Yael Maguire, the vice president and general manager of the sustainability team at Google Geo — which includes Maps, Earth, and Project Sunroof, the company’s solar calculator — stressed that Google isn’t trying to take people’s agency away with these kinds of light-green guilt trips. “We want to make the sustainable choice the easy choice,” he told me, in reference to a slew of new tools the company has been rolling out, from fuel-efficient routing in Maps (which Google estimates has eliminated the emissions equivalent of 500,000 internal combustion cars from the road since 2021), to suggesting train routes to flight-shoppers, to nudging Europeans to ditch their cars when public transportation could get them to their destinations in a comparable amount of time.
Last week, I spoke to Maguire about the sustainability projects at Google Geo, including the team’s Solar API, which provides solar-planning data for millions of buildings worldwide. Our conversation has been lightly condensed for clarity and brevity.
Do you see your job at Google Geo as passively presenting sustainability information to users, or do you see it as actively nudging people toward making better choices for the planet?
We’re not trying to take agency away from anybody. We want to make sure — whether you’re a consumer choosing an eco-friendly route, or you’re a developer who’s thinking about trying to build more sustainably, or you’re a solar developer who wants to help with that — we want the choices to be in their hands. But we want to make it the easiest choice possible because, while it’s ultimately their decision, it will lead to carbon reductions over time.
That’s the idea behind fuel efficiency suggestions in Google Maps, where a route is prominently displayed with the little leaf, right?
Exactly. We launched a capability in Google Earth last year to help real estate developers do high-level planning and building development to make the sustainable choice the easy choice. As they’re saying, “We’re trying to get this many units with these kinds of amenities, etc., etc.,” we give them the tools to optimize for all the things they want to optimize for. But we can also say, “Hey, if you also care about sustainability, you can use different materials, we can get more sunlight in the area, and you have this much potential for solar.” And that just comes bundled with the tool itself.
We always try to find the co-benefits. I know for me personally, I always try to make the sustainable choice as much as I can. But I know that other people may not be as motivated by that, and having those co-benefits — like, it saves money, or it saves time, or it saves fuel, whatever it might be. We want to try to bring those together as much as possible.
When I was in Tbilisi, Georgia, a few months ago, I was using the ride-share app Bolt, and at the time it had a feature where if you tried to book a car to a location less than a 15-minute walk away, it would suggest you walk instead. I saw in a video from Google’s sustainability summit last fall that you’re rolling out something similar in some locations in Europe — France was one. Do you find these sorts of rollouts in the U.S. are stymied at all by how un-walkable most American cities are?
We are trying to make the most of cities as they are. They’re hard to change. But one of the things I find really encouraging is there’s definitely a long timeframe for this. Mayors and the folks in their departments of transportation recognize that they have to make more options available for people to commute and move around. They’re not necessarily going to be able to change things overnight. But there are major changes that are happening — for example, in the city of London, we were able to announce hundreds of miles of new bike lanes. So a lot of changes are happening over a relatively short amount of time, too.
Sometimes it’s hard to know what is going to be the impact of those decisions, though. And so, again, with these tools, city planners have the opportunity to scenario plan and say, “Okay, we’re thinking of trying to put bike lanes in this corridor in the city, what is going to be the impact on carbon?”
I wanted to ask a similar question in the context of a new feature that suggests train routes to Europeans looking for short-haul flights. How is Google thinking about promoting low-emissions transportation options like trains to Americans, eventually, when our infrastructure often isn’t there yet? Is this a challenge you talk about internally?
It is definitely something that is top of mind. But I do think even in the U.S., there are times when taking a train is actually faster. There are actually a lot of instances where walking, cycling, and public transportation are the most effective ways to get somewhere — and that’s not even considering the cost side of it, which is also something people might want to consider. I’m actually fairly optimistic — when I worked in San Francisco, I took public transportation, and I tried to walk as much as I can in all the cities that I’ve lived in, so I feel like I have lived experience in what the reality [in the U.S.] is. And some of these alternative options can be very effective. There’s more work to do, though, to make sure that we’re doing this globally.
Arguably, Google Maps could have a significant role to play in the success of the larger EV transition in terms of making charging stations and trip planning easy and handy for drivers. I’ve been working on planning my first EV road trip this summer and have been pretty intimidated, to be honest. Can you tell me what is in Google’s pipeline to help make this process easier for drivers?
I can’t talk about things that haven’t been announced yet, but I will say that, just as an overarching goal, we want to make that as easy as possible. I’m an EV owner, I have been for a number of years, and I know sometimes it can be a cognitive task to think about, “How am I going to charge and what is that experience going to be like?” So I would just say that we are really aware and trying to deeply understand the problem as much as possible, and our goal is to really address it.
Even when someone is thinking about purchasing a car, oftentimes people go to Google Search to look for vehicles, and we can help people understand what the potential is of a particular vehicle they’re considering. What typically concerns people is a long-distance trip. So we’ve made a tool where you can plug in a familiar destination — like for me, I live in San Francisco, it might be going to Tahoe— and for a given car you can see how many charges would you have to do on the way. Being able to make that info a little bit easier for people to see before they even buy the car is a thing that we’ve tried to do.
We’re also trying to make charging experiences as positive as possible. The first thing is, honestly, just getting as many chargers on the map as possible. There are a number of different providers who have charging infrastructure and sometimes all the data isn’t widely available so we’ve tried really, really hard to work with those partners. We have information on, I believe, 360,000 chargers worldwide and we’re constantly trying to grow that. On top of that — and I hope you don’t experience this — but not all the chargers work. You’ve probably seen on Google Maps, there are reviews, right? So there’s all kinds of work happening there.
My EV doesn’t have Google Maps integrated, unfortunately, but I’m really looking forward to one day having this feature where I can search for a charger along the route. We’d like to get to that point where you don’t actually have to do all this planning in advance and you can just get in your car and plan along the way like you would if it was another type of vehicle.
It’s one thing to have a tool like the Google Tree Canopy available for cities and organizers, and it’s another thing for people to actually use that tool and act on the information. How are you measuring your success?
We measure our success ultimately by what people do with our tools. So it’s not just about putting the tool out there. We actually try to understand what people are doing. In the case of what we did with eco-friendly routing, we worked with the National Renewable Energy Laboratory in the U.S., for example, to help validate our carbon emissions model. We’re going through that process for everything we do, whether it’s Project Sunroof or the Solar API, or other things like that.
You preempted my next question, but maybe you can talk about it in a more macro sense — Google has the goal of “collectively reducing 1 gigaton of carbon equivalent emissions annually by 2030” with tools like Solar API. Can you give me any sort of progress update?
This is a project that’s been going on for some time. We’ve been working with solar developers for a while, but we’ve been pleasantly surprised not only by the solar developer community engagement, but there’s actually other industries that have shown interest. So MyHEAT — they’re not a typical solar installer, but they’re finding this data really useful to go to cities and help them with the plans that they have.
So the gigaton goal itself, there is nothing to share now other than the progress on eco-friendly routing, but it is something that we hope we’ll be able to share progress on over time. But so far, we’re quite happy.
At a time when there’s a lot of nervousness around AI — and often for good reason — you’ve been pretty vocal in your excitement about how such tools can be used for the positive purposes of sustainability. Tell me why you’re an optimist.
Here’s why I’m an optimist: Because it’s where I put all of Google’s public goals in context. We talked about the gigaton goal, we talked about the Solar API — but I think this is also a question about energy usage and carbon intensity. We will continue to invest in the infrastructure that we need — and we need that infrastructure to be able to actually help solve some of these problems, by providing information to people — but at the same time, the company has been really focused on trying to minimize the carbon intensity of the energy we produce. So, since 2017, we’ve been operating off of 100% renewable energy; this is on an annualized basis. We also have an initiative to use carbon-free energy — so the source of the energy that ultimately goes where electrons are going to our data centers, we’re actively measuring what percentage of that is carbon-free on a 24/7 basis.
With our net-zero commitments, to be on a net basis by 2030, that includes all of our AI infrastructure. That’s where I would try to separate the energy use that’s required to operate AI from the carbon intensity, which I think is very different. Our data centers, we estimate, are one-and-a-half times more efficient than your average data center. And with AI workloads themselves, in some instances, we’ve been able to get the energy usage down by 100x, and the corresponding amount of carbon intensity down by 1,000x.
But to your point, at the same time, it is very much on our minds that the carbon intensity to run all of these AI workloads — how does that compare to the benefits that they’re able to provide? I think that’s where I am. I do have a lot of optimism about the efficiency work, about the trajectory of carbon-free energy and net zero. The upsides in terms of what it does for solar, what it does for transportation — yeah, I am a big believer.
The big reason why I’m so excited about this opportunity in the Maps and Geo space is I just think there’s so much opportunity for all kinds of organizations, including individual citizens, to make these choices and changes to their environment. And I think the role that AI has is enormous — obviously not the whole thing, because it doesn’t build cycling lanes. People have to go do that. People have to change policies around how buildings are going to have less carbon intensity when they’re built. There’s tons and tons of other work that is required to actually build the future that we want, that is lower carbon intensity — ideally zero. But I do think that AI plays an enormous role as decision support for all those choices that are needed in the future.
