something I think pretty quickly and something that really shines a light on I think the right direction we’re taking. So what is Chime? Chime essentially transforms complex political content into engaging podcasts, engaging audio files. It’s an audio experience and we believe it makes democracy accessible through the power of AI and encourages conversation without judgment in a safe environment.

Terry Williams-Willcock 19:58

So these were some key things that we wanted to bring into the experience simply. You can see here, you just simply talk to it, ask it a question of any political topic you might be interested in. There’s some little kind of example questions there that you can get straight in because you might not know what you want to ask. And then it will generate an actual podcast as you can see there. And then really importantly, once you’ve listened to that podcast, it invites you to ask follow-up questions. It keeps you engaged in the conversation that you’re about to start. And that’s super important. On the right hand side, you see that it starts to show you the different sources that it’s used to create that podcast. Again, I think we’ve all seen today different LLM functionality that enables this to happen. But what I will do is I’ll give you an actual audio, an example of the audio used so you can kind of get a sense of how we prioritize the different parts of, I think the system, the AI system that we’ve created. So let me before I go into this, what it actually talks about is the Waitangi Treaty. But the Waitangi Treaty is essentially the foundational document that acts as a governing document that is an agreement between the colonizers, the British, I am British, slightly awkward, but it’s okay.

Terry Williams-Willcock 21:21

And the indigenous culture, the Māori. And they wrote a set of principles, a set of guidelines that helped them decide on how they were going to come together and make decisions. A hugely important document that has actually been held up really well within New Zealand. Super proud of that. I’ll just play the audio that starts to explain what it is based on the question that’s been asked.

Video Voice Over 21:42

Kia Ora everyone and welcome. So you’ve probably heard a bit of a kerfuffle about the treaty principles bill A, it was a real hot potato, wasn’t it? Basically it was this bill put forward by ACT’s David Seymour, remember him, that aimed to officially define what the principles of the Treaty of Waitangi actually are. Think of it as trying to nail jelly to a wall. Pretty tricky. The idea was to put these defined principles to a referendum, get the whole country to have a say and give parliament the final word on it, rather than leaving it up to the courts. Pretty bold move. Aye. Now, the supporters, mainly ACT reckoned, the current interpretations were all a bit skewed. They felt the treaty was being used in ways that weren’t really in line with what was originally intended and that it was leading to some unfair outcomes, creating different rights for different people. The whole different rights for Māori and Pākehā thing, that’s what they were on about. But heaps of people were against it. They were worried. It undermined the treaty’s whole constitutional importance. It’s standing as a foundational document for the country. They argued it ignored the treaty’s promise to protect Māori culture and allow Māori to thrive as Māori. Plus. There’s the whole historical context. The less than rosy details of colonization and the promises that weren’t entirely kept about land

Terry Williams-Willcock 23:10

Might stop it there. I think first of all, there was a slight giggle at the front and that’s fine because the actual voice isn’t the intended voice that we’re hoping for. But there is some, the way we’ve scripted the prompts, it’s actually quite a fun and engaging experience and I think that’s important, right? You don’t want it to be a dry political topic, you need to add a bit of fun into that. But yeah, there’s definitely the accent needs some work, but as you can hear, they’ve started, we’ve started to bring in elements of Te Reo, which is actually, like Anijo mentioned, we speak Te Reo within the English language and it brought in some kiwi slang, a bit of a aye every now and again, that’s a classic kiwi slang there. Well, it gives it a go anyway. The important thing in terms of how it sets it up, it starts off by giving you kind of official description of what the question was.

Terry Williams-Willcock 24:02

It doesn’t add a point of view to start off with. And then it talks about the two different kind of views and that priority that dealing with the first thing first and then thinking about each angle and the different perspectives was super important because we don’t want to favor a single view. We want it to be as unbiased as possible. And then what you start to see is that the different sources that it’s pulled that information from the legislation.gov.nz, it’s gone, right? Let’s look at official sources to get the data, also data.gov.nz and then it started to look at some trusted media sources within New Zealand. So I think, I hope that’s given you a bit of a sense of the experience and we will click again. So behind this experience is reasonably sophisticated, I suppose an AI system workflow. And before I go on, I just want to call out Derek and harnick.ai.

Terry Williams-Willcock 25:05

There he is. Thank you very much. Thank you. That’s very kind of you. I didn’t say you’d get that. Who basically helped us code this and did a lot of the API calling prompting and used what we’ve been talking about today, vibe coding to get this happening within, I think it was 60 hours, about eight days of work to get this whole thing up and running. So pretty amazing. So it starts with assembly AI using the universal modal to take your audio questions and transform it into a text file and then it uses perplexity and pushes out to perplexity your kind of question. And then goes through a lot of sources to find the most relevant content and prioritizing the local government policy data over the media content. And then it’ll use Gemini to craft that content into a beautifully scripted, a beautifully crafted script.

Terry Williams-Willcock 26:04

And that’s where it starts to inject obviously the Kiwi and the Te Reo and the slang and it takes out some of the kind of awkward nonverbal elements and the formatting issues and all of that is designed to give us the best chance to create a really engaging but accessible piece of content. We finalized it with OpenAI, but I think it was because it was a free subscription so we could quickly do it and therefore the voice is the American kind of accent. But there is great advancements of technology with a company called Kū who have actually looked at creating an AI model or they have created an AI model to basically use the correct pronunciation of Te Reo. That’s by Peter Lucas who won was on the top 100 time magazine for his work in that field. So pretty amazing. So we are hoping to kind of pull into that stuff to really connect you with the content that it gives you.

Terry Williams-Willcock 27:05

I wanted to go a bit deep into the actual prompt because I think the interesting thing about this is that it starts to really give us a direction in what we’re looking to actually achieve with the script. This is the key part of the process. So the prompt says provide factual balanced information using diverse authoritative sources. That’s the goal. That’s what we want you to do. Include recent academic research, policy information, government statistics, and reputable media when available, prioritize New Zealand Pacific data. We want to make it as regional as possible, which gives us the best chance that it’s going to be as legitimate as possible and provide. And one of the most important things here is provide the multiple viewpoints without favoring any position and clearly distinguish between facts or emerging kind of data. So all of these different prompts we believe gives us our best chance of getting us to the output that we are looking for.

Terry Williams-Willcock 28:06

And I suppose before I go into that, the important thing there is that you can’t be sure of the output. That’s the slight nervous thing that you put out to LLMs. You go, I want you to do this, but you’re not entirely sure. So those prompts were there to give us the best chance to do that. I’ll talk a bit more about that in a minute. So we tested this, we wanted to get it out as quickly as we could. And so at the moment we’ve tested on very opinionated friends and colleagues and we’ll follow up quite quickly with the Blind Citizens Foundation that we have good relationships with. So themes came through pretty quickly that actually gave us a lot of confidence. We were heading in the right direction. Convenience of the audio came up straight away and that was interesting. It came up with able people, and this was one of the key things we put in there for the not blind citizen people as well.

Terry Williams-Willcock 28:59

Points I could quickly check like political facts on my way to work sounds great as I’m time poor, these type of comments were coming out. And then transparency became the next clear kind of piece. So it wasn’t just swayed by a single source, it clearly showed you that it came from multiple sources. But what was most rewarding was that curiosity we saw. We actually couldn’t get people out of the testing rooms, they just kept asking more and more questions about these political issues, which weirdly they don’t ask about in normal life, which was kind of actually really invigorating for us to see that happen. But it’s a bit of a sad truth.

Terry Williams-Willcock 29:38

There’s lots of improvements to make. Like I said, we’re very early on in the journey. So the time to load this because it’s pulling in so many LLMs is actually quite a long time. Perplexity takes quite a long time to do what we need to do. But within the time we started this, we’ve already seen advancements that gets that going a lot quicker. The actual interface itself, not very good. Sorry I did that one. But the learnings from some of the stuff we’re seeing at the moment and how it could be a lot more intuitive, so we’d want to advance that as well. But the things that came through became I suppose what we form as our design principles for this whole project. And that was super important for us. We were in our minds, we’re going, are we actually now something that is accessible? It is safe and it is a trusted place to engage in democratic content.

Terry Williams-Willcock 30:31

And this is super important definitely when we’re trying to design with AI. A chief strategy officer at Rush, Steven Ner quoted, while deterministic design targets perfection, and we all love that as designers, we love perfection. And that’s in an idealized world. Probabilistic design acknowledges and prepares for the messy complex reality of AI. And we are trying to direct AI to go from all these almost millions of different data points, the content down to a small set of outputs, which is the script. And the thing is we can never be certain at that final output, but what we can do is prepare for it. We can give ourselves the best chance to get the output that we want.

Terry Williams-Willcock 31:22

So with Chime, what we are doing, which is super important, is we are optimizing the system towards transparency of sources, convenience of the experience, diversity of viewpoints and facts over opinions. And we believe that gives Chime the best possible chance to deliver on our goal, which is trust. So when we think about that, not just within Chime’s context, but when we’re thinking about that across other platforms we’re designing for, that approach is so important. We have seen that the commercial incentives that we put at the top of our, I suppose priorities of our decisions, they start to inform the design principles and we’ve seen some pretty bad commercial incentives happening. The attention economy is a classic example of that. And then that is informing the principles that we apply and then that also informs the systems and the patterns that as designers we design the like button as an example, those types of things, you have to do that because you’ve been incentivized from the top down and that creates commercial gains. But at what cost?

Terry Williams-Willcock 32:29

A great quote from Katrina Alcorn, GM of design for IBM, which was in one of the pamphlets. I think one of the biggest misconceptions is the belief that design is just about the surface details and not understanding that design is about the whole thing all the way from the top to the bottom there. It’s about framing the problem you are solving that sets the writing tensions. So I hope we can think a little bit more like this human values right at the top. If we instill positive human values that direct us to consider more humane design principles, these will influence the patterns and the systems that we design with AI that could have such a dramatic positive impact on humankind. And here’s the thing, I think we really need to impress on everyone that this is super important right now. We’ve heard Eric Schmidt, the ex CEO of Google talk about the arrival of artificial super intelligence.

Terry Williams-Willcock 33:23

No one knows exactly what that’s going to be, but it sounds pretty powerful. And so we think about what we’re doing today, we might be encoding into the future of that. So let’s be very, very cautious around that. So AI, to your point, Chelsea, it won’t save democracy, but the way in which we wield it might. Democracy depends on a shared truth. We’re living through what is coined the information revolution that threatens this foundation. Putting democracy itself at a crossroads. Democracy isn’t just about voting, it’s about making decisions based on shared information. When we inhabit separate realities, when my truth and yours, they never intersect, how can we possibly govern together without a shared reality? We’re not debating, we’re shouting across universes that never touch.

Terry Williams-Willcock 34:21

Sorry, could you just get to the bottom of that slide? So with Chime, we believe that there’s a potential here that we could be rebuilding this bridge by creating trust in the information we consume. We can restore the foundation that democracy needs, citizens participating in the same world even as we envision different futures for it. But like I said, we have started this journey, it’s very early on and we’d actually like and we’ve come here to invite you all to actually be part of this journey. Simply feedback would be awesome. That’s always the first thing that we ask of anyone, but we need to understand how we can take this to market. We’ve got some people in here that have scaled products to billions of people. We’d be happy to scale it to. How many have we got in New Zealand? 4 million. That’ll do. So we’re genuinely asking for your support and your help because like I say, I think we’ve got a potential to really make a difference here. So I believe together we could turbocharge democracy and not cripple it. And I’d love to just finish on a little Māori proverb that I think is super poignant to what we’re all talking about here. He tangata, he tangata, he tangata. What is the most important thing in this world? It is people. It is people. It is people. Thank you.

Key Points

  • Why “AI won’t save democracy, but design might.”
  • The democratic crisis of trust: why people aren’t disengaging from voting because they don’t care, but because they don’t know who or what information sources to trust.
  • Designing for the margins to benefit everyone: how focusing on visually impaired voters revealed universal barriers to accessing political information and participating in democracy.
  • Their work on Chime: using various AI tools to transform complex policy content into engaging, multi-perspective audio experiences.
  • Why human values must drive AI design principles and move beyond commercial incentives toward transparency, accessibility, and shared truth as the foundation for democratic participation.

Additional Resources