As we've said, this is Kara Hurst from Amazon, and one of the best things about Amazon as a tech company is that it has to work in the physical realm as well as the bit space. So they're moving trucks as well as packets of data, and because of that they come up with solutions that they also have to apply to their own situation and get to test on their own home soil before they kind of export it to the world. So it gives Kara actually a unique perspective on a lot of the issues we're talking about this week.
But I first want to ask, what is it like to be a chief sustainability officer in the United States right now? If anyone needs context, my wife, who's academic, one of her colleagues had to scrub the word emissions from an NSF proposal that had nothing to do with even green energy. It was just the word emissions. And yesterday the president said the high cost of so-called green renewable energy is destroying a large part of the free world. I just want to know what goes on in your mind when you're reading that. Are you feeling the same pressures that some of the DEI executives felt? Is it a different strategy and a different issue altogether?
There's a lot there. So, thanks for having me. I think this is—it's a really exciting week, Climate Week this year. There's a lot—I feel like the momentum's grown actually this year, bigger Climate Week than ever. And one of the really exciting things, as you started there, in my role at Amazon is that we have incredible momentum and scale and size at the company. And so for my role, I have the great privilege of actually being able to utilize that size and scale across the company to not only play in many different industries. So Amazon, we have our logistics and fulfillment business, we have devices, we have our studios business, we have retail, we have cloud computing, we own Whole Foods grocery business—I could go on, studio. So there's so many different ways that as a chief sustainability officer we get to think about sustainability and trial and test different solutions. But we also have our Climate Pledge, so our commitment to be net zero carbon by 2040. And the momentum has been really interesting. We announced this week we have 600 signatories to the Climate Pledge, 100 of which joined us this year.
And these are big companies that signed. These are not just like small—yeah.
Yeah. But yeah, so we see the momentum actually happening, and I do want to reiterate that story as well, that there's—there's progress. So with, you know, 100 more companies committing this year, I think there is also progress. And that is people taking these goals not just to take them, but to operationalize really decarbonization within their companies, to join together across industries, and send demand signals within those industries that we want better products. And that by that I mean greener concrete, better steel, better cement, things that we don't talk about as much because they're not the consumer-facing products. I'm happy to talk about those as well. But those heavy industrial industries are changing.
I mean, anyone who's ever driven an EV surely knows that it's actually a more fun driving experience than all but like the most elite gas-powered cars. I mean, it's just—why not own an EV? For $40,000 you can get something with a lot more pickup. So yeah, I get it. Like, but are you saying that corporate America is kind of looking beyond maybe what the rhetoric of the government to being like, these are our goals, you know, and this is what we want anyway?
We've been doing this for a long time and our goals are long term. So with a company of our size, our plans are very long term and we work globally as well. So we're working all around the world, we're working in Europe, in India, in Brazil, in Latin America. You know, we have so many different places in the world where we're doing this that I think the long-term nature of those goals means that we're looking, you know, across industries, across operations, across geographies. And so it's a very big view.
Right. Let's talk about—you have an approach called 3D sustainability. That's digitizing data, discovering insights, and delivering breakthroughs. Three D's there, good enough. And so what I wanted to talk about is—in other words, you're using AI for sustainability goals. A concrete example is the life cycle analysis of products that Amazon sells. We all pull out our phones, buy the thing that comes to mind. We don't give any thought to where it comes from. We definitely don't give any thought to like the fact that it had a journey even before it got to your fulfillment center. You have to think about that stuff for how many SKUs? Like—
We sell billions of products. So—
Okay. So AI's going to help you track billions.
That's a great example. So what we're talking about is the application of AI and the opportunity that AI brings to sustainability. And I think that's another really exciting thing right now. AI's been around for a long time and behind the scenes we've been using machine learning and AI to optimize our operations, the efficiency of what we do. We've been using it for a long time actually to operationalize packaging decisions. So to right-size our packaging, that's one way we've taken out 4.2 million metric tons of packaging since 2015, which is a lot of packaging material. So that's a lot of packaging savings using AI and machine learning in our operations. But now we're using it opportunistically to make changes in sustainability, which I love. And that rapid LCA tool, or life cycle assessment, is where these life cycle assessments are something companies have been doing for a long time. When you think about—we have to think the life cycle assessment of a product is when you think about how it's designed, what materials go into it, how it's manufactured, how it's used by a customer, and then how it's disposed of or reused at the end of life. So, that full cycle of a product. So if you think about a phone or a glass or a table, everything that goes into this, how it's manufactured, used, and then disposed of, that takes a long time for a company to track and do. And typically months, if not sometimes up to almost a year to get that information, to source the data, to put that into a life cycle, and it's costly. We've developed a rapid LCA tool at Amazon that now takes that from really months to weeks. We've done 4,000 of these in the last quarter. And we're so excited about what that data gives us.
So you run an analysis on a product made by someone you're buying from, essentially.
We've done it on our own products.
Yes. And we've also beta'd it with a couple of other companies now.
Okay, because that's what I was going to say is if you apply—I mean, obviously the lion's share of the billions of products are probably all, almost all, but 0.001% are made by other people. Are you going to give scores? Are you going to say—
For our products primarily first. We're doing this week. We may give this tool to others. We haven't decided that yet. I mean, there's certainly a huge appetite for it. But right now, what we're primarily interested in is for our own products that we're making. We need that information for our product.
Okay, so this is an internal like watchdog kind of product, but it could become something—
—for other products. We make clothing, we make, you know, lots of products within our own system, so.
I mean, that's another kind of overarching question I have, and an example that you guys gave last year, I think, was solar. You have solar farms where you've applied AI in order to better manage the storage of the sunlight in batteries. And because of that, it's been very effective for you. Is AI like that something that you see a demand for outside of your walls as a product, or as what would it be—as a partnership? What would you do?
There's a lot of—I think if what you're getting at, there's a lot of applications for AI in many different ways in sustainability. So we've created AI tools now that help us look at digitizing information to get at a better carbon footprint. We have tools that allow us to think about those packaging decisions. We have AI tools that are helping us to reduce food waste. So that's a really interesting use case where we've had AI tools that help us to identify food that is either slightly damaged or about to go bad, where we can then pull it off the shelves early enough to donate it and get it to people in need very quickly. We have great partnerships here in the United States with Feeding America. And so before that food, you know, may rot or go bad, we can pull it off, get it into the hands of people who need it, get it into the food banks. That's a great use case for AI. I love that one. We also use it in disaster relief, ocean cleanup. We can talk about those examples.
Right. So the ocean cleanup is an example of you guys partnering with a group outside. I think they called the—is it called The Ocean Cleanup Project? Yeah. And you know, so you're providing this tool that can identify plastic in the ocean and track it as it moves around with the Great Pacific Garbage Patch. You can see where it's moving. So the people out there with boats that can collect the plastic can know where to go, right? Literally which way the wind is blowing. Is there talk inside your walls of getting more and more of your AI out to the world?
Absolutely. Yeah, we think those kind of use cases to help—you know, I worked in nonprofits for a long time before I came to Amazon, and one thing I certainly know is those partners—the benefit that we have at Amazon, again using our size, our scale for good, is one of the critical roles that we play in society. And so I know those nonprofits, they're eager for how to teach us and help us build our capacity to use AI in those ways. They have incredibly smart people who are doing this as well. But if we can help to build tools quickly where they can utilize AI to go and deploy their scientists more efficiently and effectively, that's great. They're out there in the ocean trying to find in the great Pacific garbage patch where that plastic is, you know, and where to deploy the people to go collect it. If we can analyze that information more quickly with AI and then send people to where plastic is collecting, great. Another example is in Hurricane Helene when that hit last fall in Tennessee and North Carolina. That's a great example because that's a situation where you need information very quickly, right? When a disaster strikes, you have a critical amount of time for two reasons: one, for the people on the ground, and two, for the first responders who need information right away. So what we did there was we utilized AI and drones to collect information over a 28-mile kind of perimeter. And we analyzed 32,000 pieces of information or data points. And this is something that would have taken weeks, but it was done actually in I think days or maybe it was even hours. I don't—and that information gets so quickly, and what it allowed it to do is to say to the first responders, 'These are the critical areas that need help right now.' So we can get that information to people within hours, but it also allows them to say these are the places where the terrain is unsafe.