From the recent Decarbonize Software 2023 event, this episode showcases a fireside chat on Responsible AI with Tammy McClellan from Microsoft and Jesse McCrosky from ThoughtWorks. Jesse shares his thoughts and experiences from years of working in the field of Sustainable Tech on the topics of risks, sustainability, and more regarding AI, before answering some questions from the audience. Learn more about our people:Sophie Trinder: LinkedInJesse McCrosky: LinkedIn | WebsiteTammy McClellan: LinkedInFind out more about the GSF:The Green Software Foundation Website Sign up to the Green Software Foundation NewsletterEvents:Decarbonize Software | GSF Resources:AI Transparency in Practice | Mozilla Foundation [04:19] Reducing bias and improving safety in DALL·E 2 | OpenAI [09:31] Responsible AI: Fireside Chat | Decarb 2023 [18:41]If you enjoyed this episode then please either:Follow, rate, and review on Apple PodcastsFollow and rate on SpotifyWatch our videos on The Green Software Foundation YouTube Channel!Connect with us on Twitter, Github and LinkedIn!TRANSCRIPT BELOW:Asim Hussain: Hello, and welcome to Environment Variables brought to you by the Green Software Foundation. In each episode, we discuss the latest news and events surrounding green software. On our show, you can expect candid conversations with top experts in their field who have a passion for how to reduce the greenhouse gas emissions of software.Chris Skipper: Welcome to Environment Variables. Today, we've got another highlight from the recent Decarbonize Software 2023 event. We'll be showcasing the fireside chat on Responsible AI from Jesse McCrosky, Head of Sustainability and Social Change and Principal Data Scientist at ThoughtWorks, and Tammy McClellan, Senior Cloud Solution Architect at Microsoft and Co-Chair of the Community Working Group and Oversight Committee at the Green Software Foundation. They are introduced by our very own Senior Technical Project Manager for open source projects, Sophie Trinder, so it will be her voice that you hear first. So, without further ado, here's the Fireside Chat on Responsible AI.Sophie Trinder: Hi everyone, I'm Sophie, the technical project manager for our open source projects at the Green Software Foundation. Today I'm going to introduce a special fireside chat to our Decarbonize Software event. We're continuing the conversation that began on the 5th of October at our panel on responsible AI. The conversation surrounding responsible AI is dynamic, oscillating between optimism and skepticism. one side, practitioners believe that AI has the potential to drive sustainable development goals, from responsible consumption to waste management and energy conservation. The promise lies in our improvements in measuring software's environmental impacts, and innovation across energy-efficient algorithms, hardware optimizations, and the growing use of renewable energy sources. On the other side, the rapid expansion of AI, particularly large learning language models, and the insatiable demand for this technology, are raising concerns. If left unchecked, the energy consumption and resource utilization associated with AI make many feel like we're endangering a future where software causes zero harmful environmental impacts. To help us explore the path forward, I'm thrilled to introduce Tammy McClellan, Senior Cloud Solution Architect at Microsoft, and Jesse McCrosky, Head of Responsible Tech and Principal Data Scientist at ThoughtWorks. Thanks, both. Take it away.Tammy McClellan: Thanks, Sophie. And a hello to all you sustainability addicts. Jesse, hello. Let's start the question with, how do you see the relationship between responsible AI and sustainability?Jesse McCrosky: Hey Tammy, great question and nice to see you all. So at ThoughtWorks we use a framework that I like which we refer to as the greening of tech and greening by tech and I think this is the best lens through which to view that question. Greening of tech refers to the fact that these systems and especially generative AI as we're talking about now have serious energy consumption, they have serious sustainability issues that need to be tackled.The other side is greening by tech and recognizes the potential that this technology has to actually improve sustainability of other processes, either within or outside of the tech world. And I think what ties these, these two questions together is issues of transparency and information and ensuring that people have the information they need to make the right decisions for our environment.Tammy McClellan: I like that, greening of tech and greening for tech. It's my new mantra now. So how can, uh, we use this to make more sustainable solutions?Jesse McCrosky: So it's a big question. To begin with, I think that I refer to transparency, and when we talk about transparency, a lot of people think that means you share your source code, or you share your model weights, and then you're transparent. Or it means you have to explain the decisions the AI is making, and that's transparency. Transparency is more than that. There's a report I did with the Mozilla Foundation on AI transparency, and we talk about meaningful AI transparency that needs to be legible, auditable, and actionable. And this means that we have to consider the specific stakeholders that the information is being provided to, what are their needs, what are they going to do with this information. So it comes down to the old adage that you can't manage what you can't measure. So for example, in order to support meaningful policy, meaningful regulation, we need to have information about the sustainability characteristics of these systems.Tammy McClellan: So talk to us a little bit about some possible solutions in this area.Jesse McCrosky: Yeah, absolutely. So when we're looking at solutions, especially using the kind of transparency lens, we can think about who is the transparency being provided to. So, for example, we can talk about consumers. And right now, consumers are very excited about ChatGPT or whatever else, Stable Diffusion, DALL-E, and everything like that. It's a lot of fun to play with. And they do not have meaningful information about the carbon implications of that play. So someone was suggesting to me that ChatGPT should have a real-time counter across the top somewhere that's telling you how much carbon have you emitted so far in your session, how many, you know, gallons of water have been consumed, whatever else. And it is not a matter of just shaming people, but it's helping people make the right choices, because there might be applications for which ChatGPT is really worthwhile to use, but there's other times that somebody's just idly playing or something like that, and if they realize the implications of doing, they might make other choices. This becomes more interesting when we talk about communication between, for example, model developers and model deployers. So, for example, if somebody is using the OpenAI APIs in their product, they need to be able to have information about what the implications are of those API calls so they can make good choices in how they build their software.Tammy McClellan: So awareness is key, absolutely. So Jesse, what is the potential for Gen AI to support greater sustainability?Jesse McCrosky: Yeah, it's an exciting question, and I think there is some potential here. There's a case that ThoughtWorks took, it's a couple years back now, I think, in which we worked with a international manufacturing and services company. They were interested in finding solutions to meet their sustainability goals, and they just weren't sure which way to go.They weren't sure, "should we start sourcing our energy from a different place, or using different sorts of transportation, or using different industrial processes or offering different products?" And so what we did for them was built a mathematical model of their operations and their supply chains. once we had that mathematical model, we were able to build a sort of scenario modeling dashboard where we could show them like, "hey, if you switch to delivery trucks that are using electricity instead of gas, this is what happens to your emissions, this is what happens to your bottom line, this is what happens to your customers."And likewise, depending on / considering different product mixes, considering different sourcing, whatever else. So the mathematical model here was not rocket science, to be honest, it was fairly simple stuff. The hard part of this engagement was really understanding the business at the level that we needed to in order to build that model. There were many hours of interviews and poring over notes and internal documents and everything else, as well as actually some basic desk research to determine the necessary carbon emissions factors, that sort of thing. I'm excited at the potential of generative AI to make this sort of process more accessible and more scalable. And I think that we've seen evidence so far that these models do a very good job of looking at these sorts of documents, looking at recordings and interviews, and it may be possible that you could create this model semi automatically with far, far less of the kind of very heavyweight and expensive all sorts of interventions. As well, it was challenging to understand the exchangeability. And so, for example, if the company is buying cotton in one particular country, it might be obvious to us that they can instead buy the same cotton from some other country, and that's the only possible change that could be made. But it's not so simple for the model to figure that sort of thing out automatically.Whereas GenAI, I think when we connect to these sorts of emissions factors databases, has the potential to make this process much easier. Tammy McClellan: Yeah, awesome. Let's move a little bit and talk about risks. How do you think businesses can manage the risks of AI? Jesse McCrosky: Yeah, it's it's a big question. I think everybody's talking about this. And I think what I would say is it's critical to understand that risks must be mitigated, not removed. I think a lot of people are talking, for example, about bias and discrimination, and they say, okay, we're going to produce a model that's perfectly fair and perfectly unbiased, or we're going to eliminate this bias from our model or whatever else. And this is just not the way things work. We live in the real world, and these systems are based on data from the real world. And the real world is unjust, and so we need to be able to be ready to tackle that. So, one example that I like is OpenAI with their DALL-E interface generation system. For a while, maybe some months ago, I think, if you asked it for pictures of lawyers, it was going to give you eight pictures of white men, basically.And OpenAI recognized that there was a problem there, as did the community, of course. So eventually, OpenAI had a short blog post where they talked about how they were going to fix this. And it was apparently fixed, so when people tried to get pictures, they would see pictures of lawyers, and some of them would be women, and some of them would be of different ethnicities, and everything else. So People were curious how this had been fixed and it turned out that all that OpenAI was doing was just randomly appending words like women or black or Asian or whatever else to these prompts and people were not super impressed with this solution but I think it's an important illustrated example, because it's a mitigation, there was a problem with a model, there was a problem with the data, this is not a problem that can be solved fundamentally, it needed to be mitigated, and they found a way, they said, "here's the harm that's going to come from the system. It's going to not be producing an adequate representation, and we found a way that we can show more representation." So this is the sort of mitigation that companies need to take. So when there's issues, and this is where transparency comes in around the carbon impacts as well, so that they can be mitigated, so that if I'm an engineer sitting in front of my laptop writing some software, I need to have awareness that if I call this Gen AI call or whatever else, I have to understand this is going to spike the carbon emissions of my product, and I need to find another solution.Tammy McClellan: Gotcha. Yeah, that makes sense. Tell me. So are you optimistic or pessimistic about Gen AI at this point?Jesse McCrosky: I think I'm mixed. I think that ultimately solving the climate crisis means simultaneously solving a social crisis. And I think it's very hard to solve climate change without also solving issues of social justice globally. And I think that Gen AI is a tool that might enable some of these conversations to be tackled in a more interesting way.So I think as long as we're mindful and honest and clear eyed about how we apply this technology, there can be some optimism there. We need to ensure that we have adequate transparency so that people understand the carbon implications of the choices they're making when they're using these systems, but given that, there is potential to do better.Tammy McClellan: Gotcha. So I know when you and I chatted before, you said that you had a fun story of AI. Did you want to tell us what that is?Jesse McCrosky: Ah, so actually, I think there's a misunderstanding. The fun story was an expanded version of what I was talking about before, butTammy McClellan: Gotcha. Jesse McCrosky: if we have a moment, I think one thing I want to add when we come back to the idea of how, how transparency can help Gen AI be used more responsibly. So, a lot of people are familiar with the concepts of DevOps or MLOps or CD4ML, these sorts of processes. And I think this is a really critical place for transparency around carbon emissions to be integrated. I think the point I would make is that right now, a software developer that's working in kind of a modern setup has the ability, as they're writing code, to see immediately if the code that they're changing is causing some test to fail, or is causing some performance degradation, or is introducing some bug or whatever else. And I think we need to have the same process for carbon so that it if an engineer is making a choice and for any devs out there, maybe you have a case where you need to use a regular expression, but it seems like too much work to figure it out. "Hey, I can just call a Gen AI model and it'll do it for me as well."It'll work just fine. And you might make that choice because it saves you a couple minutes or whatever. But if you then see that all of a sudden your dashboard turns red and says, okay, your carbon has just increased like 100 percent or whatever, you're going to come back and you're going to revisit that decision. And also your team is going to see that, the trail of what's happening because of what you've done. And so it creates this sort of accountability in the development process.Tammy McClellan: All right. So I'm curious. What are the top three recommendations you would give to people who are interested in reducing carbon emissions of AI?Jesse McCrosky: Good questions. And yeah, I think that's something I didn't really touch on so far, but there are a lot of choices that can be made when applying AI. So we don't need to use the biggest general purpose models for everything. I think that there are cases where a general purpose model is really needed. But um, I think that in most cases, no. And so we can talk about using much simpler application-specific models. We can talk about using a smaller model and fine tuning it for the particular task. There's processes like quantization and distillation that can make models much more carbon efficient and nearly as effective. So investigating these options, and again, I think this kind of hinges on the MLOps setup where you need to be ready to evaluate performance. You need to be able to say "how small can I make this model and still actually meet the requirements in my product." Beyond that, I think it's a matter of providing transparency to the end user. So if you're producing something, if users understand the choices that they're making when they're using that product, there's a lot of different ways this can play out, and this can mean some Gen AI chatbot or something like that, but this also can be, maybe you have an e-commerce product. platform and you're using AI to make recommendations to your users and the recommendations that you make can influence their behavior and it can encourage them to buy more products that are disposable or made in very carbon-intensive ways, and so considering these sorts of externalities as well is really critical.Tammy McClellan: Gotcha. I'm curious, do we have any questions from the audience at this point?Sophie Trinder: Yes, we do. And thank you so much, Tammy and Jesse. It's been a really great session on AI here at Decarb, and it really shows the passion in the industry for these technologies, plus the responsibility that we all must take when it comes to AI. I know we'll be hearing a lot more in the coming months. But yes, we've got a few questions from the audience. I just want to shout out first, Jesse, thanks for the fun story on OpenAI, how they were mitigating the problem with data to show more representation through mitigation. It was a really interesting insight, thanks. So one question from the audience, how important is prompt engineering for improvement of AI efficiency?Jesse McCrosky: Great question, yes, and it's it's really extremely important, because the energy being consumed by the model is going to depend in some complex ways, depending on how many tokens are coming into it, and in quite a direct sense, how many tokens are coming out of it. So if we can reduce the number of tokens going through the system, we reduce the carbon emissions. And this again, I know I'm sounding like a stuck record, but it really depends on the MLOp setup, where we should be able to test and see how short can we make our prompts and still accomplish what we need to do. And this is both the length of the prompt itself and the length of the output. So for example, go back to that example I was talking about where maybe ChatGPT has a little indicator at the top telling you how much carbon has been emitted in your session so far. Maybe if you see that number growing as you're chatting with it, you're going to say, "hey, ChatGPT, please be a little bit more brief with your answers. I don't need the whole kind of colorful language and going on and on about everything." So yes, it's very important.Sophie Trinder: Super interesting. Thank you. We've got another one on training the AI ML model, which obviously takes a huge amount of data and processing, which in turn causes a lot of emissions. How do you think that we could best counterpart the same?Jesse McCrosky: Yeah, good question. And I think that I have an article out where I actually talk about how the comparisons are a little bit overwrought, talking about how training a model is equivalent to driving a car some distance or whatever. I think that, um, the comparison, at least so far, thankfully, is not quite accurate because we have many cars on the vehicle and a relatively small number of models being trained. I think the important thing is to keep it that way. I think the important thing is that we need to encourage use of open models and shared models rather than every single organization in the world trying to train their own LLM. And this is why I would be a strong supporter of open-source models. I think it's nice to see that movement.I think it's potential. It means that organizations, first of all, save their money, but also save their carbon when they want to be able to explore elements in their business. And there's always the potential for fine tuning, for whatever other tools need to be applied to open models to make them suit people's applications.Sophie Trinder: Amazing. Thank you. And jumping back to sort of problems on data and representation, we've got another question centered around that. So do you think we should promote digital humanism and ethical AI to raise awareness about the need for sustainable AI?Jesse McCrosky: Yeah, absolutely. I think we're existing at a moment where responsible AI and such is being discussed everywhere. There's very active regulatory work in many different regions of the world. There's many people in academia, in civil society, and in industry doing this sort of work. And I think that green AI should come along for the ride, so to speak, and it should be an important part of how we think about the risks and the potentials of these models.So, yes.Sophie Trinder: Amazing. Thanks very much.Chris Skipper: So that's all for this episode of Environment Variables. If you liked what you heard, you can actually check out the video version of this on our YouTube channel. Links to that as well as everything that we mentioned can be found in the show notes below. While you're down there, feel free to click follow so you don't miss out on the very latest in the world of sustainable software here on Environment Variables. Bye for now!Asim Hussain: Hey everyone, thanks for listening. Just a reminder to follow Environment Variables on Apple Podcasts, Spotify, Google Podcasts, or wherever you get your podcasts. And please, do leave a rating and review if you like what we're doing. It helps other people discover the show and of course, we want more listeners. To find out more about the Green Software Foundation, please visit greensoftware.foundation Thanks again and see you in the next episode.

Guests Sophie Trinder and Adam Jackson discuss the recent Decarbonize Software 2023 event and the unveiling of the GSF Impact Framework. Topics include driving climate change solutions with AI, high-quality energy data for emission optimizations, and engineering excellence with GSF principles. They also talk about the surge in interest in Green Software Practitioner courses, responsible AI and environmental sustainability, and the benefits of Carbonware SDK and carbon awareness for reducing emissions. They announce upcoming events, including a hackathon focused on the impact framework, and provide information on where to watch talks from Decarb 2023.

Ines Akrap, an experienced web designer specializing in sustainable and energy-efficient websites, joins Chris Adams to discuss the challenges of green coding in frontend development. They explore the nuances of designing energy-efficient websites, common mistakes in optimizing sites for carbon efficiency, and highlight exciting projects in the field of green software. The episode offers practical tips and explores new research horizons in the quest to decarbonize the digital world.

Asim Hussain, Speaker on the Green Software Foundation's newly introduced Impact Framework, discusses the capabilities and objectives of the framework. Project leads join to delve into its applications and potential. The Impact Framework aims to revolutionize the way we assess and mitigate the ecological footprint of software development and use.

This podcast explores the benefits and trade-offs of load shifting, modeling the European electricity grid, 24/7 carbon-free electricity matching, unbundling renewable energy generation, and optimizing energy usage through load shifting.

Anne Faubry and Alexander Dawson from the W3C Community Group discuss the Web Sustainability Guidelines, Content Accessibility Guidelines, and their roles in the group. They talk about the differences between standards and guidelines and what the Web Sustainability Guidelines aim to achieve.

TWiGS host Chris Adams is joined by special guest Tony van Swet from Electricity Maps, to talk about the mapping of the carbon intensity of electricity grid. Tony shares some of the work that Electricity Maps has been doing to make it easier to understand how clean or dirty electricity is around the world, as well as how they’re making this data more accessible and usable to consumers. Join in on this candid conversation discussing the uses of such data and how to access it, as well as Tony talking about carbon intensity, open data, and open source.Learn more about our people:Chris Adams: LinkedIn | GitHub | WebsiteTony van Swet: LinkedIn | WebsiteFind out more about the GSF:The Green Software Foundation Website Sign up to the Green Software Foundation NewsletterNews:How to trace back the origin of electricity (Smoothie Blog Post) | Electricity Maps [06:16]The value of space-time load-shifting flexibility for 24/7 carbon-free electricity procurement | Zenodo (TU Berlin’s Study with Google, using PYPSA) [12:11]Electricity Maps | Client Story: Monta (EV Smart Charging use case) [15:41]GitHub - electricitymaps/electricitymaps-contrib: A real-time visualisation of the CO2 emissions of electricity consumption [21:01]Electricity Maps | Reports - Hourly Residual Mix Methodology [27:13]Resources:Electricity Maps | Data Portal [18:29] Electricity Maps Methodology If you enjoyed this episode then please either:Follow, rate, and review on Apple PodcastsFollow and rate on SpotifyWatch our videos on The Green Software Foundation YouTube Channel!Connect with us on Twitter, Github and LinkedIn!TRANSCRIPT BELOW:Tony van Swet: Looking at Google's use case at their data centers, they have the huge potential to shift their computation based on time or location, so this enables them to manipulate their energy consumption through using our API to increase their consumption when the sun shines and the wind blows. Chris Adams: Hello, and welcome to Environment Variables, brought to you by the Green Software Foundation. On our show, you can expect candid conversations with top experts in their field who have a passion for how to reduce the greenhouse gas emissions of software. I'm your host, Chris Adams. Hello, and welcome to another episode of This Week in Green Software, where we bring you the latest news and updates from the world of sustainable software development. I'm your host, Chris Adams. When we talk about green software, it's often common to talk about energy efficiency, and one of the reasons we care about it at all, is that right now we burn a lot of fossil fuels to generate electricity used in data centers, networks, and end-user devices. But how much of that comes from fossil fuels? And is that changing? This data exists all around the world, and sometimes the data is open, but it's often very messy. In 2017, the Electricity Map project was launched to make it easier to understand how clean or dirty electricity was all around the world. And as the name suggests, it took the form of a map showing the carbon intensity of electricity in as many places around the world as possible. Over the subsequent years, an open source project has grown with hundreds of developers around the world, contributing open web scrapers for data in their parts of the world to make the data more accessible. And earlier this year, the company behind the project released a new open data portal for historical data about how clean electricity was for anyone to use how they wish. So, what does this have to do with green software? Having access to this kind of data makes it much easier to understand the carbon footprint of your software. And this week, we're joined by Tony van Swet from Electricity Maps to talk about carbon intensity, open data, and open source. Hey there, Tony.Tony van Swet: Hi, great to be here.Chris Adams: Okay, Tony, before I get ahead of myself, I think we should give you a bit of space to introduce yourself properly. So can you tell us a little bit about what you do at Electricity Maps? And for folks new to the field, what Electricity Maps does these days, please?Tony van Swet: Yeah, of course. I'm a senior software engineer in the advocacy team at Electricity Maps, and I'll give you a bit of background on what we do at Electricity Maps. So our mission is to organize the world's electricity data to drive the transition to a truly decarbonized electricity system. And as part of the advocacy team, my focus is enabling climate action with transparent insights.We do this with the help of the open source community, building products such as our map visualization and the data portal that we're here to talk about today.Chris Adams: Cool. Thanks for that, Tony. Okay. So if you're new to this podcast, um, my name is Chris Adams, as I mentioned before. Um, I work as the executive director at the Green Web Foundation, a Dutch nonprofit focused on an entirely fossil-free internet. And I also work as the chair of the policy working group inside the Green Software Foundation. And before we dive in, here's just a quick reminder, everything we talk about, we'll link to in the show notes below. So if there's a paper that caught your interest, or there's a story you've heard about, we'll do everything we can to make sure there's a helpful set of links that you can follow up for your own research a little bit later. But back to Tony. Tony, I've got to have to ask you, I know you're working in Denmark, but... I suspect you might not be coming from Denmark in the first place. What does a Kiwi end up doing on the opposite side of the world in Denmark, working for a company like Electricity Maps? I'm sure there's a story behind that.Tony van Swet: Yeah, absolutely. It's definitely a bit of a career shift for me. So I started out about 10 years ago as a truck driver in New Zealand. I was full of self doubt, a bit depressed, struggling to find my place in the world. And to lift myself up out of this, I made it my mission to create technology to combat climate change, and I identified that software was the most powerful way to effect change at scale.And this led me to enroll in a computer science degree. From there, I worked at a few cool startups in New Zealand, eventually looking to integrate electricity maps data when I saw their job postings and applied, and within a few months, I had the job and was waving goodbye to my friends and family to fly across the world to Denmark.It's definitely been full of challenges, but it's been amazing to find a company that really shares my values and aligns so perfectly with my mission.Chris Adams: Wow. So you, when you say you're a truck driver, you're talking about the massive, like 18 wheelers crossing from city to city, right? Something like that.Tony van Swet: Yeah, I actually worked with the HIAB trucks, which have a crane on the back. So I was delivering building supplies around Auckland. It definitely gave me a lot of time to think about the world and take in the kind of sights and sounds of the city.Chris Adams: Wow. Okay. So I think you may be the first former truck driver we've ever had onto this podcast. So yeah. Wow. Thank you for, thank you for coming along. That's also a fun story. I, it's, it's quite nice to hear something like that because, uh, I myself, there's a lot of us who are self taught technologists and to hear a nice story about switching careers you're going, "that's cool, actually." All right, before we digress, let's go back to what we were supposed to be here talking about, which is open data and carbon Intensity. So one thing you mentioned is that we're here to talk about open data and there's some recent work at your end that's made that visible. But before we do that, could we briefly just cover what carbon intensity means at your end, because this is something that isn't obvious to most people.And I remember there being a kind of nice introduction on your website using metaphors like blenders and so on to explain that there's more to electricity to it being just gray versus green, for example. So maybe you could just. provide a bit of a background or how you explain this to people, then we can dive into some of the details about open data.Tony van Swet: Yeah, the blender analogy is really great. We even did a smoothie maps version of our app for April fools, renaming all of the power sources to different fruits and vegetables to illustrate that. So yeah, carbon intensity to us seems like it's relatively straightforward, but if you're not familiar with this idea, it's quite hard to understand.And in this case, we refer to carbon intensity as the CO2 equivalent for a given zone where energy is being consumed. We calculate this by determining the carbon intensity for each generation type and then weigh it according to its proportion of the grid mix. We also then calculate the neighboring zones and account for all the imports and exports of the connected zones to figure out a final number for the carbon intensity where you plug into the wall and consume it.Chris Adams: So basically, if I understand that correctly, you're, what you're saying is you look at all the various parts of the world, and when you say zone, you're referring to maybe a country or a part of a country, depending on how a grid is designed. And then when you're talking about the kinds of generation, you're talking about, say a coal fired power plant or a gas fired power plant or a solar farm or something like this. So these have different levels of CO2 that get emitted for each unit of electricity and you're mixing those together, something like that. Is that correct?Tony van Swet: Yeah, definitely. When we take a look at a coal plant, it's going to emit a lot more carbon than the equivalent solar or wind farm.Chris Adams: Okay, cool. So that talks about the consumption, the how, where the electricity comes from. So maybe we can talk a little bit about, okay, how we experienced that and how, like, when I plug something into the wall, for example, what happens next?Tony van Swet: Yeah, so when we, um, plug into the wall, the energy we consume is, um, considered a mix of all the generation types of the grid you're connected to, um, and it's almost impossible to determine whether an electron comes from a wind farm or a coal plant, even though this will have a significant change in the carbon intensity of the energy you consume.So this is where it's really useful to consider the grid as a giant blender, mixing together all those generation types. And then we can evaluate the true carbon intensity of the energy that you consume.Chris Adams: Okay. If we're going to continue this blender analogy, if you put lots and lots of, say, strawberries in a blender, it's going to look one color. And if you had lots of Kiwi fruit in a blender, it's going to look another color. So that's a little bit like what you expose to people and how that might change over theday.Right. Okay, cool. I believe what we'll do is we'll share a link to the blog post, becauseI found it one of the clearest ways to actually help people get their head around this kind of concept, because it is a bit of a leap when you're first starting to get into this field. So with that, we've got a kind of grounding there. Maybe it's worth talking about this from the point of view of a software engineer. So. Let's say you do know this and you have access to this information. Why is this helpful if you're a software engineer? Like where does this fit into what you might do, for example, or affect your job?Tony van Swet: I think it's super useful as a software engineer to, to have this information and I see a few main categories where you can apply this data, particularly around raising awareness of when to consume energy. We want people to use power when the sun shines and the wind blows. So I think that there are ways to present this information so people can make decisions in their everyday lives.But particularly for me, I find it interesting of automating solutions where we can get carbon-aware products that will shift their consumption or the load of the power consumed based on how sustainable the power is available to them.Chris Adams: Okay. So in this case, this scenario here, you're basically saying, if you have an abundance of power, which is very green, you might kind of tune or change your usage to use more of that, and when the power is particularly dirty, for example, you would try to use less of it so that you're shifting your power through time or possibly through space so that the average carbon intensity might be lower than it otherwise would be.That's what I think you're saying, right?Tony van Swet: Yeah, exactly. So the two main ways to optimize your consumption here is over time or via location. Um, so we know that different grids are much cleaner and, um, some people have the luxury to be able to shift their consumption via location as well.Chris Adams: Okay, cool. So. We've got the kind of general concept for this. Are there any kind of favorite examples that you might point people to of people using this to actually demonstrate their behavior, either at a personal level or an organizational level? Because yeah, having a concrete example would be really helpful for people who are listening to this for the first time.Tony van Swet: Yeah, I think, um, my favorite example is, um, looking at Google's use case at their data centers. They have the huge potential to shift their computation based on time or location. So this enables them to manipulate their energy consumption through using our API to increase their consumption when the sun shines and the wind blows.Chris Adams: Okay, so if I understand it correctly, they're like a client of yours or a customer of yours, they pay for this, and then they use it then to essentially either scale things up or down, depending on the amount of power they might be using, depending on where the data centers are. That's, that's what it sounds like, what you're suggesting there, correct?Tony van Swet: Yeah, definitely. So there is the location aspect and we see a huge variation of the carbon intensity throughout the day. So they also do time-based or scheduled computation based on the carbon intensity available to them.Chris Adams: Cool. Okay. I'm glad you mentioned this because this is something we've had people come on the show before to talk about some of this, but since we have spoken about this, there's actually, uh, some interesting data. Uh, there was a study published with TU Berlin where we're, I'm based in Berlin so we've, I found out about this study and, uh, there's. I found this actually quite a nice example of this to talk about, because a lot of the time, when you see companies talking about this, it's quite hard to actually find meaningful numbers to say, does this actually translate to a saving in carbon? Or does it translate to a saving in even money, for example? And this is the first time I've seen with really detailed information, which has been modeled through this. Um, we'll share a link to this paper, but there's a few kind of headlines that I saw from this. And as I understand it, one thing that Google is doing, for example, they've basically set a commitment to say, "we want to have the average carbon intensity of our power to be this much." So we want to have a certain percentage coming from what they call is like carbon-free or fossil-free sources of generation. And, uh, the study that I saw basically showed that by moving the load around, it reduces the amount of renewable energy, renewable kind of generation that needs to be deployed in the first place for this.So there's an embodied carbon saving in the, in not needing to have a bunch of wind turbines or solar all around the globe. And this study that I see, it was modeling five data centers. So five out of say 14 data centers that are around there. And there were. The savings are pretty good, or actually like measurable.I think with the combination of moving things through time and moving things through space, so moving a compute load to where it was going to be greener, the figures that I saw, some of the headlines were that they're able to reduce the cost of doing this by something in the region of a third of the amount of investment that would need to be possible. And, uh, they also, this is one of the first examples I've seen, which even explains like what the costs on a yearly basis might be for this. And, uh, I think the. There was a couple of scenarios inside this. So there's maybe with zero load shifting or moving, say, about 40 percent of the compute loads that to to different parts of the data centers, where maybe one part of the world might be particularly windy or sunny. When I look at the figures here, I see something in the region of, if you, the savings that are here and we need to, and I will share a link to this, to the actual study for this, so that people can look into this a bit more detail, but with the five data centers modeled in, I think, Germany, in Denmark, in Portugal, in Ireland, and in Finland they were basically able to model savings of around at least 200 million US dollars each year by, in terms of the amount of power that you would need to be, the amount of like generation you would need to match this, to actually hit those targets. Now this is, I think this is actually useful to understand because this actually speaks to the fact that there's economic drivers as well as actually just environmental drivers for this. And this kind of speaks to the wider kind of trend, but. I think it's useful to, for this to be, people to be aware that there's actually something in the public domain to interrogate and look at some of these numbers and see how some of these are modeled and what some of the assumptions are. So we spoke about that. Are there any other use cases that you might point to that may be a little bit more closer to home, for example, or something that you might, that people might experience on a more kind of daily basis or close to themselves, for example?Tony van Swet: Yeah, absolutely. Yeah. We have a few customers in the EV smart charging space, and we have also done some research with the Frederiksburg commune here in Denmark about the benefits of smart charging. And we... We were quite impressed to see a 10 to 15 percent reduction in carbon emissions if we have grid-aware smart charging products.So this is plugging your car in the evening and letting it decide when the best time is to charge the car overnight. And even with a small shift in that load, we see a significant reduction in the carbon emissions of the energy consumed. So we were really positive with the results of that. And particularly find it a very nice use case that you put the decision-making power in the hands of the consumers here.So people can choose whether they want to use these products or not.Chris Adams: Okay, cool. All right. If you're in the UK, I believe there's a number of companies that do things like this. Octopus is one of the better known examples of this. And I think under some of the tariffs, there are scenarios where you can actually be paid to charge up a car rather than pay to charge a car or to use a car.So the cost can go negative. Because there's maybe an abundance of power in the grid or like we have here. So that's actually, okay. That's quite useful. So we've covered a couple of use cases now. Maybe it's worth talking a little bit about, little bit about what kind of software supports the use of this data. So I know that at the Green Software Foundation, there's a carbon aware SDK, which is designed to allow people to embed this in some of their software. And where I work at my nonprofit, the Green Web Foundation, we have a library, a Golang library, which is used in a project called Carmado, which is a kind of federated Kubernetes operator. Could you talk a little bit about some of your experiences of what you've seen people use for some of this stuff? For example, maybe you could talk a little bit about some of the pieces of software that you've seen in the wild using some of these tools or using some of this data, for example.Tony van Swet: Yeah, definitely. Firstly, yeah, we're hugely appreciative of the Green Software Foundation and their work to make it easier for developers to use data like this. We do our best to enable developers and hobbyists with our free data through our API. Previously, it was known as CO2 Signal and we've now incorporated that into the Electricity Maps API.And we see lots of amazing tools being built. We see people building dashboards so they can make decisions around which data centers they use. And we do see a big community from Home Assistant also integrating our data. So people can connect their smart homes to become carbon aware and give information on the carbon intensity of their homes.Chris Adams: All right. So we've got some, some stuff like that. And I think we've done a decent job of now establishing what carbon intensity is and how some people might be using it so far. And, uh, we spoke about this idea. There's a, like a free tier, which basically implies that people pay for a data service. But one of the things that we're here to talk about today is open data and this open data portal.And as I understand it, this is your baby, so to speak, right? So maybe you could talk a little bit about, okay. What is this that we, that that's actually gone live because I've got a history with open data, but I suspect it'd be useful for people who are coming to this to understand what this data portal is and why it's useful and what it lets people do, for example.Tony van Swet: Yeah, I was super excited to take the lead on the Data Portal project and really happy to come on the show today to talk about it. Providing free and open data really motivates me. And the Data Portal is a product on our website where anyone can download free carbon intensity data for over 50 countries in hourly, daily, monthly, and yearly for both 2021 and 2022.Chris Adams: Okay. So let me just check if I understand that. So, uh, if people want to start using or experimenting with this data, there's a free tier which you, which folks like yourselves provide. Uh, there's another provider called Watttime that does a, of a free, a free tier. And there's commercial kind of real time feeds from both yourself.And, uh, this part here is this high resolution historical data that has typically been quite hard for people to give access to. And this is openly licensed in the sense that people are free to use this how, however they wish, is that the case or is there any, or maybe we could talk a little bit about the licensing part so people understand how they could use some of this.Tony van Swet: Yes, so we have provided the data free for anyone to use. We particularly look at Carbon Accountants and researchers to use the data. People are welcome to use it under our license, as long as they, if they're building a new product with our data, then they'll be required to open source that new product, but if you're using this data for Carbon Accountant, then you're fine to use it and charge for that accordingly.Chris Adams: Okay, cool. All right. Uh, what we'll do is we'll share a link to the message, to, to the licensing. So people have an understanding for this. So I think when I looked at it was the open database license. So you're able to use it for free in any, in any form, as long as, uh, you're prepared to share under similar terms yourself.That's basically the kind of general approach that I understand for that. And you, you spoke a little bit about there's an intended audience of people who might be carbon accountants or researchers or energy geeks. Can you talk a little bit about how this data gets published in the first place, where it comes from? Because as I understand it, the data can be quite messy to actually put into a kind of API for someone to consume.Tony van Swet: Yeah, yeah. It's, it's a huge challenge to collect all the data. So we have an open source repository full of parsers that collect this data from TSOs and data providers around the world. We have an incredible open source community that helps us to maintain those parsers. We then process this raw data with the kind of smoothie idea that we talked about earlier, run data quality checks on top of the data, and then present it in a way that's easy to navigate and consume.Chris Adams: Okay. So you've used a bit of jargon that I'll need to unpack on there. So you said that you're getting data from a few places and you mentioned a TSO. I'm assuming a TSO is a transmission service operator, like someone who operates part of the grid and they publish information. So that's where some of the data might be coming from.Is that correct?Tony van Swet: Yeah. Yeah. Spot on.Chris Adams: Okay. And one of the challenges is that not every, so maybe I, as I understand it, when I've looked at this data, the data comes out in like grams per kilowatt-hour, what I would typically be paying for, but different places might have different ways of reporting it or different units. Is that the kind of stuff that you, that ends up having to be munged so that there's a kind of clean interface for people to consume?Tony van Swet: Yeah. So the data providers, the TSOs tend to give the data in the format of a energy breakdown. So the various production types, whether it's wind, solar, coal, gas, and we then process this data and apply emission factors. So we add a direct and life cycle emission factors to each of the generation types, and then compute that to give a final carbon intensity number for each zone.Chris Adams: Okay. All right. So if I understand that correctly, you're basically saying we know what this kind of coal power station is likely to be doing for each unit of coal. And because we might have some information about it being an old machine, old one or a younger power station. So you'll have some figures like that, and you essentially run through every single form of generation so that you've got a kind of up-to-date, accurate number for that based on what, what people are doing rather than have to look that up because yeah, it's quite hard to find.So. You've, you've created this data portal. People are able to download it for a set of countries or different parts of the world, and you said that there's data for 2021 and 2022, and this kind of begs the question, what happens next? Is this, is the idea, is the intention to keep having this available on an, on a, on a annual basis so that next year there'll be data for 2023, for example?Tony van Swet: Yeah, absolutely. I listened to your podcast a few weeks ago and I heard you mention that we were looking to raise the bar of energy data available out there. And I really like that term. It's exactly what we want to do. We plan to release new data early 2023. We want to enable carbon accountants to do granular carbon accounting based on our data.And we really hope that providing this data for free gives the industry a push to be more open and transparent around what energy data is available.Chris Adams: Okay, cool. All right. So for the energy nerds here, I, it might be worth just briefly talking about the fact that this currently provides average carbon intensity data. Is that correct? So that's basically the kind of location-based figure. So there, this isn't trying to take into account water or anything to do with market-based figures at present.That's something that might be on the horizon in future. Could you maybe talk a little bit about what things are on the wishlist or what people are asking about What would they like to use in future from here? Because you alluded to some things about, uh, the life cycle intensity of, of, of energy, for example, and there's a whole other set of footprint impacts that people often ask about when they talk about carbon intensity, or even just the environmental impact of the use of electricity in any kind of service.Tony van Swet: Yeah, absolutely. So carbon accountants are most interested in the direct emissions that we provide in this data because they're doing their accounting based on the Scope 2 emissions of a company. Um, we do also provide the life cycle analysis emissions for each zone as well. And this is taking a cradle to grave approach of the emissions.We use the numbers from the IPCC and the,Chris Adams: So IPCC in this case is the Intergovernmental Panel on Climate Change.So that's one thing. And then the UNEC, so I'm guessing it's United Nations.Tony van Swet: United Nations Economic Commission for Europe. Chris Adams: Okay, great. Okay. So, so that's basically the kind of bona fide place where you're taking some of these numbers from. And when you talk about the life cycle emissions there, that means that let's say you're talking about solar or wind, for example. That includes the fact that someone has to make the panels in the first place, and there's going to be some pollution that may be caused there, carbon pollution from making the kind of silicon panels or the turbines.Is that correct? And then the dispose disposal.Tony van Swet: Yeah, exactly. Yeah. And even in the case of nuclear, the lifecycle analysis takes into account the storage and disposal of nuclear waste over hundreds of years and applies the costs of that to a carbon equivalent.Chris Adams: Okay, cool. All right. So we've spoken about carbon and we will talk about carbon a bit more, but. It's very, one thing that has come up when people talk about the environmental impact of digital services, there's this term called carbon tunnel vision, where people only look at the one figure, or the one kind of dimension. Is this actually something that, is this on your wishlist, for example, because we know there's a, say, there's a water impact. People talk a lot about machine learning and AI and tools like that, having a water impact, and there's also an impact from the actual generation, for example. Could you maybe talk about a little bit like that?Is that something that you'd like to be heading towards, or is that on the roadmap, for example, in the, in the long run?Tony van Swet: I think we'd love to take a step back and, and have a broader look at the impacts. We're relatively limited with our capacity, so, so we do focus on what we know and what we're experts in. But I would love to see us work with partners to be able to provide our data alongside other sources to take a bigger picture approach to this.Chris Adams: Okay, cool, Tony. So back to carbon then. You spoke a little bit about working with other providers and I realized just as I was doing some research for this recording, this podcast, there was a new paper that came out from Electricity Maps specifically about, this is a, this is a really nerdy, I'm afraid, residual carbon emissions when you look at the environmental impact of electricity and If I understand it, I'm just going to try and run my understanding by you if I can, and then if you can give me an idea about if it's in the right direction, that'd be really helpful.So, as it stands, electricity maps gives you figures for location-based data. So you look at the carbon intensity of the generation all around the world through, and like dams or wind or solar, uh, you'll look at that part, green energy, they often talk about, say, using green energy in some parts of the world where they've purchased essentially certificates to count electricity as green.And this is a kind of like a market-based approach that people use. And this is the basis that various companies use to say, "we're using a hundred percent green energy," for example. Now, if I understand it, this paper that goes into this and basically says, if you're going to look at the carbon intensity of electricity, you need to account where these certificates that people use, where they're actually being used in various parts, because that's going to have an impact if, because if someone is claiming green energy in Ireland, for example, and they're claiming that on the basis of certificates that were traded from Norway, that's going to have an impact on how green the power might look in Norway compared to Ireland, for example. This is what I think some of the research is that was in this paper. Is that directionally correct? Is that moving in the general direction of correctness for this stuff?Tony van Swet: Yeah, absolutely. Yes. I think a lot of companies are buying renewable energy certificates and it has to be a zero-sum game. So the residual mixed paper that our policy team has just released goes into a huge amount of detail into how you calculate the carbon intensity after you have sold those renewable energy certificates for each zone.Chris Adams: And as I understand it, this is something that's on the roadmap that will be looked at is A, a thing that is unsolved right now, but people are looking to figure out how to incorporate into how they work. I know that in the Green Software Foundation, there's a group called Realtime Cloud, who are working to come up with hourly figures to make it possible to provide this kind of reporting. This seems to be one thing that comes up because when I was looking through this paper just last night, actually, there was a few things which are really eye opening for me. So Ireland and Germany are two large markets in Europe, for example. And as I understand it, something like eight times the certificates were consumed as were issued in Ireland, for example.So this basically means as I understand it, that eight times as much green energy is being claimed as is generated in Ireland. So therefore you've got a bunch of generation in somewhere else in the world that needs to be accounted for when you look at the carbon intensity of say, a place like Iceland or Norway, for example, but the same things seem to be in Germany as well. Germany has something like seven times the certificates consumed as were issued in Germany. So that suggests to me that seven times as much green energy has been claimed as is being generated. So if Germany had to have an entirely green grid, you would need something like a sevenfold increase in order for them to be saying, "yes, we're running entirely on green energy." That seems, this is pretty eyeopening. I'm really glad this is actually out there because I haven't seen this data provided in this resolution, particularly in hourly level before.Tony van Swet: Yeah, I think it's really fascinating and definitely highlights why we need the transparency around this market based approach. And it's very early days, so we are hoping to inform the methodology of how we approach this in the future.Chris Adams: Great. Okay. So what we'll do, we'll share a link to that. The things I've just shared are in the first 10 pages of this paper. It's about 80 to 90 pages long, and it's a really impressive tour de force. So Cyril, I'm really impressed with this work. Really mad props for you to actually get this together. Cyril is the policy lead, Electricity Maps, and he's also on some of these working groups, which is why it really caught my eye. So Tony, we've just spoken a little bit about Open data, different ways of measuring the carbon intensity of electricity here for informing your decisions as a software developer. Is there anything that you would like to draw people's attention to? Any projects or things that you are particularly interested or that you're excited about right now?Tony van Swet: Yeah, absolutely. Yeah. First off, I'd welcome people to jump on and take a look at the data portal and I would appreciate any feedback around that. And. If anyone would like to contribute to our open-source work, we're also always looking for contributors there. To find out more, jump on our website at electricitymaps.com.Chris Adams: And if I understand it correctly, you folks are still, it's still mostly Python scrapers and a kind of React app that you had before. Is that still the case for people?Tony van Swet: Yeah, definitely. Yeah. Python and JavaScript.Chris Adams: Okay. So common languages that people know their way around. Okay. Brilliant. I think that's pretty much it. I've, I've really enjoyed this, actually. Thank you so much for giving us your time and diving into some of the finer points of carbon intensity of electricity and, uh, some of the nerdery there. And, uh, Tony, thank you so much.I've really, I've really enjoyed this. Cheers, mate.Tony van Swet: Yeah. Thank you, Chris. It was really great to be here. I also wanted to say I went digging through our open source repo and found your name on there. So I want to give you a personal thank you for contributing in the past.Chris Adams: Thank you very much. Um, there, I think there'll be more PRs coming in future when I find the time. Okay. Cheers, Tony.Tony van Swet: Amazing. Thank you.Chris Adams: Hey everyone, thanks for listening. Just a reminder to follow Environment Variables on Apple Podcasts, Spotify, Google Podcasts, or wherever you get your podcasts.And please, do leave a rating and review if you like what we're doing, it helps other people discover the show, and of course, we'd love to have more listeners. To find out more about the Green Software Foundation, please visit greensoftware.foundation. That's greensoftware.foundation in any browser.Thanks again, and see you in the next episode! 

Dr. Daniel Schien from the University of Bristol, UK, joins host Chris Adams to discuss digital sustainability. They cover topics such as streaming's environmental impact, the carbon footprint of digital services, and the importance of reducing carbon emissions in ICT. The conversation explores different approaches to measuring carbon intensity and emphasizes the need for long-term decision-making in green software.

Join host Chris Adams and Tereze Gaile, global Sustainability SME at MuleSoft, as they discuss sustainability tools, resources, and bringing sustainability into organizations. Topics include the Green Code initiative, Developer Carbon Dashboard, generating customer demand for sustainability, change models, measuring organizational emissions, and self-care in the sustainability space.

Arne Tarara, CEO of Green Coding Berlin, joins host Chris Adams to discuss Mojo, a new programming language that achieves a performance boost of 68,000 times over Python. They also talk about progress in Grid Forecasting and Apple's carbon neutral Apple Watch. The Wagtail 5.1 project's greening efforts and AVIF encoding are explored, along with initiatives for energy conservation and expanding the measurement of environmental impact in software lifecycle assessment.