Community Technology is the Future with Dave Griffiths

What happens when technology is built with communities instead of just for them? In this episode, Eleanor Drage speaks with David Griffiths, founding director of the nonprofit Then Try This, about participatory design, accessible tech, and the hidden histories shaping modern computing.

The conversation explores how textiles and programming are deeply connected, and why innovation is often rooted in grassroots, community-led creativity rather than top-down systems. Griffiths shares projects like Sonic Kayaks, which turns underwater soundscapes into accessible environmental data, and Nurgle, a game that uses audio cues to track public health trends. Together, they discuss how inclusive design can make technology more accessible, sustainable, and responsive to real-world needs.

David Griffiths is the founding director of Then Try This, a nonprofit organization that develops creative, community-driven approaches to technology, science, and public engagement. His work focuses on participatory design, digital inclusion, environmental sustainability, and the social histories of technology. Through interdisciplinary projects that combine art, science, and technology, Griffiths explores how innovation can be made more accessible, collaborative, and responsive to community needs.

Reading List:

• Then Try This

• Indigenous Circuits: Navajo Women and the Racialization of Early Electronic Manufacture

• Queer In AI: A Case Study in Community-Led Participatory AI 

• Sensing Bodies: Engaging Postcolonial Histories through More-than-Human Interactions

Transcript:

Eleanor Drage (00:53)

Hello and welcome to The Good Robot. This is Eleanor Drage. And today we're going to be talking to the awesome Dave Griffiths, who's the founding director of Then Try This, a nonprofit research, design, and development organization based in Cornwall, UK, that works on environmental and social projects. So the buzzwords around this will be citizen science, community technology, and participatory science and technology.

And one of the reasons I'm talking to him today is that I wrote an article in The Guardian a couple of weeks ago about frugal tech, and it went down really well. It argued that the most amazing innovations of the moment aren't just further iterations of GPT, but really cool stuff that's being designed in low-resource environments by people with less. So I'm really excited to speak to David today.

Our first question is: what do you do, and what brings you to the subject of technology and justice?

Dave Griffiths (01:56)

So I run this independent research organization with my partner, Amber. I think it's always very difficult to explain, but it's best to explain it in that we are trying to do research a bit differently, or just do things a bit differently. My partner's background was in academia, and my background was in commercial corporate research and development.

So we were both doing things how you're supposed to do them, and we both sort of reached the point in our careers where it wasn't making any sense anymore. We had to just try something else, and we decided to try to build something that did things more interestingly. In terms of your first question around gender, feminism, and justice...

I suppose my personal background is that I was part of that quite annoying generation of people who learned to program when I was about 11 in the '80s on a ZX Spectrum computer in the front room of our house, plugged into our TV. But in the same room as that was this enormous floor loom because my mom was a weaver.

And so on a Saturday morning, I would be sitting there trying to make different patterns of pixels on the screen, and my mom would be there weaving patterns on her loom. It was just immediately obvious to both of us that we were doing the same thing. Obviously, there's a superficial, surface-level similarity between the stitches in textiles and pixels on the screen, but it was also the way that you have to approach it. She was really more interested in essentially programming the loom, setting up and warping the loom, and setting it all up to do a series of patterns. I was more interested in writing the code and understanding it, making some patterns, and then deleting them to start again, rather than focusing on a finalized product at the end of the process.

Eleanor Drage (03:57)

I love that. Because you were doing these two different activities in the same space, it allowed you to see the points of connection between the two. Weaving and tech development are usually done in many different places, but they are rarely done together. That proximity isn't allowed to emerge in the same way. So, can you tell us then, we are The Good Robot, so this is our kind of pithy provocation. What is good technology to you? And is it even possible? How do we work towards it?

Dave Griffiths (04:33)

So yeah, I've listened to a few of your podcasts, and it's difficult to add much more to what people have said already on your show. But I suppose for me, it's thinking about technology that acknowledges the social and political environment that it's made in, who made it, and who it's for. I think this is what makes it different from the idea that technology just exists in isolation, as if it's completely detached from everything else and is a subject in its own right. It doesn't really make sense in that completely isolated world.

The problem with that, or maybe the opportunity of accepting that it's a human thing, like politics and history and all the things around it, is that it allows for criticism. Then you can ask questions as to why this piece of technology is shaping my behavior in this way, or why things are built in a certain way. Do they have to be? It requires a bit more confidence in technology. It seems like, with the way a lot of things are built, there's not a lot of trust in people. There's an assumption that you already understand all the uses that people are going to have for it. So I think good technology accepts that maybe the people using it know more than you do, and it allows the technology to be reconfigured, changed, and questioned in a different way.

Eleanor Drage (06:07)

I really like how you said that if we design and see technology as a social artifact, then it opens it to critique. This seems exactly right, because the reason why the assumption that technology is neutral or objective exists is that it stops people from critiquing it. How can you possibly critique an objective, neutral tool? So it opens it up to all sorts of things that science, or the scientific context, I should say, has been afraid of for a long time.

The trust element is really important. What I hear a lot, particularly from governmental funding bodies and foundations, is "how can we improve public trust in AI?" But you can only get people to trust trustworthy technology, which leads us to the next question of participatory design. Do you think that trust and trustworthiness have to come together? What's the relationship between trust and trustworthiness, and how does participatory design help us work towards both?

Dave Griffiths (07:05)

So I think participatory design is part of a long series of these kinds of approaches, like co-design and user-centered design. I guess they are trying to put names on a way of getting around this challenge: how do we build these things into how we make things? It tends to come more from the academic side, but it's certainly something that we pick up a lot and use in the kind of work that we do. We just try to use it as a way of bringing a bit more common sense to this kind of thing.

Eleanor Drage (07:58)

"Participatory design" has become a bit of a buzzword. And, interestingly, you list buzzwords on your website as things that you're aware of. My colleague, Tom Holeňák, is a specialist in looking at all the participatory design methods there are around and distilling them into something good, useful, and workable. For you, what is participatory design, and what are the pitfalls that should be avoided?

Dave Griffiths (08:27)

I think for us, we use a lot of open-source and open-hardware approaches. That's so much a part of what we do that we don't actually talk about it very much, but I think it serves as a really good underpinning. Because we don't seek to have exclusive ownership over anything that we're making, and because we're a non-profit interested in fostering communities rather than developing tools to sell and own, people are much happier to work with us and contribute.

This goes not just for the technological side of things, but also when we're working with community groups. When we publish papers, we invite them to be co-authors on the open-access paper. So it's really about ownership. If people have ownership over something, they trust it. All of these methods are great ways of encouraging that, I think.

Eleanor Drage (09:37)

Which totally makes sense. I mean, if you're going to contribute to something, you want to feel like it belongs to you in a certain way. And that's why a lot of people feel like engaging in democratic processes isn't working, because they feel they have no ownership over the result, or over the state of the country once they've cast their vote, and nothing happens, nothing changes, and they have no control.

But you've come up with an incredible project called the Sonic Kayaks. Can you tell us what a Sonic Kayak is and what you built them for?

Dave Griffiths (10:14)

So, Sonic Kayaks started right back at the beginning as an arts project with Kaffe Matthews, where we were actually putting sound onto bikes. We were messing about with GPS and putting really big speakers on the handlebars. It's still a project that she runs in different cities around the world, layering sounds over different urban locations so you can cycle around and explore them. The maps of the city were like sonic scores that you could navigate on a bike.

We are based in Cornwall, and we were always talking about trying it out here because I was testing the system locally while building it. But Cornwall is a terrible place for cycling; it's just really steep hills, narrow roads, and fast drivers. So we thought maybe we could try this with kayaks instead.

When I worked with Amber on this project, she noted that from a scientist's perspective, it was only really going to work if we could turn it into a citizen science project as well. There's a huge amount of information you can gather when you're out on a kayak. A Sonic Kayak doesn't just have GPS and sound; it also has a temperature sensor, a turbidity sensor, and we've even put air pollution sensors on them. We take all of this information, turn it into sound, and play it out of waterproof Bluetooth speakers on the kayak.

Essentially, a Sonic Kayak extends your senses under the water so you can hear what's going on as you're paddling around. You can actually hear things like temperature gradients happening in real time. I've also added hydrophones so you can hear any underwater life forms making sound beneath your kayak, or approaching boats with engines, which is a bit more common.

Eleanor Drage (12:19)

And who were they for?

Dave Griffiths (12:21)

Originally, they were for kayakers who wanted to collect data on the marine environment. It's interesting because we don't actually have a lot of data on tidal estuaries, places that are incredibly important for fish spawning, especially in the context of climate change. We have a lot of information further out to sea, where there is satellite coverage, but estuaries are quite difficult to monitor, so researchers were looking for ways to gather data from those locations.

So yes, we worked with people who were into kayaking and wanted to collect data to understand the world better. Then we encountered another group who became interested while we were testing it out to see what data we could get. A local kayak group called Access Lizard Adventure, which works with people experiencing sight loss, was looking for ways to give people more confidence on the water so they could kayak around on their own. They saw that we were working with sensors and sounds and realized this would work incredibly well for the people they were supporting. That gave us a very specific audience to target.

Eleanor Drage (13:39)

And these are blind people, right? I love what you were saying before that people don't always want to go out in tandem kayaks. People want to go out on their own and experience being out on the water, where it's just you, the water, and the soundscape. We all have a right to experience that feeling of solitude, quiet, and being by ourselves. So I love this. Yeah, go on.

Dave Griffiths (14:05)

This was something we had to work really hard on. When we were doing this with bikes in cities, we had to battle traffic and play lots of loud sounds, so you could kind of get away with anything. But when you're on a kayak, the very first feedback we got from testers was actually, "We want less sound." They wanted it to go completely silent when nothing was changing, so they could just stop and listen to their own thoughts, and only have the sounds change when they paddled or when an environmental reading shifted.

Eleanor Drage (14:33)

Yeah, I guess there's a temptation to feel like you have to do more when you create a soundscape to ensure there's always something to tune into. But really, it's about the silence. And the ability to enjoy that silence is only possible if you know you're not about to hit a rock or head in the wrong direction.

You also had another project called Nurgle, named after the Mesopotamian god of disease. Tell us about that.

Dave Griffiths (15:11)

This project was designed to explore the relationship between disease and social networks. Despite having just gone through a global pandemic, that is actually a very difficult thing to study because of ethics; you obviously can't take a population, track everyone's friend groups, and then introduce a disease.

So, we worked with Matthew Silk at the University of Edinburgh and decided to build a computer game where we could emulate some of these behaviors. We didn't need to map it out in exhaustive real-world detail; we just needed enough data generated from people playing a simple game. In a game environment, you can introduce a disease and track the social network exactly.

However, we had to be quite careful when designing the game world. Because the topic is so serious, we needed a light touch, so we used hand-drawn characters and a nice, familiar world that wasn't too realistic. It's a world where you can wander around, talk to plants, chat with friends or strangers, make new connections, and just explore.

Then, at some point, a disease is introduced, and characters start sneezing. We are specifically interested in how your behavior changes. Some players immediately unfriend everyone and go off to talk to the plants by themselves, while others try to make as many friends as possible because that allows them to gather a lot of food. There's a trade-off to it. We all experienced that during the pandemic, you could completely self-isolate, but that came with a social cost, or you could form a bubble with someone to maintain a connection. There is always a trade-off in how we behave. And how we manage those trade-offs.

Eleanor Drage (17:13)

Yeah, absolutely. You mentioned that this was about collecting data on how disease spreads through social networks, both for sighted and non-sighted people. What is the relationship there?

Dave Griffiths (17:15)

So, we worked with the same people at iSight Cornwall, the charity we collaborated with on the kayaking project. We contacted them again because we wanted to see if we could make a game like this work for people without sight.

There are plenty of citizen science projects and public health policies out there, but as far as we are aware, very few of them truly consider disability. Disabled individuals tend to be a forgotten group in a lot of public health policy.

We hosted a workshop with participants from iSight Cornwall and asked them about a range of things, how they play computer games, what matters to them in terms of audio cues, and how they navigate text-to-speech tools. We also talked about their experiences during the pandemic, such as how incredibly difficult it is to social distance when you can't see anyone. The reactions they received from the public because they couldn't perfectly social distance led to a lot of added isolation, making it a very challenging time.

Through this, we learned fascinating things, like how some users turn their text-to-speech software up to maximum speed, and how you train your brain to hear and comprehend it at that pace just by doing it. We also discussed how, even though most games aren't designed to be accessible, users find ways to hack and manipulate them using different tools. In our second workshop, we took that feedback, built a working prototype based on their input, and ran a playtesting session so they could tell us what they thought of it.

Eleanor Drage (19:25)

Which is very cool, because you're working with the ingenuity of how non-sighted or partially sighted people are managing to make these games work for them. Because, you know, they are fun.

Dave Griffiths (19:30)

Yeah, and it makes the game better for everybody as well. I think that's the thing with these sorts of approaches, it's like working with people who are experts at listening, and you can find out how to make that experience better for everyone.

Eleanor Drage (19:48)

Yeah, there's an artist called Carmen Papalia who's blind, and he has lots of different ways of getting people to experience New York in a way that matches his experience as a non-sighted person. One of the things he gets people to do is go onto the High Line in New York and hold the shoulder of the person in front of them. The idea is that you get a more immersive experience of New York anyway, because you realize, "Oh, I'm under a tunnel now because I can hear my voice bouncing back up from the ground." It opens up these different ways of experiencing your senses, which is so amazing to get that insight into the world and enrich your own way of experiencing it, even if you can't see.

So, what was something you learned about how you could make a game more interesting and more sensorial through the input of the people you were working with?

Dave Griffiths (20:26)

Yeah. One of the things we worked on. Actually, I think it came originally from when we were working on the kayaks, which was the realization that you can learn to hear the sound of different trees. The way leaves move is actually different for different trees, and you can identify the tree based on that sound. It was really amazing to walk around with someone who can do that.

And so that was one of the things we put into the game: different sounds for different plants, really filling the world with that. The other thing we did that was really important was creating a button you can press to describe the world around you, which reads it out like a clock face. It says, "At 12 o'clock, there's a plant, one of these specific plants. At 3 o'clock, there's a friend of yours," and so on. It also describes it in terms of paces to show how far away things are. You really get into describing the world in that way, while keeping it quite short as well, so it's not a really lengthy description, but still trying to get the maximum amount of detail in. That was one of the really fun things to put in there.

Eleanor Drage (22:02)

That's very cool. And it speaks to a current trend in interactive design and human-computer interaction of involving plants and biosensors in data displays. If listeners are interested in this, Nassim Parvin at the University of Washington does some really cool projects called Sensing Bodies with these kinds of human/non-human interactions.

I wanted to ask next about big data, and that phase we went through a while back, where big data, or just "more data," was seen as the answer. Kevin Guyan, who is a scholar of queer data and collecting queer data in censuses, has a lot to say about this. But as we've seen with climate change, more data doesn't necessarily result in better policy, because both data and policy are political. So, what kind of environmental data are you collecting, and how are you collecting it?

Dave Griffiths (23:05)

So, one of our projects is actually an offshoot of the Sonic Kayaks project. I mentioned briefly that we were collecting air pollution data for that, and it turned out to be quite interesting. We developed this separate thread, a project we're just coming to the end of at the moment, where we've been building air pollution sensors, giving them to community groups, and helping them process and collect the data.

Now, this is an area where there are already hundreds of open-source air pollution sensors and lots of existing work we could base this on to make yet another open-hardware air pollution sensor. But our approach, and I suppose this comes down to being aware of the politics of technology, is shaped by the fact that we do a lot of work with local government. We have an understanding of the limitations and the possibilities of working in that space.

What we wanted was an air pollution project where there was actually a tangible chance that we could change local laws based on the data collected. And that can be as simple as collecting the right units, collecting data that is directly comparable to what your local council is collecting. That way, they can look at it and say, "Actually, yeah, I can compare this with what we're measuring, and this is quite bad; we need to do something." A lot of air quality sensors you can buy just give you a rating from 1 to 10, like "bad air quality" or "good air quality." That acts as a measurement to inform you as an individual, but it doesn't really have much utility outside of that.

This is, of course, because calibrating sensors is a really hard thing to do. We've had to calibrate our sensors against local government sensors by putting them in the same place and checking that our data actually matches what they're collecting. That requires having a local government that is quite friendly and willing to let you do that, which we luckily have with the Cornwall Council.

Partly, and shout out to Cornwall Council for this, they're really happy to have a community group working on something like this with their residents. They can help us, and even if people don't see that direct collaboration, it helps raise awareness of the issue. Then, residents can go and talk to their elected officials about it.

Eleanor Drage (25:41)

Incredible. And it's super important that people aren't comparing apples and oranges. Obviously, you need to have the same metrics; otherwise, how can you possibly interpret the information you receive?

Dave Griffiths (25:53)

Yeah, exactly. And it's also just a case of empowering people. We did a trial of this a few years ago, where we gave a local community group some data, just some graphs we plotted showing the specific times at which air pollution was above the WHO limits for a 24-hour exposure. They were able to take that data to a local consultation session about proposed road changes. Afterward, they told us, "It just made us feel like we weren't just a bunch of hippies. We actually had hard data." Suddenly, everybody listened to them, and they were taken much more seriously.

That's really what it's about, elevating the ability for everyday people to actually effect change. I think everybody knows what needs to be done; that isn't really the problem. The challenge is pushing that awareness onto the people who have the power to make a change, which happens when the people voting for them are visibly worried about it.

Eleanor Drage (26:54)

Yeah, it's amazing how powerful statistics are. My colleagues who work with femicide data collection constantly battle against the need to create statistics that will always be inherently insufficient, because they can never fully reflect the reality of femicide in Latin America. But they find that collecting stories alongside those statistics is key...

Stories are also things that can be passed on and speak to an issue more qualitatively. I guess, you know, in Cornwall, when people experience pollution, they tell stories about it. I don't know exactly what people might say, but having the data to show the reality is so important, especially for a region like Cornwall, which I think is generally assumed by the rest of the UK to be this seaside-y, outdoorsy, countryside, healthy, non-polluted area. Their data is vital.

Dave Griffiths (27:50)

Yes, exactly. It's just that on busy roads in these places, it is just as bad as it is in cities; it's just super, super localized. I think the same is true in cities, a lot of places aren't as bad as you probably think, while other specific spots are real hotspots for it.

But yeah, in terms of stories, one of the nice things we were doing with this Smogrote project at a school is changing how the sensors look. A lot of air quality sensors are kind of hidden away; they're built into gray boxes and concealed so they don't get stolen, I suppose. We decided to go the other way and make them highly visible. We put screens on them that you can read, which explicitly state: "This is an air quality sensor. This is what we're measuring. This is a URL you can visit to find out more." We also included a simple smiley, sad, or neutral face as a kind of executive overview of the live situation.

We found that it was really interesting because just by putting it out there in public, you get groups of people gathering around it, wanting to know more and talking about it. Some people even said they were going to change their cycling route because they hadn't realized a particular road was so bad. People really want to talk about this because a lot of them are genuinely worried.

Eleanor Drage (28:43)

Right.

Dave Griffiths (29:10)

It's just as much about reassuring people when it isn't a problem as it is about raising the alarm when it is. It's a complicated situation.

Eleanor Drage (29:19)

We shouldn't be gathering data in the darkness of labs away from people. We should be doing it in public, where people can see and talk about the findings we have. Data is not as complicated as people might think it is.

As we're running out of time, can you take the last couple of minutes to tell us about some of your work on unraveling the hidden histories of technology? We know a lot about Ada Lovelace, or you should do if you're listening and haven't heard of her, and the Luddite rebellions, and the Jacquard looms, but there's so much we don't know about. You've investigated the Navajo women designing Fairchild semiconductors. What does AI innovation owe to them?

Dave Griffiths (29:53)

So, this is part of a project that began with a researcher we work with, named Ellen Harlizius-Klück, called the Penelope Project, which looked at the history of digital thinking. This goes right back to ancient Greek times, examining how weavers of antiquity were manipulating information and patterns in the same way we do today when we're programming.

This is based on the fact that it's often difficult for people to conceptualize that digital technology didn't just start in the '90s. It didn't start with transistors, and it didn't even start with Jacquard looms and punch cards; it actually goes right back to prehistory. The concept of "digital" is simply a combination of discrete elements rather than continuous ones. The thought process involved with threads, which can only go over or under each other, is the same as voltages that we read as ones or zeros.

A lot of this history is hidden because it's a feminine technology, largely involving people in craft, which we now view purely as craft, which gets sidelined. A great example of this is the research done by Lisa Nakamura on Navajo weavers. Fairchild was one of the first producers of integrated circuits, and they happened to set up their factory on a Navajo reservation. While this may have been motivated by a desire to pay lower wages, they were actually quite proud of the connection and even put the weavers on their catalog covers. The incredible skills these Navajo women used in their traditional weaving were being directly translated into placing transistors on circuits.

And if you look at them, they look like they come from the same place. If you've seen Navajo weavings, they are incredible, and they truly look just like integrated circuits. This is an example of how these foundational ideas flow directly into the digital devices we use today.

That work has continued with Alex McLean, who is doing a Future Leaders Fellowship with us. We're looking at how heritage crafts and digital technology share this common history, and how much modern digital technology can learn from a fuller understanding of creativity and these sorts of feminist approaches. It's a massive area.

Eleanor Drage (32:37)

Where can we find the pictures of this incredible project?

Dave Griffiths (32:42)

So, our organization's website is thentrythis.org. From there, you can find links to all of these different projects.

Eleanor Drage (32:45)

Say that again.

Dave Griffiths (32:48)

It's thentrythis.org.

Eleanor Drage (32:51)

Wonderful. Well, we've begun and ended with this beautiful scene of you and your mother weaving and coding together in the same room, highlighting the proximity between those two things, which is especially relevant today. I don't know whether you've listened to it, but we have an episode coming out just before this one with Betti Matsumoto, who also incorporates weaving and weaving techniques into tech design and coding.

But this is beautiful work. Thank you so much, Dave, for coming on and sharing your stories and your projects with us.

Dave Griffiths (33:26)

Thank you so much for having me on. It's been really fun.