(00:01):
Well, I'm in over my head. No one told me trying to keep my footprint ball always harder than I thought it could be. I'm in over my head. What do I really need Trying to save the planet over someone? Please save me. Trying to save the planet over someone. Please save me.

(00:24):
Welcome to you over my head. I'm Michael Bartz. My guest today is Dr. Kelly Widdicks. Kelly is a lecturer in computer science in the School of Computing and communications at Lancaster University. Her work to date has focused on uncovering the environmental impacts of digital technologies as well as the ethical implications for digital wellbeing tools and promoting equality, diversity and inclusion and computer science education. She is most known for her research on video streaming and sustainable design of online services, which has gained international and national media attention, including interviews for BBC Radio five Live and BBC Radio Scotland. Welcome to in Over my head Kelly.

(01:00):
Hi Michael. Thank you very much for having me here.

(01:02):
In talking about our digital life, the idea of sufficiency interested me, we always seem to want more, but I'm wondering if this is sustainable. Your work focuses on this idea, so I'm excited to discuss this with you. For your PhD, you did a study on household data demand. Can you briefly outline the study?

(01:17):
Yeah, so I guess for some context about my PhD was all about basically how the design of internet services kind of encourages time online and this thing called data demand, which basically puts pressure on the underlying internet infrastructure to expand and therefore expansions on that energy consumptions associated with that. And so what I was doing in my PhD was looking at how these internet services are being used and how we might design services in more sustainable ways. And this particular study that I did on household data demand, it was this kind of larger scale studies already out there that lock up network use across different countries, et cetera. What I really wanted to do was kind of understand this nuances around what users do and those designs leave the greater data intensity. So with this household study, I actually installed logging routes into people's homes for a month, which was super intrusive but obviously all participants had consent and everything and this logs what they did in terms of the network impacts as well as I also did qualitative interviews with those participants and they were really just to have this in depth understanding about what they, what they do online and how that impacts those and what designs we might change.

(02:30):
So yeah, interviews with participants before install the logging routers and then kind of took data from those logging routers back to participants to discuss with them about what common trends and things that might have been a bit irregular to understand how they used the internet in their everyday life.

(02:45):
So what exactly generally did you find with these participants in their internet use?

(02:50):
There's quite common kind of activities that people do online that are quite data intensive. For example communication and and social networking and gaming. But the kind of main thing was around watching, so streaming video now, and this actually formed around 72% of the data demand for my participants. And this coincides with that large scale studies that are out there as well. We found that YouTube was quite a large contributed to that demand as well. So about 50% of the streaming that my participants did was around YouTube. So this gave us a big quantitative kind of element. And then looking at those qualitative data interviews we see new ways of which people are watching and streaming was becoming that more common way of watching rather than DVDs and and broadcast media like you'd have in in the past potentially. And also this thing called multi watching which oh, media multitasking and multi watching is what we defined it in the paper was around basically people watching their video streams on their own separately from other household members.

(03:54):
So you might have somebody watching the television and in one room and streaming one type of video and then someone else watching a video on their tablet for example, this could be in the same room even so people listening maybe with their headphones as well. And I actually noticed that this started happening to my family as well when I was doing the study and I was thinking this is strange, while studying it, you know, with football streams or fighting for the remote and the audio on that, but a communications regulator in the UK actually found that people were watching their own content in the same room for about third of UK households. So this was kind of like a big thing that we were seeing. But yeah, so I guess that was kind of one of the things around multi watching and there was also a bit of some of the qualitative data helped them cover that some of the watching people were doing wasn't always meaningful to them. So it might be getting into those kind of binge watching traps or just watching content to kind of fall asleep.

(04:48):
Yeah, I think those things really interest me cuz it's great to use the internet to entertain ourselves or like we're chatting now through the internet, it's a very useful tool but at what point is it not helpful or not sustainable? And so I guess with that, if people were just watching one television, one film together, that would obviously lower the data demand if I'm understanding that correctly.

(05:11):
Yes. So if you watch it kind of one stream, but also I guess there's a wider point here about the emissions associated with devices as well. So if households just went back to have one device, then obviously that would be more sustainable than having all these different technologies created for people to use and then dispose of and change. But I think that's not really the reality of where the sector is heading and and we're, we've all got these mobile devices and it's kind of becomes easier for us to embed this digital connectivity within our everyday lives.

(05:41):
Yeah. And one part of your study I found interesting talking about how you can mitigate that, like you mentioned just watching something to fall asleep or just not really engaging content. What are some ways that you found that you could mitigate that?

(05:56):
Yeah, so there was two things here. So when I was doing this research there was clear obvious links with other research areas on like the impact of technology and our digital wellbeing and our work productivity and there's lots of research around how we can encourage less time online. So there was this element of my work that kind of went down this area of sufficient internet use. So how if we needed to transition users to demand less data or use digital technology in a way that's sufficient, we can meet their needs without having this continuous growth. This was kind of drawing on the benefits of digital technology and spending time away from that how that can bring people's lives, those benefits. But in terms of kind of the main thing about how we can actually mitigate some of the impacts from data demand was really focused on how we can have more pressure on internet service providers to rethink how they design their technology so that the demand less data or consider digital sustainability more generally.

(06:52):
For example, one design that we might remove would be his potential of autoplay. For those that don't know what autoplay is, it's basically starting the next episode or the next video as you finish something that you've been watching. And this just encourages the users to just watch it and it's already started and sometimes you might be like, oh well it's on so I'll just carry on watching this now. Whereas if we kind of remove that feature and and the user actually has to click, then the video that they want to engage with kind of creates like an additional friction that makes people think, oh do I want to do this? Rather than it being just created for you. This is what I mean about that. There needs to be pressure on the internet service providers to change design rather than the users changing how they use technology in that sense. It's easy to just sit here and say oh we should demand less. But lots of people use the technology for their jobs and for entertainment, you know, it was like such a key thing during the pandemic.

(07:49):
Yeah, people just use it in their everyday lives and it's part of our life and it's wonderful in a lot of ways. So you're right, I don't think people can suddenly just shut off the internet cuz they use it a lot. Like just that idea of using it for what it is and not just letting it be a distraction. Because I think one thing that I read if I remember correctly in your study was talking about streaming music through YouTube and how that's more data demanding cuz you're now using video to stream music. So I think there's a little bit of education as well around that is not a bad thing.

(08:20):
Definitely. And I, I think, I mean I, I don't know what it's like now. So from when I did the study my participants would use YouTube for listening to music, their primary action would be listening. They just wanna have that in the background and they're not using that video content so that content is kind of wasted. So I think at the time when I was writing my thesis that it was kind of a paid service for YouTube where you could remove that video and actually listen to the audio only but then you have to pay to be more sustainable in that sense. There was actually some research from the University of Bristol that took this example of digital wasted video content. If users just listen to the audio YouTube videos that this could be kind of a comparable emission reduction to running a data center renewable energy. So there can be quite significant impacts if those designs were catered for for users.

(09:10):
And you've mentioned a few times the larger implications. So what sort of changes need to happen with the data centers and the providers and that larger societal change to reduce our data demand?

(09:24):
So I think in terms of kind has just stepped back a bit and think data demand is kind of one portion of the pie when it comes to considering the carbon footprint of our digital technology sector or the information communication i c t. And so when we think about ICT, that's kind of the networks and the data centers and user devices and and things like artificial intelligence and internet of things and blockchain, you know all these different trends and I C T has this climate impact from the embodied emissions released from the extraction of raw materials required for technology as well as the manufacturing processes and the transport to business or user. There's also the use phase in operational emissions and then the end of life emissions of digital technology is disposed of. So there's quite a big life cycle that digital technology companies need to be thinking about when designing their technologies more sustainably. So in terms of data demand, this idea of reduced data through internet service designers one area and then thinking of this larger ecosystem

(10:27):
With all those different elements, where are most of the emissions coming from? Is it from making the devices or running them?

(10:33):
So me and some colleagues were actually in a research project looking at the climate impacts of the I C T generally and there's actually quite limited studies that estimate the whole sector's emissions and they don't come to a consensus actually either. So the overall estimate lies between 1.8 and 2.8% of global greenhouse gas emissions. And whilst they did have a consensus about some of the categories that they include in that, so they've got data centers, networks and user devices, these were seen as key components of I C T that should be studied but they did disagree in terms of the share of each of those areas about what is the largest portion. So some experts argue that data centers, those were the majority of our emissions lie, others say it's user devices and then others they think it's more level across the three categories. So I don't actually have an answer unfortunately and I think this links quite directly in terms of that we need more data around this in the sector so that we can actually figure out what are the key challenges that we need to be making changes for sustainability.

(11:39):
For sure. And I mean one to 2% of our emissions doesn't sound like very much but I think every industry needs to lower their emissions so why not try to lower ICTs emissions? Would you say that we could just try to target all those different areas or would it be better to focus on one area I guess once we find out where most of the emissions are coming from?

(11:57):
Yeah, definitely. I think it's an element that we need to have this more data first and we need to see transparent accounting because with this research that we did, whilst it was 1.8 to 2.8% of global greenhouse gas emissions when we were critiquing these studies in depth, they did contrast you to the different kind of data sources that they used, the recency of the data that they collected. Also the different decisions about what classes is in scope for ICTs climate impacts. So whether scope one, two or three emissions are included or even what type of digital technologies to include. So whether television for example classes as I C t and so these different calculation measures then mean that we have these variants in calculations and so we really need to make sure that across the digital sector we're using the same measuring and scoping for all of this that's inclusive of all free scopes and technology generally so that we can actually start to develop and understanding about how we're gonna work towards our global carbon targets.

(12:55):
Yeah, no it does seem like a difficult thing to pin down for sure. Cuz even in that study, if I remember correctly, you talked about the truncation effect. What is that?

(13:03):
Yeah, so I should say that Lancaster, we collaborated with small world consulting so they're kind of experts in carbon accounting and we found that with them that these estimates that we could teach, they actually underestimate the true environmental impact of I C T because of those differences that as mentioned but also truncation error as you mentioned. So this is where the full supply chain emissions for ICT are not considered in carbon accounting. So when you actually take this truncation error into account, we found that the real share of global emissions for ICT is actually somewhere between 2.1 and 3.9%, which is kind of comparable to the aviation industry. So it's not a small percentage in that sense.

(13:42):
And in that study you also covered efficiencies, which is something that was, I was curious about cuz I think a lot of people would say that well as technology gets more efficient it's it's not going to have as big of an environmental impact. Is that the case?

(13:56):
In this project we try to really uncover the different debates in this area and there are some researchers out there and research that says that there's been huge efficiency gains and carbon savings for example in data centers despite global compute instances rising significantly. So efficiency has continued to improve as well in in I C T, which is great, but if you look at historical data, the demand for computations and the number of digital devices has actually or per person has actually outpaced those efficiency improvements. So this has resulted in ICTs and you consumption of caral footprint to grow year on year. And so that kind of relation between increasing efficiencies isn't really linked directly to that consumption. And I guess the, what we describe in the work that we've been doing is that this is a pattern explained through what is called a rebound effect. So where ICTs introduction or efficiencies actually can offset any emission savings from those efficiencies or even lead to an increase in emissions overall.

(14:55):
You also mentioned I C T lowering emissions possibly in other industries. Is that part of the the picture as well?

(15:03):
So there's big debate about this actually is, and there are studies out there that estimate that I C T could save quite significant reductions in global emissions through efficiencies in its introduction in other sectors, for example food or transport, this classic example of video conferencing technologies. So we are meeting virtually instead of flying, but if you look at that data, you know, video traffic has increased rapidly and it's highlighted through my studies as well. But also emissions from flights have increased too. So I C T is not substituting flying for example, it's that being added alongside travel. And so there's little evidence in the past that efficiencies have led to this decrease in global emissions and instead they've risen hand in hand of each other. So whilst there are studies and and examples out there how it could support the global greening in that sense, it's kind of risky that I C T would just start to introduce emissions now if we just assumed it would rather, it's likely that ICT is creating those kind of rebound effects in other sectors.

(16:05):
So it's leading to emissions growth in those sectors that it delivers efficiency gains. So we need more evidence on this if there are actually solutions out there that help us reduce our emissions and that's great, we just need the ICTs own emissions to be accounted for as well and having that clearly and transparently evidence for the sector. And I think unless it's kind of important to remember that not all ICT is used for the good of the environment of society as well. So whilst there might be options for this, there's lots of tech that will be useful for things as well.

(16:37):
Ok, that's super interesting. So it sounds to me like, like when you did your PhD you were looking at personal use and sustainability, but it sounds like you've broadened things out to how do we change the industry and how do we accurately find out what the emissions are for I C T. So do you feel like your, your scope has broadened as you've been working through your career?

(16:56):
Yeah, definitely. So my PhD was quite niche, as you said, on user devices or user service design. These things are still needed, you know, it's really important that we have these in depth understandings but also having that wider understanding about who actually has agency at this point. So whilst I suggested design changes, it's quite often when I talk about this work then the people focus on that consumer change, but really it's, they don't really have a lot of agency about what we do and and what we control and that's where the designers of these technologies come in.

(17:27):
I mean the show is about empowering citizens to take action on the climate crisis. Is there anything that people can do on the individual level to affect ICTs emissions?

(17:36):
Yeah, so there are small things we can, we can all do. The key one is can keep in devices for as long as possible so that we are not got the embodied emissions and disposal emissions from changing our technologies all the time. Then things like, you know, streaming video in lower resolution, we've mentioned streaming and listening to music rather than maybe listening through video. And I think a really key one is don't replace the use of technology with a more carbon intensive activity because I think it's clear to remember the bigger picture in terms of technology. It's a smaller portion of our emissions than say for example going out and and driving. So, but I think the key thing that users could do is kind of lobby technology companies for change so that it's not bond consumers. So asking them to be more honest and open on their accounting, making sure that they're using real green renewable energy and, and not just offsetting and yeah, getting them to design sustainability into technology. So making it more transparent to users about what they're buying in terms of their environmental footprint of technology and things like the right to repair technology as well so that we don't have to completely get rid of our technology once one part has been made obsolete.

(18:48):
Yeah, I get that's very empowering that people could have that impact on I C T and that users can hopefully have a voice and get technology that they want that is more sustainable, that lasts longer and they're able to use the internet as they need to and not have to maybe worry so much about is this affecting emissions and should I be using it less? So yeah, it's been really an interesting conversation. Kelly, thanks so much for coming on the show.

(19:11):
Thanks so much for having me. It's been, yeah, really interesting conversation, so thank you.

(19:16):
Well that was my conversation with Kelly. I found it interesting to talk about how people are using technology in their homes like multitasking, but also that larger conversation about how can we affect I C T and it's not necessarily about always limiting our internet use, although that can be good for us personally. It's really about that larger conversation. How can we get those industries to improve so that I c t can lower their mission. So to me that was really helpful. Well that's all for me. I'm Michael Bartz. Here's the feeling a little less even over our head when it comes to saving the planet. We'll see you again soon. In Over My Head was produced and hosted by Michael Bartz original theme song by Gabriel Thaine. If you would like to get in touch with us, email info@inovermyheadpodcast.com. Special thanks to Telus STORYHIVE for making this show possible.

(20:02):
I'm trying and save the planet. Oh, will someone please save me?