Ryan Panchadsaram, co-author of Speed and Scale, sits down on the podcast to explore the role of developers in combating climate change, how efficient coding practices help lower emissions, and how developers can contribute to open-source projects through GitHub’s Climate Action Plan that help drive forward sustainable technologies.
The Speed & Scale Plan is a global initiative to move leaders to act on the climate crisis, tracking progress towards net zero by 2050.
Explore Speed & Scale’s resources for combating climate change, and connect with Ryan on LinkedIn and X.
Congratulations to Stellar Answer badge winner Christian C. Salvadó who won the badge for answering the question What's a quick way to comment/uncomment lines in Vim?.
[intro music plays]
Ryan Donovan: Hello everyone and welcome to the Stack Overflow Podcast, a place to talk all things software and technology. I am your humble host, Ryan Donovan, and today we're talking about what developers can do about climate change. My guest today is Ryan Panchadsaram, who is the co-author of the book Speed and Scale. Welcome to the show, Ryan.
Ryan Panchadsaram: Thank you for having me, Ryan.
RD: Yeah, so at the beginning of the show, we like to get a little bit of the guest origin story. How did you get into software and technology?
RP: That goes way back. You know, I always have to thank my dad in many ways, right? He always had a computer in the house growing up, so I was always there and my brother and I got really into it. We were really into the early, call it, Microsoft Access and Visual Basic in the early days 'cause that's what my dad was using. I think growing up I got really into Macromedia director and that kind of led down that path and I very vividly remember that bringing things closer to Flash and then ultimately to Fireworks and then realizing I actually quite enjoy the user experience side of product development and creation. And so I went that route. My brother went more, call it hard code hardcore backend engineering. And you know, the two of us make actually a pretty good duo, but we haven't worked on a project in a long while.
RD: You came across my inbox through a partnership with GitHub, with this, was it Clean Developer Initiative? What's the title?
RP: It's the Climate Action Plan. It's actually the climate Action Plan for developers to be even more specific. And you know, this came about really being inspired by the folks at GitHub. Paul Young, who's the Chief Sustainability Officer over there, saw the work that we were doing on Speed and Scale, which was mapping out solutions to the climate crisis. And he clearly saw the biggest circle, which is how do we decarbonize the grid? It was a pretty spot on place for engineers and developers to be able to help out on. And then from there, very quickly realizing actually there are repositories and activities happening that cut at the perfect intersection of engineers and decarbonization, not just across the grid, but for the transportation industry and other places. And so it's kind of amazing. Brian, there's 60,000 repositories on GitHub that have some sort of relation to climate or emissions and tracking and it's pretty cool. It's a really interesting project.
RD: So 60,000 repositories. How does code affect climate change? How does code reduce climate change?
RP: There's many ways to look at it. First, it's like directly, how can code reduce emissions? And when you think about how, when we program something and we have it run on our computers or run in the cloud, it uses energy. And so you can think about it very tactically, well, how can I do that more efficiently? How can I distribute my load in places where, well, that grid's running on more clean energy. And so that's kind of a very tactical way to think about as an engineer or developer, how can I directly impact emissions from the projects that I do. But then when you step out of that circle, Ryan, there actually is a world of projects out there that are helping with call it the 'push to decarbonize. One of the really awesome projects that was featured as part of this action plan is electricity maps, and here you have a wonderful API that tells you the carbon intensity and energy mix around the world, but more specifically where you're running projects. And so it's kind of neat. You can tap into those APIs, but they also have an open source project that allows you as an open source contributor to help them map more grids around the world. And so you can see how the French grid, how data comes from there and you can help improve how that happens and so forth.
RD: Seems like developers are getting a greater awareness of the climate impacts. One of the former co-hosts of this podcast, Paul Ford, he left to work on climate change initiatives, and we've talked to folks doing fusion reactors and stuff to reduce it. Is there one simple thing that every developer today can do to sort of minimize the impact of their code on the climate?
RP: Ooh, great question. Just think about efficiency, right? If you can run your code more efficiently, you are using less energy. And if you have to use compute, how can you put it in places that are cleaner? I mean, that is truly Ryan, the most direct way that you can do this. I'd even pull on that question and say, well, you know, maybe the project you're working on is running actually quite well, and if you have extra cycles, you know, take a look at the Climate Action Plan as well as the GitHub's Green Software directory. There's a lot of really interesting projects out there that are looking for open source contributors. And so a few more other things to kind of share as well. It's like when you think about how we look at the world and measure emissions and measure the effects of it, they're just a wonderful set of repositories out there that you can jump into. When you think about the policy side, there's all these different models that model, ‘what if the US enacts this or China does that, or Europe does this?’, and these different groups put their models out on GitHub for people to look at and use, but also to help improve. And so you could be an active direct participant when you think about your code, or you can be, call it, a broader participant and be an open source contributor to call it the movement.
RD: Yeah. Well, you talk about being more efficient with your code, like do you mean things like refactoring or using serverless technologies or is it sort of more broad, like just trying to minimize the compute?
RP: You know, there are a few good libraries that are as part of the guide that you actually can run on your code and it can tell you different ways and techniques of how to be more efficient or use different compute clusters that are in better places. But it really comes down to that. I mean, I think that's kind of an interesting motivator because maybe let's say someone listening doesn't quite prioritize, call it climate or emissions reduction. I think what that person probably would care a lot about is, well, how much am I spending on compute? Like if I'm spending a lot on compute, I probably want to cut that down. And so what's neat is that if you're even pulling on that lever, you're gonna be likely using less energy. And then when you think about these grids that are operating on lots of renewables and lots of excess solar and wind, you can imagine the compute prices on those grids becoming lower over time. Even if you just have a cost motive in this, like the neat thing about that is that likely means you're gonna try to use less energy to move your load in places that meet that target.
RD: Right. Having a greener electricity system is gonna be cost effective in the long run.
RP: Absolutely. In the long run, without a doubt. I mean, especially, you know, Ryan, we're living in a very interesting time with the growth of data centers. Just changing the relationship that we've all had with our grid for the past, you know, call it decade or two. You know, I live in the US. I think I'm on the west coast, you live on the East Coast, and call it, before this AI wave, the grid was kind of stable, like very few additions every year, you know, you look at that line, it looked kind of flat with a tiny little increase. But what we're seeing now with just the sheer demand from data centers and electrification and, and, and, right. You are actually starting to see a slope that's putting, call it pressures, on our grid and what that means is unfortunately, you have natural gas projects that are now staying on longer or coal projects that are staying on longer, while more solar and wind is being added to it.
RD: And it does seem like a lot of these initiatives are coming from corporations, industries, right? You partnered with GitHub. I've talked to the AWS Climate Center and all these folks are trying to figure out how to get more sustainable energy. In part, I would say because the US has not done its fair share of regulation, do you think that industry is gonna be the best path forward or do we need a greater sort of state interest in this?
RP: I think both have a pretty important role to play. I think what's special about all the names you mentioned, Microsoft, Amazon, and you could put Facebook in this category as well too, and Google as well. When they're building out their grids, they have a very strong climate commitment, right? Most of those organization, all those organizations, have a 2030 net zero target. This is them saying we wanna be by 2030, which is like five years from now, emitting net zero. So we're emitting less than, sorry, netting out what we emit with reducing carbon emissions, but also then buying carbon removal and pieces like that. And what's important about that is those companies, as they're building out their data centers, are trying to get more clean energy added to the grid. They're going out of their way, spending more, financing, doing these things called PPAs, which are power purchase agreements. And that's really exciting 'cause it's the private sector really leading the way in this net clean energy additions. If you're right on the policy side, that is a lever that can be pulled on. But I would say we're in this, you know, ‘reset’ moment where you actually had all this incredible investment from the federal government during the Biden administration on deploying solar, deploying wind, getting these incredibly new clean energy projects to be put across the country in all 50 states. You know, when you look at the money that was going out, it was like it was going to more red states than blue states. 'cause you had manufacturing jobs and supply chains out there. And so that part is getting a bit of a reset in the US unfortunately, that step back. Policy is something that can happen at the state level and local level. So while maybe the federal government isn't supporting it at the level of magnitude, it doesn't mean a state has to stop. But also, I should say to anyone listening, a lot of the pieces of the Inflation Reduction Act and others haven't been completely rolled back yet. And so there's still time to say, Hey, this is important to me. I want clean energy on my grid. I want that new battery project or fusion project in my backyard. Right?
RD: Yeah. The new technologies that have been developed, AI being very prominent among them, it's a hungry compute user. Very, very ravenous. Do you know if there's best practices for minimizing that compute use for AI?
RP: The most direct way you can approach the Hungry Compute is just deciding when and where you run it, and so that's a big important piece. I think then the next piece is that you get to really figure out is how efficient can you make it, right? I mean, you know, back at the start of the year when Deep Seek came out, it was this sort of reminder that, hey, things can be done more efficiently. So I think this, you know, maybe call it the earthquake that happened at the start was you would always see, in order for these models to get smarter, you needed to spend more and more on compute. And that sort of got reset in some way that ‘hey, you can do more with less’. But as we all know, sometimes when that happens, you just, you more of it. And so I would say here in the world of AI, it's like really just as you are consuming compute, you know, run it on a grid, run it in a data center that's using clean energy as much as you can. If you are a project or a company that's like building out data centers, really prioritizing the addition of solar, wind, batteries. You know, you can see a lot of the big tech companies buying ahead of time, like geothermal, Alphabet with [inaudible] or Microsoft with like a nuclear project. Like this stuff is really catalytic. I would say that tech companies are catalyzing so many neat, innovative things on the clean tech front.
RD: Right? I mean, you know, necessity is the mother of invention, right? They need all this energy -
RP: They need it.
RD: Yeah. You mentioned nuclear, do you think that's sort of a semi green technology? I think it's considered renewable, but also, you know, it can destroy an area if it goes wrong. Do you think technologies like that and fusion will have a place in the sort of green energy environment?
RP: I absolutely do. Like, you know, and speaking on behalf of the Speed and Scale Plan and the Speed and Scale team, right? We put together Ryan this way to go from that 59 billion tons of emissions down to zero. And the biggest category of emissions that we have today is on the grid, and then it's transportation, and then industry and then nature, right? It's like it goes down that little curve and you have to find solutions. And so for the folks listening out there, I'm sure a lot of your audience knows about objectives and key results, you know, OKRs. And so we applied OKRs to this climate crisis. And on the grid, you know, we really need to get to a place in space where you've got a lot of clean, durable energy. And that includes not just the solar and winds, it includes geothermal, it includes nuclear. You know, nuclear fission right now, you have advanced projects that are happening outside of the US, but in the US we have a handful of nuclear reactors that we need to keep online. Like you saw what happened in the Diablo Canyon in California where it's going to shut down and then be replaced by two fossil fuel facilities, and they said, no, no, we're gonna keep this on. And so I think keeping the existing ones on is really important. But then also adding more, trying to invest in things like fusion, which would really make clean, abundant energy fuel source that isn't [inaudible] you know, right? That is safe and can't have those worries that you do have in the back of your mind. Like that's why we have to invest in those things. And I would say right now in Fusion, it's such an exciting moment. I know there's always that saying, 20 more years. 20 more years. 20 more years. But we might be in that 20 year moment right now with Commonwealth Fusion out of MIT or folks at Pacific Fusion who are, you know, outta Sandia Labs. And there's like five or six approaches in the US that are being undertaken today that are just well financed. And I think confidently in the next five to 10 years, I think you're going to see a reactor at a demonstration stage without a doubt and spitting distance to a commercial one.
RD: Yeah. I talked to the folks at Realtor Labs, which I think are based, they're spun off the University of Wisconsin. And they're working on fusion, and it's still a thing in the future, right? It's still a hopeful thing in the future. So for today, we need ways of reducing our energy use and increasing the output.
RP: We absolutely do. I think one thing to be very clear about, we have, John and I like to say we, there's the now and the new. There's the technologies that we have today, solar, wind, storage. Those pieces and electric vehicles that could help cut emissions by half, right? The now technologies, but we need a whole handful of new technologies to get all the way there, right? We need energy running when the wind isn't blowing and the sun isn't shining, so we need investments in those area. When you look at steel and cement, very highly emitting categories, you need solutions there. And so you need to invest in them today for them to be ready for us in 10, 20 years from now. And so fusion fits that, right? Folks like to say, we need fossil fuels. And I would say, no, no. We need energy. We need abundant energy. And for the moment, you are correct, like natural gas and coal in places of the world provide that durable base layer of power. But the purpose of a lot of these investments are to find ways to create that base load with geothermal, with new types of fission and fusion, with storage, to create that, to fill in those gaps. I'd like for us to always think about, like right now, while fossil fuels are primary on the grid, how do they become the backup generators? How are they there for us? In the moments when you need them at the peak or when you need them, because, you know, you're in the middle of winter and the sun hasn't been shining so much and the wind isn't happening. And so it's like thinking about how do we get to that moment would be pretty exciting.
RD: Right? There's a finite amount of it, you know, we keep discovering more, but eventually it's gonna run out.
RP: That's one aspect. I think the other one too is it's, it's a volatile energy source. You know, as we all saw, maybe not as much in the US because we are so rich in natural gas here and it's hard to move around. But when, you know, the Ukraine, Russia War started. You saw a whole country of Europe, basically natural gas prices, not only five x six x, you saw other parts of the world being affected. Places that when you pick on South Asia, which imports a lot of its fertilizer, you know, the source of fertilizer are natural gas, you know, and you could see the ripple effects happening. And so I think for anybody that's interested in like stability [laughter] and domestic energy, you know, this is like a pursuit worth running afterwards.
RD: The grid is the big one, right?
RP: Yes -
RD: Are there other areas of the 10 point like Speed and Scale plan where developers can make an impact?
RP: Yeah, there are. I mean, in many ways also when you think about being an engineer and the developer where you work is an interesting proposition, right? You know, when you look at all of the solutions across the six objectives, there are technical teams at these companies doing fantastic work. So maybe I'll pick on an unsuspecting one, like Fix Food. You know, one of the KRS that could help us save a gigaton a year is on food waste, right? We waste in the developing world and the developed world, so in the US and in someplace else in the world, about 40% of the food we eat. And so figuring out ways that that doesn't end up in the landfill where it can decompose and create, you know, excess emissions is a company called Mill, right? This was founded by Matt Rogers, who's a co-founder of Nest as well, and worked at Apple on the original iPod. He's got a bin that you put your excess food in and it runs at night and it has software that runs that bin. There's a whole UI component on a phone, and you think about as an engineer, you know, you either could work on the device or work on that phone. That's one example, Ryan, another one. Picking on the scale of things, when you look at any of the hard, hard to abate sectors like steel or industrial heat, those companies need engineers on staff to basically create the tools that help those technologies work and run reliably. And so I think there's lots of opportunities and options there. If folks are looking for maybe more of the software fix, you know, you think about a company like Watershed, which is a company that helps other companies not only account for their emissions, but help draw it down. And so those are all on the company sides. But sometimes we love the place we work, so if you look at the Climate Action Guide and look through the Green software directory, you'll see those 60,000 repositories. And a lot of those developers have issues listed of like, Hey, I need help here or there. So that's a way too, I, I think like we're not so far away from being able to help on this.
RD: Yeah. Working at the right place certainly helps, but I think a lot of developers are personally motivated, so having that list of repositories where they can actually contribute. Great, great resource -
RP: I also love how you say that, personally motivated. I could also, you know, when you think about the community of folks that have put solar on their roofs. And batteries in their homes and have hyper optimized, you know, when to, like, that's our community. You know, it starts here with a side project to help yourself and then you realize, whoa, like I should share this with others. Right? And so when you think about like installing a heat pump in your house or figuring out how to add windows that are better for insulation and keeping air from leaking, like these are all things that folks who listen to this podcast, we naturally are like inclined to do. And so, to your point, personally motivated. Stepping away from the company and other things, but thinking about like our own lives and saying, well, how do we decarbonize this? How do I make myself more reliant on the sun, the battery that I have, in many ways getting off the grid. And that's an exciting challenge.
RD: You know, I read something where it's like, you have this knowledge, this power that's sort of overwhelming to be like, I can build anything I want. I can fix anything I want. But I think something like climate change is such a big, big project. How can they break that down to the smaller pieces that they can actually tackle?
RP: I would actually say go for the big pieces first. In some ways it's like flipping it around. Because when you think about where the biggest sources of our emissions are that we can control, right? It's the power we use in our home. So figuring out how to decarbonize that either directly, you know, solar storage or picking a plan that your energy provider has that has more clean energy on it. So that's the big one. The next one is how we move, right? Are there ways for me to get from point A to point B? Not needing a car. Like hopping on a bike or if I have to take a car, maybe it's an electric vehicle, right? There's a lot of electric vehicles that you can buy today that are at a great price compared to just a year or two ago, and it will only get better in the years ahead with just the amount of innovation that's happening. And then food waste is the big one. And thinking about like consumption there and, and how you tackle that. I will say the hardest and biggest piece of the pie for folks that live in the US and travel is, you know, flight. And when you think about flight, you know, how do you make a difference there? I think the big piece is how do you get more companies and airlines to buy sustainable air fuel? And this is where you as an individual have the ability to encourage the place you work for to be a part of some of these buying coalitions. Because the sustainable airfield piece, Ryan, it's like just starting in the sense that it's not at the scale of solar and wind and things like that. It's at the scale of, well, you know, you're talking about a tanker or two here, we need a lot more so the price goes down. And the only way for that to happen is to start to do it at scale. For someone to be willing to pay that premium on it, that green premium for the moment.
RD: That is interesting. I don't think I've heard too much about the sustainable air fuel. What is that? Is that the sort of ethanol mixes or -
RP: Yeah, there's a few approaches. There's the, call it the easy stuff that you can do today that's not too expensive, but those are supply constraint. This is like using waste oils or using AG, creating ethanol and pieces that that's one way, but there's never gonna be enough of that for the, call it the global flying need. What we really need to do is find a way to turn carbon that's captured into fuel and then that becomes really expensive. And so I'm talking about like fuel that's eight times the cost of jet fuel today. But if you find ways to scale those processes, if you find ways to use and power them by excess solar and wind, you start to get into a place where this fuel isn't 12 times as much, but maybe it's double the cost or triple the cost. And when it's in that range, you can start to think about how consumers and businesses can start to pay for it. I think also the other angle we may find out for flying is that maybe carbon removal alone can get cheap enough. Today it's really expensive, it's like 600 to almost a thousand dollars a ton. Can innovators drive that cost down to a hundred dollars a ton? Can it get down to $50 a ton? You know, flying may be a small portion of global emissions. It's like one to two gigatons of that 59 billion ton pie. But it's one where, call it the more developed world is just emitting in excess. While if you live somewhere in the developing world, you look up and you see folks flying and you say, well, if they do that, why do I have to emit less or pick a cleaner stove or do these things? And so I think it's like, it's almost our Achilles heel in the developed world. We've gotta fix flying. So yeah -
RD: I think I read the stat today that 80% of the people in the world have not flown on a plane -
RP: That's right. And then I think the other stat too, it's like 1 to 2% of flyers make up the majority of flying emissions, and so tends to be for business travel, it tends to be in places where people are spending a premium already. So there's a way to tackle this. It just will take a concentrated effort in us saying that this is important. You know, the folks at GitHub put together an incredible resource looking through these repositories, and so if you're approaching it from the engineering mindset, absolutely go there and look at that. If you're looking to get a primer on the climate crisis, the emissions crisis, and how to tackle it, you know check us out and the work that we do at Speed and Scale, right? I mean, Speed and Scale isn't our day job. Our day job, the team at Speed and Scale is investing and finding and funding companies that are trying to decarbonize and tackle the emissions crisis. And so for this, this is our map. This is our tool that we use to help orient us to, well, where do we have to spend our time? And so we've got that whole plan online as well as a newsletter that folks can see all the nerdy news items we're reading that attach to each of these. I think what folks will realize is this touches every part of the way we live, the way we move, the way we power, the way we eat, and that the solutions aren't completely far off. And if we start to look for them and implement them we'll start to make the progress that we need. Ryan.
RD: You said you're investing in decarbonization solutions. Is there something that you've invested in that you're really excited about?
RP: Ooh. Oh my God, there's so many of them. You know, I wear a few hats, like of course, the one that we're spending time together is the co-author in Speed and Scale. But I get the joy of working for the chairman of Kleiner Perkins, John Doer, and he has been a nonstop prolific investor in clean tech during the first wave. And then the second wave. Right there is almost this lost moment of time of just realizing, well, you know, it's not that it's lost, it's that it just takes longer to build these kinds of companies. You know, bits and bytes, the world that maybe you and I deal with on, on software is fast. It's just easy to deploy when you're talking about changing the way you drive and power, those projects take a little little more time. And so if I had to share one project that's coming to mind, that's a fun one is one's called Rondo. It's a thermal battery. It is a way to take solar and wind excess energy from that and store it as heat, you know, a chemical facility could use. Right? So they're not burning natural gas. And it's a fun one to look at as well too 'cause it's these bricks that kind of like a toaster oven get powered via electricity and it's just like, wow, that seems to be a very practical thing. And if you could build it and deploy it, like you can start to see that, you know, decarbonizing can be possible. And so that's one that comes to mind is a fun one.
RD: Well, ladies and gentlemen, it is that time of the show where we shout out somebody who came on Stack Overflow, dropped the little knowledge, shared some curiosity, and earned a badge. Today we're shouting out the winner, be stellar, answer, somebody who answered a question that was saved by a hundred users. And today's badge goes to Christian C Salvado for answering ‘what's a quick way to comment on comment lines in Vim?’ Vim is a perennial topic, so if you're curious, get there. It'll be in the show notes. I am Ryan Donovan. I edit the blog, host the podcast here at Stack Overflow if you liked what you heard, have questions, concerns, topics, et cetera. Email me at podcast@stackoverflow.com. And if you wanna reach out to me directly, you can find me on LinkedIn.
RP: And this is Ryan Panchos from Co-Author of Speed and Scale. And if you want to find me in our resources there at speedandscale.com and you can find me on LinkedIn and Twitter at ripan as well too.
RD: Well, thank you very much for listening, everyone, and we'll talk to you next time.