top of page

Episode 36 Transcript: Wading Into AI and Biodiversity

This is the transcript for Episode 36.

Episode 36 Transcript: Wading Into AI and Biodiversity

Molly Wood Voice-Over:
Welcome to Everybody in the Pool, the podcast for the climate economy. We dive deep into the climate crisis and come up with solutions. I'm Molly Wood. This week … a little something special from a conference I recently attended in Phoenix called GreenBiz … which is a gathering for sustainability professionals … people who work in the climate part of whatever business they’re at … be it AT&T or energy companies like NextEra or Wells Fargo or any company you can think of, really …


And while I know conferences THEMSELVES are not the most sustainable business …

the fact is … it’s still magical and valuable to meet in person with people doing the same kinds of work … to learn about trends and developments … and in my case … to moderate a fascinating mainstage conversation that brings together the two parts of my journalism career …


tech … and climate … in this case … AI … and biodiversity. And GreenBiz was kind enough to let me share the session with all of you.


It was called AI for Nature, Climate and Beyond: How New Tech Is Transforming Sustainability …

And you’ll hear me speaking with Elizabeth Hunter … co-founder and COO of a biodiversity tech startup called TreeSwift … Melanie Nakagawa … chief sustainability officer at Microsoft … and Millie Chapman … a postdoctoral fellow at the National Center for Ecological Analysis and Synthesis and Climate Change AI … which is honestly … kind of a job I wish I had but probably am not smart enough for …


You’ll also hear a brief introduction at the top from Dylan Siegler … the senior vice president of sustainability at GreenBiz and host of the conference … who I adore.

Anyway … here’s our conversation … please enjoy!


Speaker:
Has anyone been in a room at green biz where AI was not mentioned? We are going to be leveraging our influence in this year and in years to come through technology. And so this next panel is going to look at what AI and other new tech can do for sustainability.

It's led by one of my very favorite people in the entire world, Molly Wood, and it features, uh, Well, I'll let them introduce themselves, but uh, Elizabeth, Molly, and Millie, I, Elizabeth, Molly, Millie, and Melanie. It's a lot of Ms, um, who are here to, um, to educate us here. So thank you.


Molly Wood:
Can this be the walk on music for the rest of everything that we ever do ever again? How wonderful is this event? Welcome everyone. We do, we do sort of sound like. A girl band, like just a tiny bit . I'm just gonna own that upfront. Um, I am thrilled. I am Molly Wood. I'm thrilled to be on this panel. I am gonna have each of you introduce yourselves one at a time.

I know that AI and biodiversity on its face sounds like a slightly, uh, unlikely set of bedfellows, but we're gonna explain why it actually all really does come together. And, um, as we mentioned, please, I'm gonna have each of you introduce yourselves, tell me what you're doing and how it intersects, um, with biodiversity.


And Melanie, I'll start with you. Great. Uh, thank you. And hi everyone. I'm Melanie Nakagawa. I'm Microsoft's Chief Sustainability Officer and thrilled to be on this panel. Uh, really excited to get into the discussion about the role of AI in these critical ecosystems. And we'll get into where the data is, where the technology is, but needless to say, hopefully by the end of this 17 minutes, you'll walk out with a couple of really great concrete examples of how technology solutions are really closing important data gaps.


Elizabeth Hunter:
Great. Elizabeth. Um, hi everyone. I'm equally thrilled to be here on this panel. Um, I'm Elizabeth Hunter. I am one of the co-founders and also the chief operating officer of TreeSwift. Uh, TreeSwift is a startup company using robotics and AI to capture landscape data, uh, and really excited to, um, just, you know, share with you as part of the panel, just some of these new and innovative ways that we can think about using technology for nature-based solutions.


Thanks. Awesome. Hi, everyone. I'm Millie. I am a postdoc at the National Center for Ecological Analysis and Synthesis and a core team member at Climate Change AI. A lot of my work thinks about how we can leverage AI not only to understand our ecosystems and the way they're changing, but to come up with strategies to mitigate impacts on nature.


Um, so it seems like for AI to do its best work, it needs a lot of data. So I actually want to start with the data challenge around biodiversity. Um, and what, and I'll, I'll probably Melanie, have you address like what we can do with what we have, but Elizabeth, it's my understanding that, I mean, actually we lack a lot of data to even pull insights from at this stage in the game, right?

That's, um, that's exactly right. So I think about like, and how the landscape. is changing for innovation in this space, there's two things that I'm really excited about. There's AI, um, but, and that is really crunching and, and synthesizing, coalescing, and, and looking at the insights that we can get from all of these really incredible data sources that are available to us.


But on the flip side of that, there's the source. And I think one of the things as a technologist I'm super excited about is using technologies like robotics, um, that we're really putting into practice at tree swift to fill in those gaps and think about new sources, uh, new views, new perspectives of how we can actually interact and gather this data across the landscape.


Um, a really concrete example of this is at tree swift. We focus on data in forests, so above ground biomass, um, these really important carbon stocks that are, are critical to, um, right, net zero goals, et cetera. And so one of the things that, um, has been longstanding, at least in forestry and some of these solutions is, is looking at sources of data from, from kind of, um, a bird's eye view.

So over the canopy. No pun intended. No pun intended. Um, and at Truthswift we're really filling that gap. So with robotics we can come in, we can actually work in concert with boots on the ground to have this view and kind of this new data source of under canopy. So looking at the crowns. of trees and how that, um, how we can infer and extract insights about these ecosystems and to think about how these new sources actually, um, fuel how we think about measuring, measuring, um, uh, these, these things about biodiversity.


Right. Um, Millie, I want to come to you about what we don't know and then Melanie on what we are doing to solve that. Yeah. I think that's a great question of what we don't know. And that's, I think your work is really inspirational and super powerful and going to be incredibly critical. So when we look at biodiversity data, which is collected from all different sorts of streams from people taking pictures on their iPhones and uploading them to participatory science platforms like.

iNaturalist, um, and otherwise, to camera traps and acoustic monitors, imagery from drones and satellites, um, We think that maybe we have a lot of information about the world, but really that, that information is incredibly biased, um, and it's bias along social and political dimensions. Um, we can see armed conflict in when and where we have data, colonialism in what information we have and who owns that information.


We can see redlining in U. S. cities in where we have biodiversity data. We have almost double the amount of biodiversity data in neighborhoods that were not historically redlined or segregated. Um, and so we have to think a lot moving forward about how we can ensure that we fill those data gaps in such a way that we don't propagate these social inequities moving forward when we start to leverage all of this data and throw it into our AI models, um, and our decision making algorithms.


Like, how can we ensure that those biases. Um, are not reflected, no pressure, Microsoft. So talk to us about filling those gaps and, and what AI can do with the information it has now, and then how we get more and get it as, you know, as clean as it can be, I guess, for lack of a better way of putting that.


Sure. So I think we haven't used the 7 percent figure yet. So let's put that out there. I think you go skirted past it. So 7 percent is the amount of actual data we have for our biodiversity and our ecosystems to date. So you talked about all these incredible technologies that are out there from Landsat to satellites to soil and that's fantastic.


We're getting a lot of data from that, but that only covers around 7 percent of the data. So we're using AI to fill some of those data gaps that are out there. But frankly, the next piece is Why should we be doing this? What are the insights you can get from this information? I thought you did a great job talking about in the land space of understanding where those trees are growing and how they're growing.


You can also use satellite imagery to know where deforestation could happen or is happening because you could see when roads are being cut through forest. To do, to do greater deforestation and greater tree cutting. And you can see those things in advance before they actually get to the cutting of the forests and alert the authorities and start to actually address deforestation earlier in the game.


So there's an organization we're working with called Amazon, and that's one that's looking at Brazilian deforestation and trying to catch that before it happens. Another area is in poaching. Uh, wildlife trafficking and poaching that happens, which, um, is a, you know, sort of a shocking figure about the number of animals that are poached on an annual basis.

But you can use satellite imagery and this type of technology to start to see where humans are entering these different areas in these forest landscapes or these ecosystems to, again, alert the authorities as to where humans are entering these areas to get the authorities there earlier to respond. And then another area where you can use technology and satellite imagery and, and, you know, ecosystem analysis.


is around wildfires, and you can see that on two sides. So you talk about the importance of trees. Um, the other side of it, you know, the importance of trees for, uh, conservation, for environmental protection, for biodiversity and ecosystem services. Also, as carbon sinks, which are really important for the climate crisis that we are all working.


So, especially those in the room to combat, but the other side of forces when they burn, they emit carbon. And so, um, last year, I think it's around seven gigatons, I think is the stat that wildfires, the forest, the carbon from wildfires emitted. So if we can do better resilience and optimization of prediction around responding to and addressing wildfires.


We also have multiple benefits that come from that. And so one of the ways we've been doing that is you can use technology and AI to look at large ecosystems to understand what's at risk, what are the factors that, you know, predict or increase the chances of hotter fires or more complex fires. And again, allocate resources directed at that to get out in front.


The other piece too to the, to the wildfire space is recently after the devastation in Maui and the Lahaina fires, Microsoft partnered with Planet Labs to be able to provide responders and policy makers and others with information using, you know, using AI to analyze the maps. to understand exactly where the fire, how far the fire went, what was affected, to really think about the post disaster recovery side and how do you use these type of map imagery when you couldn't physically get access to certain areas.


You can use that imagery to help people better understand what's happening. So there's really interesting, you know, kind of examples of, of how do you bring technology together with AI to close some of those gaps, to yield, you know, benefits for both people and planet. Yeah, I love these, I love these specific examples because as we mentioned, these can seem like strange bedfellows, but all of you have effectively determined that, you know, we should stipulate technology is going to be crucial in terms of gathering and the types of insights you mentioned.


Molly Wood Voice-Over:
Time for a quick break. When we come back, the second half of my GreenBiz panel … where we get more specific … about what we can accomplish … with the right tech in the right place.


Molly Wood Voice-Over:
Welcome back to Everybody in the Pool. I’m sharing my mainstage session from the GreenBiz conference last week in Phoenix … and we’re about to get specific about what AI and other new technologies can do to help us improve our relationship with nature.


Molly Wood:
Let's turn to this idea of insights and action because it sort of feels like what the, you know, we've all known that a lot of surveillance, a lot of satellite technology can gather data. The question that AI is poised to answer is. Cool. Now what? Um, so maybe I'll just come back this way. I'll start with you, Millie.


Millie:
Cool. Now what, what, what can we do in terms of these insights and actions that we maybe couldn't do before? Yeah. Just a small question, but I think going back to this, this you have eight minutes, um, going back to this example of wildfires, if we all of a sudden now, uh, can understand where wildfires are, we can predict their spread at kind of unprecedented scales.

We can then leverage this information and kind of the other flip side of. AI, which is coming up with strategies to kind of suppress those fires or to mitigate their spread, um, using like optimization algorithms. Um, a lot of my work is thinking about how we can use all of this massive amounts of biodiversity data to think about how we prioritize investment in conservation, in restoration, in sustainable management on agricultural landscapes.


So if we have all of this information. Um, maybe it can help us make more strategic investments, uh, in these conservation measures. Um, so I think, um, you know, we, we kind of voiced the 7 percent statistic. Another kind of stat or, or thinking about the wildfire specific, um, kind of use case is forests cover 30 percent of our planet.


Um, and when we think about AI and filling those gaps and all of the, um, kind of the, the source, the new sources of data coming in, like the ones that we're developing at TreeSwift, um, what's striking to me and something that that I've, I've kind of like learned kind of on this, on this journey of kind of putting this technology into the hands of kind of boots on the ground and, and local folks working is that a lot of these, uh, kind of insights, um, whether they be coming from AI, uh, kind of other analyses still rely on.


kind of boots on the ground people going and still looking and verifying. And there is a level of rigor that is still required to, um, kind of kind of pull back the curtain, if you will, to not make the kind of this AI or machine learning process so opaque, but actually, um, going on the ground and verifying and, um, making things.


more transparent. So in the case of wildfires, one of the things that's really important is to understand the health of kind of, um, the health of trees, how that put that those pose a risk to various assets that can start wildfires. And so one of the things that's really powerful is that if you have Again, new ways to source and collect the data, you can work in concert with, with people whose job it is every day to, uh, ensure the safety, um, and mitigate the risk of these disasters.

Um, and so I think it's, a, a lot of times we think about, Um, or, or there's a lot of, uh, kind of conversations about what does this AI mean? Is it kind of replacing or kind of, uh, how is it coming into the conversation with, with folks kind of doing this most important work? And I think one of the things that I've been seeing and like really like, But how, how, how the data, what the data is really showing is that, um, it's AI really working in concert with folks who have these insights, but making it more scalable, more systematic, um, to, to kind of extrapolate and really scale over the whole landscape.


Molly Wood:
I'd say to add to that, uh, it's always hard to go with a third in a panel. Uh, I would say what to add to that. Think of yourself as first all over again. Exactly. Perfect. I don't know what the next question is. Um, but on the, so now what I'd say on the insights piece, there's two areas. One is how do you take all these data and put these insights and increase productivity, increase.

So, the phrase we often use is, how do you grow more crop per drop? So right now, I think it's around, um, the typical farmer will, or the farm will receive around 500, 000 data points on a daily basis. That's expected to grow to millions of data points over the next decade. And so with all that insight, all that data, how do you create insights that you're allowed to, you know, grow more crop per drop, more efficiency, and more, and higher productivity?


Output to feed the world that is, you know, obviously a growing opportunity for all of us for all these technologies to be a part of so that's one area is how you try to turn those insights into economic productivity. Another one is how do you turn those insights to build trust? In both the measuring, the verification, and the reporting in these systems that are so important to what we need from solving and addressing the climate challenge, but also to investing in these solutions.

Microsoft has a climate innovation fund. It's a billion dollar innovation fund that's investing in innovative technologies that not only will Microsoft need towards our own sustainability commitments, but also that the world will need. And so we've recently invested in two companies that are addressing some of this, both the data challenge and the insight opportunity.


So a company like Vibrant Planet or Yardstick, Yardstick is looking at soil carbon measurement and really, really being able to provide the, the measurement and the verification about what is the actual soil carbon content in these types of projects that then allow people to then share that with the marketplace to say, this is what the carbon enhancement looks like for these types of.

Uh, for this, for this practice that I'm using, and that really helps then on the reporting side. Uh, you're seeing a growing interest from policy makers and regulators. Whether it's a task force on nature based disclosures, or the European Corporate Sustainability Reporting Directive, the CSRD, that's asking for more information from corporations and others on nature based disclosures and reporting.


And so how do these tools help improve that? And recently with our partners at Planet Labs. We produce a white paper looking at what some of the existing tools today, what they exist on biodiversity and nature based and how they can be used in current reporting, um, future reporting requirements that are out there to get companies and others ready for it.

But between insights for the, from an economic productivity perspective, but also insights for ensuring that there's. There's trust and validity and verification in these systems that are so critical for the ecosystems we're going to need. I think there's a great opportunity here and we're seeing that in practice today.


Molly Wood:
Right. And such a crucial foundational layer for policy, for enabling financial tools, all of the kind of tools we need. In our last minute or so, the theme of this event is of course, listening and influence. There's, you know, ultimately you're trying to get insights and analysis to create, uh, influence.


Millie:
Millie, you have written though about. Um, I think another thing that is, um, I think another thing that is, um, important to think about is, um, how do we make sure that we're listening to the right sources? Yeah. I mean, I think that's a great question. I think it was brought up kind of time and time again in this conversation about, uh, collaboration with local communities, with, with people who are actually feeling the impacts of environmental change, of ecosystem degradation.

Um, I think another thing that is, um, It's really important and really exciting about this conversation is kind of moving from this idea that like all we can do with AI and big data is to document degradation. Um, that we can do more than that, that we can come up with solutions that really there is kind of a positive lens on all of this and, and moving towards that with local communities and with.


With, with big companies that give us the capacity to scale up, um, which is ultimately what we need to do, uh, in the context of these planetary crises. Amazing. Melanie, Elizabeth, Millie, thank you so much for an incredible panel. We wish we had an hour, but we don't.

to Melanie Nakagawa, Elizabeth Hunter, Millie Chapman, and the inimitable Molly Wood.


Molly Wood Voice-Over:

That's it for this episode of Everybody in the Pool. Thank you so much for listening!

Email me your thoughts and suggestions to in at everybody in the pool dot com and find all the latest episodes and more at everybody in the pool dot com, the website. And if you want to become a subscriber and get an ad free version of the show, hit the link in the description in your podcast app of choice.


Thank you to those of you who already have. See you next week.

bottom of page