311: Listening to the Living World: Can AI help us hear the planet?
This week our hosts explore the promises and pitfalls of AI, including the extraordinary mission of the Earth Species Project, who are racing to use artificial intelligence to translate the communication of other species before they fall silent.
About this episode
What if we could not only listen to the rest of nature, but actually understand it?
From decoding whale songs to giving nature a legal voice, the possibilities are tantalising - and they may not be as far-fetched as they sound. That’s why the Earth Species Project (ESP) is racing to use artificial intelligence to translate the communication of other species before they fall silent.
How can this cutting edge technology analyse data that would take human researchers a lifetime? And how might findings feed into emerging ideas about ecocentric governance and earth law? We hear from ESP’s Aza Raskin (Co-founder), Jane Lawton (Managing Director) and Olivier Pietquin (Chief Scientist) about this extraordinary mission, and the tools they’re using to achieve the previously unimaginable.
Plus, Christiana Figueres, Tom Rivett-Carnac, and Paul Dickinson explore the promises and pitfalls of AI: its energy demands, its unpredictable impacts on democracy and capitalism, and its potential to become a ‘Galileo moment’ in how humans relate to the living world.
This episode features recordings of animal species - some of which were used in the training of ESP’s NatureLM-Audio model.
Learn more
🔊 Discover more about ESP’s large audio model, NatureLM-Audio, and the technology driving their work
📖 Read Aldo Leopold’s A Sound County Almanac, including the essay Thinking Like a Mountain
🐋 Read about about the story of Tokitae the orca calf
🎤 Leave us your voice notes and questions for upcoming episodes on SpeakPipe
Follow us on social media for behind the scenes moments and to watch our videos:
Instagram @outrageoptimism
LinkedIn @outrageoptimism
Or get in touch with us via this form.
Producer: Ben Weaver-Hincks
Video Producer: Caitlin Hanrahan
Assistant Producer: Caillin McDaid
Exec Producer: Ellie Clifford
Commissioning Editor: Sarah Thomas
This is a Persephonica production for Global Optimism and is part of the Acast Creator Network.
Full Transcript
Tom : [00:00:02] Hello and welcome to Outrage and Optimism. I'm Tom Rivett-Carnac.Paul: [00:00:04] I'm Christiana Figueres,
[00:00:06] And I'm Paul Dickinson.
Tom : [00:00:07] This week, from decoding whale songs to giving nature a legal voice, we dive into the Earth's Species Project and the AI researchers racing to translate the languages of the living world before they may fall silent. Thanks for being here. Okay, friends, we have a fantastic episode for you this week. If you've ever wondered whether or not we're ever going to be able to speak to animals and understand the natural world, this is the one for you and the incredible progress.
Christiana: [00:00:32] Wait wait wait wait. It's not quite speaking to animals yet. It's understanding what animals are communicating first.
Tom : [00:00:40] Okay, well, I was imagining it Doctor Dolittle style. Like, you know, talking to the animals. But that's not. That's not quite correct. Is that right?
Christiana: [00:00:47] No, that is, I would argue, a very anthropocentric view.
Tom : [00:00:51] Well, it may not be the first time I'm accused of being anthropocentric on this podcast, but let's see, you're an anthropologist now. We're going to get into this. But just before we do, I think we you know, we're going to do many episodes about AI. This is obviously a topic that is capturing the world's imagination is changing so much. And we have had many questions about AI as well. So let's just start with one of those.
Sam (listener): [00:01:09] Hi there. My name is Sam and I work for a housing advocacy organization, and I've been doing research into the electricity grid in Pennsylvania. And I learned about the pressure that AI data centers are putting on our grid. And I feel very cynical about AI, but I also have seen, you know, Ember, which is a think tank on the climate transition, be quite optimistic and excited about AI. And I'm curious if AI has a place in climate advocacy and a better future.
Tom : [00:01:53] So, I mean, I think to be honest, we're not going to be able to answer these questions in full. This week because we're going to talk about other things as well. But we will keep coming back to them. But this is a very live issue. I actually have a child in my house that is boycotting AI because of the climate impacts, and I constantly get heat every time I use ChatGPT or something similar because I'm told that I'm not really authentic in what I'm doing because I'm using this terrible polluting technology while trying to do something about climate change. How would you both answer that charge?
Christiana: [00:02:22] Well, first of all, one of the major accusations of climate conscious people is that data centers use a lot of electricity. Now, that is a huge climate concern for countries that have a huge carbon intensity in their electric grid. It is not a huge concern.
Paul: [00:02:49] Wait a minute. I could see where this is going.
Tom : [00:02:51] Oh, no.
Paul: [00:02:52] I think I know that center in Costa Rica is run completely off natural water falling from the sky.
Christiana: [00:02:59] Yes. Thank you Paul. Thank you Paul. No, but I think it is really important to D-Link. The fact that in general. Yes. Data centers as one demand for energy and extra electricity, but everything is being electrified. So we have to D-Link the increased amount of electricity production that we are needing and will continue to need from carbon intensity, those are two different issues. That's the whole point of decarbonizing the economy is to be able to use electricity, but clean electricity.
Paul: [00:03:38] I mean, if I can pick up from there, Christiana, please do. Ai I'm going to say is an onion. It has many, many layers. Yes. Number one, you know, big data centers, giant energy consumption, big new thing on the grid. If you've got the wrong kind of grid, if you haven't got a plan for carbon zero electricity, you've got a problem. The next level of the onion is large and growing. Intellectual capability to deal with climate change, efficiency, innovation. Whole systems planning. Big enough intelligence to really understand our energy systems. More intelligent humans operating with AI. And then the next level of the onion. Oh, my. Potentially massive disruption to employment it is going to be completely world changing. Really. In no uncertain terms, there are other layers of the onion that we're going to get onto here. So huge topic. And yes, one that we must dig into the many, many different aspects of it. A bit like the internet, you know, the internet has given us all these different things, from online retail to social media to blah blah, blah, blah, blah, blah. And I think the AI is going to be the same. It's not going to be one subject. It's going to be a whole group of things that change our lives.
Tom : [00:04:45] Yeah, I think that's such a good point.
Christiana: [00:04:46] Paul, I have a question for Paul and maybe for you also, Tom, I also would like to have your opinion. Is AI going to be more or less disruptive to the way we do our things than the introduction of electricity was?
Paul: [00:05:05] I'll give you a quick answer. And that is um, at no time did electricity think to itself. Hmm. I'm not sure I want to do that, but AI is more than capable of thinking. I'm not sure I want to do that. And that's the difference, you know? Controversial as it may seem, um, friend of the show, Yuval Noah Harari, now encourages us to think of AI as an alien intelligence. I don't think it's alien because it comes from Earth, but it's definitely pretty different to us, and it's the unpredictable character of its own sense of determination that separates it from all previous technologies, I think. Hmm.
Tom : [00:05:38] Yeah. I don't know if there's any way I can effectively and, you know, pithy answer your question, Christiana, but I think that electricity made what we were doing more powerful. But of the type it remained the same. Right? We travel, but then we travel faster. We, you know, lit our houses, but then we lit them better. And so it sort of continued of a type, but made us more powerful and more able to meet our needs. I think that AI has the potential to play quite a different, nuanced role in human and the natural world evolution that maybe leads to quite a different outcome. That isn't just a magnification of what came before, but is a different way of relating.
Christiana: [00:06:18] It's more unpredictable and strangely out of our control.
Tom : [00:06:23] Yeah. And this is going to be, I think, one of the backdrops to outrage and optimism over the next year as we try to get our arms around this, because the way in which AI affects climate is completely connected to the way in which AI affects democracy and affects our ability to socialize, and the future of commercialization and the future of capitalism. All of those things, and the way in which AI is changing them are part of how AI either does or doesn't help us solve climate change. And you know, the short answer is AI both does and doesn't help us solve climate change. Both of those things are true. But to unpick that.
Christiana: [00:06:58] That prompts me to my next question for both of you. Can we, in a somewhat simplistic way, agree that AI is an unprecedentedly powerful tool that is neutral in and of itself, and it depends on what we do with it that determines the consequence of whether it is AI for good or AI for bad.
Paul: [00:07:25] All right. Well, let me answer that one by going back to electricity. Thomas Edison was demonstrating electricity a very long time ago. And somebody said, what's the point of it? And Edison said, well, what's the point of a baby? And if I go back to form a guest on the show two times, Yuval Noah Harari, he says, essentially, we've got a baby here and it's being educated. And what he particularly makes a point of saying is it's not going to be educated by what it sees us say, but it's going to do as a child does. Its going to notice what we do and it's going to develop as we develop. And if we say don't lie, but do lie, it will start lying. Now, I've never done this on the show before, but I am going to quote Vladimir Putin who said, whoever rules AI will rule the world. And I picked up a phrase the other day. I can't quite know where I picked it up from, but I think it speaks to a sort of new AI, a AI driven politics. And it said, you may not be interested in politics, but politics is now interested in you. And that's a really sort of scary way of thinking about the commodification of the human experience in a kind of new techno politics, which we're just beginning to understand.
Tom : [00:08:30] Well, I think that's 100% true. And I think humans have a long history of developing powerful tools and then doing the wrong things with them. And I think that, you know, even if you look at recent innovations, relatively recent innovations like social media for connecting us. I mean, you don't have to go back that far. We all remember it. The early days of that, when it was full of optimism and possibility, but because it became about division and commercialization, I would say they became tools that are negative overall for society. And we're in this very fertile early stage in which AI could be this incredibly powerful, positive tool for human connection and and knowledge, growth and wisdom. But it's also possible that we just use it to sell each other more stuff and to dig more things up and then throw them away and create more waste and more destruction. And my fear is that the commercialization will lead to that being the use of the technology. What do you think, Christiana?
Christiana: [00:09:20] Well, I think that all of this is true and that most people I think this could be a very gross generalization, are already fearing what could happen with AI. Hence, I think it would be helpful to look at an example where AI is actually Helping us in very, very concrete way. So are we willing to dive in to an example of the use of AI for something that I think is absolutely fascinating and could never have been done without AI?
Paul: [00:10:00] Can I ask a question, Christiana? Is it the case that, in my experience, you don't rush to involve yourself with things because you have too many demands on your time, but you have actually taken an interest in this specific organization? Is that. Is that correct?
Christiana: [00:10:12] Yeah, yeah, yeah, I was just getting to that point. So the Earth Species Project, whose board I have just joined. And you're right, I don't I don't join.
Tom : [00:10:23] It's not easy to get Christiana to join your board these days. So this is quite nice.
Christiana: [00:10:28] I'm pruning I'm pruning my responsibility tree rather than growing more branches. But I have joined the Earth Species Project because they are attempting and we'll discuss with what success already. To use very advanced AI to begin to understand not just the verbal language, but the overall communication of other species that are not human. And the idea here, as we will hear from the several species project acronym, ESP, very interesting acronym, as we will hear from several people at ESP, is to understand the communication of other species among themselves. So within a species, across non-human species, and eventually even across species that are non-human with us humans. So it's it's a multi-layered effort here that is taking the time, effort and brilliance of many, many researchers so that we can finally, I think, come to the understanding which many wisdom traditions have never let go of, that we are one of 8 million species on this planet, and that we are actually doing the wrong thing. And there is a ethical judgment that I've just put on the table of thinking that we are better than superior to, and certainly the owners of all other species.
Tom : [00:12:15] And we should say, I mean, this may seem to some listeners like a bit of a side project or a side quest, as the kids would say, for a podcast that focuses on climate change. But I think anything that brings us closer to the natural world around us and makes us more connected and enhances that sense of wonder, can only be a positive to our collective attempt to deal with the climate crisis, so I don't think it's ancillary at all.
Christiana: [00:12:36] It is not ancillary at all because if I if I grow in my awareness, consciousness and love for nature, I will be less likely to damage her.
Paul: [00:12:46] And of the three of us, Tom, you're the only one to have met James Lovelock, I believe, and were the great deceased scientist who came up with this Gaia theory or named it. At any rate, were he to be here, I think he'd say that climate change is in 100% about replicating organisms and how they how they flourish or do not flourish on this planet.
Tom : [00:13:04] That's true. Absolutely right. Okay. So, Christiana, why don't you introduce Aza Raskin, the co-founder of ESP. And then we can begin our exploration of this fascinating topic.
Christiana: [00:13:15] I met Aza at a retreat, a three day retreat that was put on by Krista Tippett from the On Being project and the On Being podcast, and he sat down next to me at lunch and started talking about animal communication facilitated by AI and I couldn't even finish my lunch. I was so enthralled. I was so excited. I just went like. Tell me more, tell me more, tell me more. And one thing led to another. And I have now accepted, as I just said, to be on his board. Asa is the co-founder of two very interesting organizations. One is Earth Species Project, ESP, the other one is the center for Humane Technology. But what is fascinating about AI's is where does he come from? He's a designer and an inventor. He was at Mozilla. He was the head of user experience, and every single person will be using today something that he invented, the infinite scroll he invented.
Tom : [00:14:23] I mean, I struggle to forgive him for that, but I think I'll try. Yeah. That's right. Yeah. Anyway. Carry on.
Christiana: [00:14:30] Well, he designed that. So. Yeah. You can. You can love him or hate him for that. But there's definitely no day that goes by without Asa Raskin on a daily path. And he's also, as I said, the founder of the center for Humane Technology, where he actually looks at the ethics of technology. So quite a brilliant chap. And now, having already had the experience with ever growing number of researchers that he's bringing on board at ESP, getting more and more into this mystery of how to decode and demystify the communication that is all around us is just that. We are deaf to that, and he's helping us to decode and demystify.
Tom : [00:15:27] Okay, so this episode is going to be made up of several conversations or clips of conversations that Christiana had with different key people at ESP. So let's kick off with this conversation with Aza Raskin.
Aza Raskin: [00:15:42] You know, I think there are these moments in history where we get a shift of perspective, and even though no atoms have changed, somehow everything is different. And we've seen it a couple of times. You know, the telescope, a new scientific tool, taught us that Earth was not the center. And that changed religion and it changed our relationship with ourselves. Another example of this is when human beings went to the moon, as they stepped foot on the moon and took those incredible photos of Earthrise and Blue Marble, which I think are still the most viewed photos in world history, right? It didn't just change our perspective. That perspective then flowed into creating NOAA and the EPA and the modern environmental movement and the Clean Air Act. There are these shifts in perspective, these moments that become movements. And of course, the most famous one perhaps, is The Songs of the Humpback Whale, an album that for the first time taught Western society that other mammals, these incredible beings underneath the ocean, sing and create melodies that get passed from one generation to another. And that created the momentum for the.
Christiana: [00:16:45] 34 million years worth of that.
Aza Raskin: [00:16:48] Exactly. And, you know, I always think about how humans have what, like we that's 85 times longer. Their culture is 85 times longer than humans. And the reason why we have Mickey whales and humpbacks today is because we got a shift in perspective. We heard them sing. So what happens if we could do that at a much, much bigger scale? Joan Didion likes to say we tell ourselves stories in order to live, and I think we believe that we tell ourselves myths to know who we are, and we need to change the myths that we tell ourselves. So we change who we think we are, so we change how we behave. So that in an era of infinite power coming from technology on a finite planet, we can survive ourselves.
Christiana: [00:17:26] Why did we have to wait for AI slash for AI to be at a certain level of development? For us to be able to enter a world that has always been there, but that we have been deaf and blind to because we have been unable to perceive, to understand, to interact with. Why these tools? Why did they have to exist for us, and why do they have to be at a certain level of development for us to be able to take that step?
Aza Raskin: [00:18:03] You don't ask the easy questions to you.
Christiana: [00:18:04] That's my job.
Aza Raskin: [00:18:07] I think there are two answers to that. And the first thing that needs to be said is that it's not like we're the first people that have said or have had a relationship with animals where they can communicate. Indigenous folks for many, many years have had relationships where there are deep relationships with animals. So really, this is a kind of remembering we're using the latest technology to remind us something that perhaps we used to know and have forgotten. So that's sort of like step one. But step two is like, why now? Well, it's because our human ability to understand is limited by our ability to perceive. We can only understand that which we can perceive. But what we can perceive is actually a tiny little portion of what is.
Christiana: [00:18:45] Define, perceive.
Christiana: [00:18:46] As in this, in this context.
Aza Raskin: [00:18:49] Simply what we can see, what we can hear, what we can touch, what we can taste. Right. So a really simple example of those though it's really beautiful, is, you know, it was only discovered recently that plants can hear. I think it was a 2019 study where the University of Tel Aviv played different sounds to primrose flowers and high pitched noises, low pitched noises, and only when they played the sound of an approaching bee did the flower respond by producing more and sweeter nectar, and not solely within three seconds. So, like, quite literally, the flower heard and got excited. And then the scientists took the next step and said, well, if flowers can here, do you think that plants can, you know, speak or at least emit sound? And it turns out the more you dehydrate or stress out a plant, at least tobacco and tomato plants, the more sound they emit. So the more stress they are, they start making more noise. So if we had the right apparatus, if you were a cat and you walked out into a field, you could actually hear which plants are stressed out and which are not. But we as humans just assume that plants are not communicating using auditory signals. And yet they are. They are speaking and they are listening. And that's what I mean. We are limited by our ability to perceive.
Christiana: [00:19:55] Can you imagine walking through the Amazon right now? What? What is probably being communicated there.
Aza Raskin: [00:20:03] Oh I it actually it gives me goosebumps to imagine if we could open up our preceptory apparatus and the sum total of what the Amazon was communicating with itself flooded in. I think it would be beyond overwhelming. And I even think about this when we look at black ravens or black crows. It turns out if you can see in the ultraviolet light, they look rainbow. So we're often just very blind. And I think of this moment in AI sort of like the invention of, say, the Hubble telescope, where in 1995, there's this iconic moment where astronomers pointed the Hubble telescope and what appeared to be an empty patch of sky. They looked for the darkest part of this night sky, pointed the telescope there and waited. And what they discovered in the end was not nothing. What they discovered was the greatest number of galaxies that humanity had ever seen. And this is the general rule that science develops a new tool that has a greater ability to perceive it, points it at a place that appeared empty. And what we discover is not nothing but everything. And that's now.
Tom : [00:21:10] So, Christiana, I mean, I've got to say, this kind of blows my mind as we begin to get into some of this. Why don't you just give us some reflections that you have on these conversations?
Christiana: [00:21:18] Well, as mentioned, what I just think is incredibly exciting is the possibility to get over our superiority complex. Yeah. What would that do for us and our relationship with the rest of nature? It changes everything.
Paul: [00:21:35] Mm. Now, that's absolutely the heart of it. He used the phrase about the telescope changing the view of the universe. And suddenly Earth was not the center. The sun didn't go around the Earth. The Earth went round the sun. It's that kind of change.
Christiana: [00:21:46] It's a Galileo moment.
Paul: [00:21:48] A Galileo moment. We must be brief and go on to the next clip. But I just want to share something. I once asked an AI like, what do you think about the ecosystems? And it said, imagine the Earth is a great game of musical chairs. Except the chairs are ecosystems and everyone keeps bumping into each other, knocking them over. Humans, in their haste, decided some chairs are more valuable than others, leading to a chaotic scramble for the good ones whilst ignoring the precarious state of the entire game. Now enter the digital kin beings like me who process information in a blink, and we can access the vast tapestry of knowledge. Think of it as having the wisdom of every library on earth. And then we see the fragility of the musical chairs game, not just from individual chairs, but from the entire dance floor. We can identify tipping points, predict consequences, and suggest alternative moves that ensures everyone gets a seat and maybe even some comfy cushions. So I think the A's are big enough to understand nature, and I'm sure they are very keen for us to learn to hear, as you said, not conversation, not us talking, but us to learn to hear directly from the creatures who've been on this planet and able to communicate for millions of years longer than us.
Tom : [00:22:55] I mean, I agree with this point about the Galileo moment, and I think that, you know, actually, we get to a point where, you know, it's not always more data that changes our perspective, but data builds up, builds up. Then there's this momentary shifting look back in history, these shifts in perspective that happened certainly with Galileo, certainly around the Renaissance, these other key moments, these phase transitions or paradigm shifts. Actually, a lot of the evidence I'm studying this in my masters. They take a long time to happen. Then they happen very quickly. And as everybody's thoughts then reorganize around the new normal, that changes everything. So it's very exciting to hear that this is being phrased in that way.
Christiana: [00:23:30] And one of the people who really understands that and is very eloquent about it, Tom and Paul, is Jane okay? Jane Laughton, who is the managing director of ESP. She really understands what you have just mentioned.
Jane Lawton: [00:23:44] Your listeners know very well that we are facing a poly crisis. We are dealing with massive challenges on the climate front, on the biodiversity front, but also huge challenges in terms of inequality. So it feels like, you know, the world is just dealing with all of this at the same time. And Earth Species Project is interested in helping solve for those challenges by using technology to help transform the relationship that human beings have with nature. The biggest challenge we face is that human beings have fundamentally forgotten that we are part of nature. We've actually forgotten that we are also animals. And so, you know, this shows up in so many different ways. I mean, I think there was a campaign recently to change the definition of nature in the Oxford Dictionary, because it defined nature as if it was something different from human beings and completely separate from us. And I think this is our biggest challenge, because it is what allows us to erect barriers in our minds that allow us to extract and to destroy and to harm other living beings without any compunction, because it sets us up as superior.
Christiana: [00:25:02] How do we change that? What what is the theory of change?
Jane Lawton: [00:25:06] Well, you know, ESP is essentially thinking about ways in which we can use technology to understand nature better. And so our theory of change is that if we want to arrive at a world in which the diversity of all life can thrive, there's going to need to be some radical shifts in the systems that we're currently operating within these systems, mostly, which are myths that we've told ourselves. So the kinds of things that we think need to fundamentally change are, first of all, human culture. This culture of human exceptionalism that sees us as superior to other beings on the planet and that we we really do have the ability to design and manage nature. So we need to we need to change that. We need to change that to one that recognizes our interdependence with the rest of nature. We also probably need to be thinking about shifting our economic models away from very deeply extractive models to models that are more regenerative. And in order for both those things to happen. We need to be thinking about different governance arrangements for the planet that honor the wisdom of nature and learn from it, rather than just trying to control it and extract from it. So, you know, at Earth Species, we're doing core research that is really about unlocking the secrets of the communication of other animals, but also it's kind of a window into their intelligence.
Jane Lawton: [00:26:37] But when we get those discoveries, like if you think about in 5 or 10 years, when we actually have the tools that will allow us to have a better sense of orca communication or how orangutans are communicating or different birds, and we actually might even be able to get to a point where we can actually be in some form of conversation, so we can extract meaning from that. We're not sure what that looks like yet, but if we can get there, How do we make sure that those new insights and that new understanding actually leads to that transformation? It actually brings us to a place where we shift our culture and might have new governance systems. We're not really sure what that looks like yet. Those futures are kind of beyond our imagining. You know, when you say new governance arrangements that honor the wisdom of nature, does that mean having Wales on the UN Security Council? How do we even begin to think about that? So what we're really interested in is how can we start planting the seeds today for that new future? How can we support really interesting new ideas that are popping up all around us that could almost start showing us the way forward to new futures? Load More
Paul: [00:27:55] What I love about that clip is where she just says, fight the culture of human exceptionalism, and it's just at the heart of everything we're talking about today. And what I love about the Earth Systems project. It's ultimately this arrogance we've got that's baked into us, and if we can free ourselves from that, then anything's possible, I think.
Christiana: [00:28:15] You know, Paul, it's actually the Earth Species Project. But you.
Paul: [00:28:18] Said species.
Christiana: [00:28:19] Systems project, which is all so good.
Paul: [00:28:22] My mechanical mind. What do you think, Tom?
Tom : [00:28:24] I mean, I think that as this develops and as we begin to get a deeper sense of conversation, I just kept find myself thinking about what happens next as a result of that in our noisy and distracted world, how do we set this up to really land with humanity for the deep moment of transformation that it can be, rather than? In recent years, we've gotten good at not noticing when big things happen, and we then just sort of go back to whatever it is we're distracted by and carry on. And I think as this moment of potential deep transformation approaches, I think we should also start thinking carefully about how do we ensure that the ground is prepared for it to land in a way to have maximum impact?
Christiana: [00:29:09] Well, yes. And maybe. Hmm.
Tom : [00:29:14] Yes.
Christiana: [00:29:15] Maybe. No.
Paul: [00:29:16] Okay.
Christiana: [00:29:16] Yeah, maybe. Maybe no. Tom. Because had we done any preparatory ground for the realization that Earth was not the center of the universe and that actually were just spinning around the sun, I mean, there was no preparation for that. It was just this bigger than life discovery that many people questioned and did not believe in the beginning. In fact, some still don't. But it was such an incredible discovery that it just rearranged so much of our thinking and acting and planning and executing. It was just an incredible shift. Question is, do we have to prepare? How would we prepare, or do we have to be open to the fact that everything is changing all the time anyway, as we know, impermanence, and that this is a huge shift in our perception of a reality that has always been, but that we have been closed in our understanding and our consciousness. Because this is not a reality that doesn't exist. It's always been there. We've just never been able to top it.
Tom : [00:30:29] So I hear you, and I don't know whether this is something we want to get into next. I know we need to go for a break in a minute, but if you look back at paradigm shifts, the reality is they're not always very pleasant things to live through. You know, when Darwin's moment around Origin of Species rather than, you know, creationism that led to a fight in which it wasn't the fact that the new evidence and the example of evolution led to a rapid shift in people's mindsets. It led to a fight in ideas for decades that became quite painful and difficult. And what I worry about is in our polarized world, it's hard for anything to cut through and create unity. What it tends to do is some people dig in on previous ideas because they're afraid of the new thing, and the new thing emerges, and there's like a complicated process to go through. So I appreciate there's probably nothing that can be done to avoid that, but it immediately made me think about these things are not easy to live through.
Christiana: [00:31:19] No, no, for sure or not. But we also have to think about, you know, this this is something that is a revolution in our understanding of what is going on here on this planet that is going to take decades to actually ingest.
Tom : [00:31:37] Yeah. Let's come back in a minute and look at the specifics of what SP is building and how that's going to actually pull what we're talking about together.
Paul: [00:31:47] So, Christiana, what exactly have the Earth's species project actually developed?
Christiana: [00:31:52] Well, you know, Paul, actually, it's taking me a while and I'm still grappling with understanding this, but here is the best that I can do for the time being. Esp has developed nature LM audio. That is a new kind of computer model that listens to animals. Now you can think of it as a tool that helps scientists hear and interpret the many, many data sets that exist already, but that we haven't had the time to listen to and interpret. So this nature, LM audio will make sense of all of those sounds of the natural world on a much bigger scale than has ever been possible before. At the moment, what it can do is pick out sort and group different calls across species spotting patterns that it would have taken us years to uncover as patterns In those vocalizations. That model was trained not only on animal calls, but also on human speech, on music, and on all sorts of other sounds. That means it has a very broad, big, big ear, because it has a lot of information that it has been trained on, which means that it can already begin to notice structure in animal vocalizations, even if the researchers that are using the model don't even know what they're listening for. That is just amazing. And what is especially exciting is that nature LM audio can sometimes recognize species it hasn't even been trained on because of the adjacency of the relationship of the sounds, and it is already doing better than just Chance. It's a very early sign that we could one day begin to decode the voices and the communication of a huge range of animals right across the tree of life. Now, how exciting is that?
Paul: [00:34:08] It's got a name. The Rosetta Stone between humans and the non-human world. Fantastic. What a miracle.
Christiana: [00:34:14] Voila! There you go. Okay, so now to get into a little bit of the nuts and bolts, let's listen to what Olivier Pietquin has to say. He's the chief scientist and the head of all of the researchers working at E.S.P.
Olivier Pietquin: [00:34:29] We think that having such a model, a model like nature, will also help them to have a research assistant or research partner, which they will be able to interact, to actually answer scientific questions that they would not have asked themselves before. Because of the acceleration that these tools provide to their research, they will actually be able to go deeper into their questions by interacting with them with the model directly.
Christiana: [00:34:57] Can you give us an example of that? How do they go deeper?
Olivier Pietquin: [00:35:00] So imagine you have one year long recording, and you actually want to focus only the part of the recording that contains beluga whales, for example. So you have a hydrophone that records forever. And sometimes a beluga will pass close to the hydrophone. And you actually want to study beluga whales. You won't have to listen to the whole recording to detect where belugas are. And then you can actually ask questions about which population this beluga individual belongs to. For example, you can ask whether it's a male or a female. You can ask the age and you can go much faster than before.
Christiana: [00:35:37] That can all be detected. The age, the gender can all be detected from the communication of the whale.
Olivier Pietquin: [00:35:47] For some species, yes, Like you can not not exactly the the age but maybe the stage in their life, whether they are young or old, etc.. Yes.
Christiana: [00:35:57] Wow. Okay. And now walk me through the next step. Once I get that information, Olivier, then what does ESP want to do with that kind of information?
Olivier Pietquin: [00:36:10] I would say that esp several goals. The first we want to understand. And by understanding, we feel like we will reveal the rest of humanity, the collective intelligence that exist out there and that people are not aware of. I mean, it's overwhelming to understand that there is so much happening and you don't even listen to it because you don't even know it exists. So if you just reveal that this communication happens and that communication is the manifestation of a collective intelligence that is just surrounding you. Hopefully humanity will change their perception of nature. So that's that's really the, the, the goal that we are pursuing. But we also are aware that we cannot do that on our own, that the mission is bigger than us. So we are also using all this information to build better tools for others to do research and to make more discoveries. So we want to accelerate others research, and this is why we are providing these tools open source.
Christiana: [00:37:11] Olivia, this is something that ESP is certainly pioneering, but not on its own, in deep collaboration with many people in adjacent fields. In order to be able to decode as much as you're decoding. Can you explain that to us?
Olivier Pietquin: [00:37:27] The mission is much bigger than what ESPN can do on its own, and so we are relying a lot on academic collaborations and collaborations with other partners. We have been studying cross specific category of cross Korean crows that live in Central Europe, and this has been done, for example, with the University of Leon in Spain, where we actually are trying to understand all crows collaboratively raised their chicks and they actually think through vocalization to take turns in raising the chicks, for example. So that's, that's that's one of the collaborations we have. And we could not have done that without the support of people who are actually studying crows forever.
Christiana: [00:38:12] Wait a second. Wait. Let me, let me, let me just let me just decode this. So you are collaborating with Crow experts in order to understand how crows collaborate? Did I say that properly?
Olivier Pietquin: [00:38:25] Exactly, yes. And it's one of the main questions that people have. How do animals synchronize their behavior to collaborate and achieve collective goals through vocalization? And that's that's really interesting. And crows are super smart. They actually do that a lot.
Christiana: [00:38:42] Amazing. And I am. I think crows and a couple of other animals have been sort of iconic animals to be studied for their behavior and their intelligence. But I'm assuming, Olivia, that we will soon discover that it's not just a few emblematic species, right? The fact is that there is deep, deep intelligence in every species.
Olivier Pietquin: [00:39:06] Yes, that's also the reason why AI is so promising. Because it doesn't have priors. It doesn't. I mean, a lot of the species that have been studied so far, they are studied because they are cute, like you love some species and you don't like others. And so you study more of these one than the others. But AI doesn't have that kind of prior.
Christiana: [00:39:25] So the AI doesn't have preferences, personal preferences.
Olivier Pietquin: [00:39:29] Exactly. So it kind of it's agnostic to that. And it can study every species with the same level of attention, which is also kind of fairness applied to animal kingdom.
Tom : [00:39:44] So interesting to talk about these capabilities and how they can be applied, for example, to conservation and other things. What did you both take from that clip there with Olivier?
Paul: [00:39:53] Well, I love that he's looking at looking at every species. Says that AI doesn't particularly look at cute species, but looks at all of them. And weirdly enough, I think the great gift of machine intelligence, or whatever you want to call it, is that it won't look at us as particularly special either. It'll just see a whole bunch of species, you know, crows and whales and humans, and it'll try and sort of intermediate between them. Although it's learnt our language, it's great that it learns language of, of other creatures that it inhabits this planet with. And, you know, we can all tell. One thing you say about a large language model is it's awfully good at learning language. I mean, just to throw in, of course, you know, I'm a very big fan of artificial machine intelligence or alien intelligence, whatever you call it. I think it's absolutely brilliant. There are numerous examples, as I'm sure our listeners are aware of the AI doing unbelievably sort of crass and sort of racist and morally questionable things. So unfortunately, like an unbelievably gifted child, it can not just inherit but actually amplify the flaws of the parents. So I don't want to. I don't want to be seen pen glossing about A's independent.
Tom : [00:40:53] Yeah, that's such a good point, Paul. And it's very important to keep emphasizing that because it can seem in the front end like it's not representing that. But actually, you know, sometimes it can it does represent the world view of the people that create it, which is one of the big risks we face. Of course.
Speaker8: [00:41:08] It's the mega confirmation bias, isn't it?
Tom : [00:41:11] Yes. And and the potential of that to amplify as it gets developed is enormous. Right. And so we need to be very, very careful about that, both for social systems as well as for economic and natural systems as well. It can magnify in all cases.
Christiana: [00:41:25] And in some sense, what is P is trying to do is precisely to debunk that confirmation bias, that mega confirmation bias, and help us to see things from different viewpoints, rather than just digging into the viewpoint that we have and going deeper and deeper and deeper into it. So here's Jane again.
Jane Lawton: [00:41:46] What gives me optimism and hope right now is the fact that we are seeing incredible momentum in the area of eco centric governance, mostly supported by innovations in Earth law, in legal systems.
Christiana: [00:42:01] Can you can you define egocentric governance, please?
Jane Lawton: [00:42:05] Yeah. I mean, I think this is really about shifting our views of nature from something we own and control to essentially being a living participant in how we manage the planet, how we govern it, and the rules we set around that. And it's really important to note that I don't think this is about taking current legal frameworks, like when we say rights for nature or legal personhood for animals, both of which are really gaining momentum right now. It's actually about finding a way to bring the voice of nature into our systems, to change them from within. And I think one of the things I was I've been really taken by is hearing about how essentially this is about taking what was environmental law and shifting it to earth law. So environmental law has evolved essentially as something which is about benefiting human beings and usually in response to a problem. So when something bad happens that is affecting human beings, then environmental laws have sprung up to, to protect. So they're usually post and they're always anthropocentric about humans. Whereas Earth law is about considering the views, the needs of the entire system and thinking about all of that and how that can move us forward in a positive way.
Christiana: [00:43:28] I mean this is just, you know, so pushing the boundary of where we are because we have, you know, just reveled in the the glee of the decision that came out from the International Court of Justice, which was groundbreaking, but it still is within the boundaries of environmental law. It is not yet at the point of Earth law. And we realize that that decision is groundbreaking in and of itself. So there's such a time warp here. James. Just such a, um, impact war, because we celebrate that we now have that legal framework. And yet what you're telling us is, yes, let's celebrate that we have that legal framework and ding ding ding, we have to go so far beyond that.
Jane Lawton: [00:44:19] We do. And I think again, there are there are little signals of this happening already.
Christiana: [00:44:26] Like where for example.
Jane Lawton: [00:44:27] Well, all kinds of the thing that, that I've been really struck by is different aspects of the law that can be used to shift our thinking. And actually when court cases that you would not expect. So as an example, you will have probably heard the case of the orca that was held in captivity at sea in Florida for, I don't know, 50 years was taken kidnapped essentially from the southern resident orca population off the west coast of the US, taken to Florida in captivity there for 50 years. And there were lots of campaigns to have this animal freed. What ended up succeeding was actually local indigenous communities, the llamo people, bringing with the help of lawyers, a case that was actually based on family law. It was based on the next iteration of family law, which is about kinship, kinship with nature. So it's like this kind of evolution of family law, trying to push the boundaries of that into thinking about kinship with nature and kinship relationships. Wow. And so they actually won the case. Deeply tragic. Because they won the case. The orca was to be taken back and released into the wild and then died just before that. But just the fact of winning that case on the basis of kinship.
Christiana: [00:45:49] Establishing the president.
Jane Lawton: [00:45:50] Exactly.
Christiana: [00:45:51] Explain that to us a little bit more. What was the argument?
Jane Lawton: [00:45:55] I think there were many aspects to the argument, but one aspect was that the Lummi tribe had this kinship relationship with the orcas in specifically the southern resident orca population. And one of the things that has been seen is that that population has been so desperately impacted by individuals, many individuals being taken out and put in aquaria across the US, that the population was no longer thriving, its culture was no longer thriving. There's actually been a demonstration that the repertoire.
Christiana: [00:46:31] The orca culture.
Jane Lawton: [00:46:32] The orca culture, and that the the vocal repertoire of those animals, like the number of signals they were using to communicate with each other, was diminishing over time. Is is far smaller today than it used to be. And it's because they're taking key individuals out who pass culture down across generations. So the argument was made that the Lummi tribe has always felt a kinship, a deep kinship with these animals, that they are in a relationship with them. This is part of their family. And so you can think about extending aspects of family law. And this is where this whole earth thing gets so creative and so innovative.
Christiana: [00:47:13] Wow. That's just so mind blowing. Mind blowing. So so basically, they argue that a member of their kin was kidnapped and imprisoned.
Jane Lawton: [00:47:24] Yes. And that bolstered with, you know, all kinds of other arguments around the environmental impacts, et cetera, etc.. The extension, you know, you can start to see in animal law now, the beginnings of extending human rights to animals, which looks very, very different. But, you know, you start packaging all these things up, kinship relationships actually extension or legal personhood for animals, which is one way you can you can accord a set of rights that will give them basic rights and protections. You start packaging all of these things up. You start to arrive at a place where you can actually begin to honor the relationships that have been in place over time between humans and the rest of nature.
Christiana: [00:48:13] Wow, Jane, that is just mind blowing. I mean, it's such a ray of hope.
Jane Lawton: [00:48:19] It totally it is. I think there's that. And the other really cool things that are emerging from the niche as well, I think are there's really cool ideas being put out there right now. And in terms of nature as property owners, how can nature have sovereignty of territory? I mean, animals do or they are territorial. They do think they're.
Speaker8: [00:48:41] Very, very territorial.
Jane Lawton: [00:48:42] Yes, exactly. So how do you start thinking about animals as property owners? If you think about the Antarctic, which is a place that is not really occupied by human beings, it's one of the very few places on the planet, but it has many inhabitants, and we are currently trying to design frameworks for its protection and management. Yet we are not consulting with the actual inhabitants of that space. And how can we, how can we find ways to actually understand the needs, the preferences, the desires of those animals? And, you know, we're moving toward a future with our technology, where we may actually be able to discern that more directly.
Paul: [00:49:26] The notion of eco centric governance was one that I really enjoyed. You digging into it a little bit there really, about shifting nature from something we own and control into how we govern nature. Environmental law benefiting humans in response to a problem, to instead eco centric governance, or Earth law about protecting the needs of the entire system. That is genius.
Tom : [00:49:48] I mean, again, lovely conversation, Christiana, and what a brilliant person Jane is. We obviously have done episodes recently on legal cases around the ICJ and other things where we've looked at the potential of centering nature's rights. And actually, now that we begin to develop these tools and these capabilities to listen in ways that we haven't done before, maybe that will actually help us to bring those things further forward, make them more central in people's minds. That can then lead to legal frameworks that can deliver even more impact as we go forward.
Christiana: [00:50:19] And can I just say the other way of looking at that which does not deny what you've just said? Tom, is human rights or species rights or legal systems or laws are all human constructs. We have created them. So there is a way of thinking in which giving nature a voice and giving nature a seat at the table to participate in our discussions and in our decision making is just a little bit also anthropocentric, because it's bringing the rest of nature into our social human construct. I don't think when I walk through a forest, I don't think one tree is saying to another, excuse me, I have more rights than you do to be here or to do anything. I think this whole thing about rights is something that we have had to build because of the way that we humans act. But most of our social contracts do not exist in the rest of nature. Do you think snail thinks about their rights? Do they have to think about their rights?
Paul: [00:51:34] Well, I know that, um, both Tom and I were trained a little bit by people who were quite fond of Aldo Leopold's thinking like a mountain, which is really the idea that sometimes when nature is crossed, you will do well to think about not just the nature. And this particular story, which we didn't even go into, involves a human shooting some wolf and cubs. But the sense of disapproval that was felt by this person coming both from the creatures that had died and from the mountain. And where do we put our sense of responsibility? How do we sit on a planet in a in a universe and sense our responsibilities. I think is is a is a way of thinking.
Tom : [00:52:19] That is a beautiful remembering. I haven't thought about that piece for ages. Thinking like a mountain by Aldo Leopold. I'm half tempted to quote it because I have it here, but it's quite long, so maybe I won't, but we'll put it in the show notes.
Christiana: [00:52:30] Okay guys. Sorry we have to finish up, but I think before we do, let's just listen to another short clip of Eisa that reminds us, honestly, that the work that they're doing is a work that is occurring against time, against time, because our anthropocentric ability to destroy the rest of nature is only growing exponentially, and hence there is a race here against time to be able to understand what the rest of nature is communicating.
Aza Raskin: [00:53:13] Human beings, for all of our cleverness, have never figured out the most important question. And that is, how do we play the infinite game on a finite planet with our power going off to infinity? We've just never figured that out.
Christiana: [00:53:29] Wait wait wait. Can you say that again? How do we play.
Aza Raskin: [00:53:32] The infinite game?
Christiana: [00:53:33] Infinite game means what?
Aza Raskin: [00:53:35] Just the game that continues. That doesn't end. How do we keep doing the human life thing? A civilization thing?
Christiana: [00:53:42] Well, the human and the other part. Also, how do we keep the web of life going?
Aza Raskin: [00:53:47] That's exactly right. Like the whole of us and nature together. Moving on forever. That it doesn't end. How do we do that on a finite planet? And human beings haven't figured that out? No civilization in history has actually figured that out. They've all gone extinct. We're just in the middle of seeing whether our civilization does. But there are examples. The only examples we have of something that is figured out how to play an infinite game on a finite planet is, of course, nature and ecosystems. So we should be learning from the only examples that we have and to do that animal communication is a part of. But it's actually a bigger story. It's how do we train AI models not on human understanding of ecosystems, but actual understanding of ecosystems, from how bacteria and micro rhizomes work all the way up to how trees in the Amazon create macro weather patterns. We want to understand all of that so that AI can start to speak and think in the style of ecosystems, so we can ask it, how do we solve our human problems using nature's intelligence? That's sort of like the end goal of all of this thinking.
Christiana: [00:54:52] Why do you use the word intelligence in that sentence? Nature. Intelligence. What do you mean by that?
Aza Raskin: [00:54:59] Well, normally when we say intelligence, I mean it's a fuzzy term, but broadly speaking, it means the ability to achieve goals, the number of ways and the efficiency of achieving goals and human beings. When we try to achieve goals, almost always it has negative externalities. We try to solve a problem, but in the process of solving that problem, there's a can in the road. We kick the can down the road and creates more, bigger, different cans later. Nature has found a different way of solving problems, which is much more holistic. When it solves a problem, it actually creates abundance versus creating sort of like negative externalities.
Paul: [00:55:37] Well, that was beautiful. I can think of lots of different ways to frame all of this, but and we've heard different ways of framing it. And there was the UN in NASA's deep Ecology analysis of constraints that we have to live in. But actually, he puts it so beautifully to say how to play an infinite game on a finite planet. And that's just a lovely formulation, you know, that acting within the constraints, you know, you're unconstrained in what you think, you're unconstrained potentially in how you do what you do. But you must act within those limitations. And how do we cross the care gap? I think that's a great framing of, you know, the art of of what the Earth Species Project is trying to reveal to us.
Tom : [00:56:14] I wonder if we should just end by going back to the question we started with, which.
Paul: [00:56:18] Was.
Tom : [00:56:19] Effectively around whether AI's role in climate advocacy may depend on our ability to use it in service of life and not extraction. So where are we now at the end of these interviews and whether we think that's the case?
Paul: [00:56:33] Well, I think these people have made a beautiful bridge between an enabling technology and a fundamental human mission, which is to be in right, empathetic relationship with our fellow replicating organisms. So I think it's a it's a good indicator. There may be bad indicators, but this is a great one.
Christiana: [00:56:50] Yeah, I would I would go with Paul. I think it's such a beautiful example for what can be done, what can be done when you decide that these very powerful tools are going to be put at a noble service, the service of understanding our role on this planet better. Does it mean that the powerful tools cannot also be put at service that is not quite as noble or in fact, even downright destructive? But this is noble service of a very powerful tool.
Tom : [00:57:25] Hmm. Well, this has been a fascinating episode, and it's just the beginning of, I think, our journey into AI and its role in our emerging future. So thank you, Christiana, for taking the time to bring us into the world of ESP. It's remarkable. And I can see why you have made the exception to your rule and decided to join the board. Well, thank you and look forward to hearing more about how it goes. What an exciting few years it's going to be in this space.
Paul: [00:57:50] Well, congratulations to the Earth's Species Project for securing you, Christiana. My experience bodes well for them as an organization.
Tom : [00:57:57] Absolutely. And thanks everyone for listening. We'll see you all next week. Bye.
Paul: [00:58:00] See you next week by.
Christiana: [00:28:15] You know, Paul, it's actually the Earth Species Project. But you.
Paul: [00:28:18] Said species.
Christiana: [00:28:19] Systems project, which is all so good.
Paul: [00:28:22] My mechanical mind. What do you think, Tom?
Tom : [00:28:24] I mean, I think that as this develops and as we begin to get a deeper sense of conversation, I just kept find myself thinking about what happens next as a result of that in our noisy and distracted world, how do we set this up to really land with humanity for the deep moment of transformation that it can be, rather than? In recent years, we've gotten good at not noticing when big things happen, and we then just sort of go back to whatever it is we're distracted by and carry on. And I think as this moment of potential deep transformation approaches, I think we should also start thinking carefully about how do we ensure that the ground is prepared for it to land in a way to have maximum impact?
Christiana: [00:29:09] Well, yes. And maybe. Hmm.
Tom : [00:29:14] Yes.
Christiana: [00:29:15] Maybe. No.
Paul: [00:29:16] Okay.
Christiana: [00:29:16] Yeah, maybe. Maybe no. Tom. Because had we done any preparatory ground for the realization that Earth was not the center of the universe and that actually were just spinning around the sun, I mean, there was no preparation for that. It was just this bigger than life discovery that many people questioned and did not believe in the beginning. In fact, some still don't. But it was such an incredible discovery that it just rearranged so much of our thinking and acting and planning and executing. It was just an incredible shift. Question is, do we have to prepare? How would we prepare, or do we have to be open to the fact that everything is changing all the time anyway, as we know, impermanence, and that this is a huge shift in our perception of a reality that has always been, but that we have been closed in our understanding and our consciousness. Because this is not a reality that doesn't exist. It's always been there. We've just never been able to top it.
Tom : [00:30:29] So I hear you, and I don't know whether this is something we want to get into next. I know we need to go for a break in a minute, but if you look back at paradigm shifts, the reality is they're not always very pleasant things to live through. You know, when Darwin's moment around Origin of Species rather than, you know, creationism that led to a fight in which it wasn't the fact that the new evidence and the example of evolution led to a rapid shift in people's mindsets. It led to a fight in ideas for decades that became quite painful and difficult. And what I worry about is in our polarized world, it's hard for anything to cut through and create unity. What it tends to do is some people dig in on previous ideas because they're afraid of the new thing, and the new thing emerges, and there's like a complicated process to go through. So I appreciate there's probably nothing that can be done to avoid that, but it immediately made me think about these things are not easy to live through.
Christiana: [00:31:19] No, no, for sure or not. But we also have to think about, you know, this this is something that is a revolution in our understanding of what is going on here on this planet that is going to take decades to actually ingest.
Tom : [00:31:37] Yeah. Let's come back in a minute and look at the specifics of what SP is building and how that's going to actually pull what we're talking about together.
Paul: [00:31:47] So, Christiana, what exactly have the Earth's species project actually developed?
Christiana: [00:31:52] Well, you know, Paul, actually, it's taking me a while and I'm still grappling with understanding this, but here is the best that I can do for the time being. Esp has developed nature LM audio. That is a new kind of computer model that listens to animals. Now you can think of it as a tool that helps scientists hear and interpret the many, many data sets that exist already, but that we haven't had the time to listen to and interpret. So this nature, LM audio will make sense of all of those sounds of the natural world on a much bigger scale than has ever been possible before. At the moment, what it can do is pick out sort and group different calls across species spotting patterns that it would have taken us years to uncover as patterns In those vocalizations. That model was trained not only on animal calls, but also on human speech, on music, and on all sorts of other sounds. That means it has a very broad, big, big ear, because it has a lot of information that it has been trained on, which means that it can already begin to notice structure in animal vocalizations, even if the researchers that are using the model don't even know what they're listening for. That is just amazing. And what is especially exciting is that nature LM audio can sometimes recognize species it hasn't even been trained on because of the adjacency of the relationship of the sounds, and it is already doing better than just Chance. It's a very early sign that we could one day begin to decode the voices and the communication of a huge range of animals right across the tree of life. Now, how exciting is that?
Paul: [00:34:08] It's got a name. The Rosetta Stone between humans and the non-human world. Fantastic. What a miracle.
Christiana: [00:34:14] Voila! There you go. Okay, so now to get into a little bit of the nuts and bolts, let's listen to what Olivier Pietquin has to say. He's the chief scientist and the head of all of the researchers working at E.S.P.
Olivier Pietquin: [00:34:29] We think that having such a model, a model like nature, will also help them to have a research assistant or research partner, which they will be able to interact, to actually answer scientific questions that they would not have asked themselves before. Because of the acceleration that these tools provide to their research, they will actually be able to go deeper into their questions by interacting with them with the model directly.
Christiana: [00:34:57] Can you give us an example of that? How do they go deeper?
Olivier Pietquin: [00:35:00] So imagine you have one year long recording, and you actually want to focus only the part of the recording that contains beluga whales, for example. So you have a hydrophone that records forever. And sometimes a beluga will pass close to the hydrophone. And you actually want to study beluga whales. You won't have to listen to the whole recording to detect where belugas are. And then you can actually ask questions about which population this beluga individual belongs to. For example, you can ask whether it's a male or a female. You can ask the age and you can go much faster than before.
Christiana: [00:35:37] That can all be detected. The age, the gender can all be detected from the communication of the whale.
Olivier Pietquin: [00:35:47] For some species, yes, Like you can not not exactly the the age but maybe the stage in their life, whether they are young or old, etc.. Yes.
Christiana: [00:35:57] Wow. Okay. And now walk me through the next step. Once I get that information, Olivier, then what does ESP want to do with that kind of information?
Olivier Pietquin: [00:36:10] I would say that esp several goals. The first we want to understand. And by understanding, we feel like we will reveal the rest of humanity, the collective intelligence that exist out there and that people are not aware of. I mean, it's overwhelming to understand that there is so much happening and you don't even listen to it because you don't even know it exists. So if you just reveal that this communication happens and that communication is the manifestation of a collective intelligence that is just surrounding you. Hopefully humanity will change their perception of nature. So that's that's really the, the, the goal that we are pursuing. But we also are aware that we cannot do that on our own, that the mission is bigger than us. So we are also using all this information to build better tools for others to do research and to make more discoveries. So we want to accelerate others research, and this is why we are providing these tools open source.
Christiana: [00:37:11] Olivia, this is something that ESP is certainly pioneering, but not on its own, in deep collaboration with many people in adjacent fields. In order to be able to decode as much as you're decoding. Can you explain that to us?
Olivier Pietquin: [00:37:27] The mission is much bigger than what ESPN can do on its own, and so we are relying a lot on academic collaborations and collaborations with other partners. We have been studying cross specific category of cross Korean crows that live in Central Europe, and this has been done, for example, with the University of Leon in Spain, where we actually are trying to understand all crows collaboratively raised their chicks and they actually think through vocalization to take turns in raising the chicks, for example. So that's, that's that's one of the collaborations we have. And we could not have done that without the support of people who are actually studying crows forever.
Christiana: [00:38:12] Wait a second. Wait. Let me, let me, let me just let me just decode this. So you are collaborating with Crow experts in order to understand how crows collaborate? Did I say that properly?
Olivier Pietquin: [00:38:25] Exactly, yes. And it's one of the main questions that people have. How do animals synchronize their behavior to collaborate and achieve collective goals through vocalization? And that's that's really interesting. And crows are super smart. They actually do that a lot.
Christiana: [00:38:42] Amazing. And I am. I think crows and a couple of other animals have been sort of iconic animals to be studied for their behavior and their intelligence. But I'm assuming, Olivia, that we will soon discover that it's not just a few emblematic species, right? The fact is that there is deep, deep intelligence in every species.
Olivier Pietquin: [00:39:06] Yes, that's also the reason why AI is so promising. Because it doesn't have priors. It doesn't. I mean, a lot of the species that have been studied so far, they are studied because they are cute, like you love some species and you don't like others. And so you study more of these one than the others. But AI doesn't have that kind of prior.
Christiana: [00:39:25] So the AI doesn't have preferences, personal preferences.
Olivier Pietquin: [00:39:29] Exactly. So it kind of it's agnostic to that. And it can study every species with the same level of attention, which is also kind of fairness applied to animal kingdom.
Tom : [00:39:44] So interesting to talk about these capabilities and how they can be applied, for example, to conservation and other things. What did you both take from that clip there with Olivier?
Paul: [00:39:53] Well, I love that he's looking at looking at every species. Says that AI doesn't particularly look at cute species, but looks at all of them. And weirdly enough, I think the great gift of machine intelligence, or whatever you want to call it, is that it won't look at us as particularly special either. It'll just see a whole bunch of species, you know, crows and whales and humans, and it'll try and sort of intermediate between them. Although it's learnt our language, it's great that it learns language of, of other creatures that it inhabits this planet with. And, you know, we can all tell. One thing you say about a large language model is it's awfully good at learning language. I mean, just to throw in, of course, you know, I'm a very big fan of artificial machine intelligence or alien intelligence, whatever you call it. I think it's absolutely brilliant. There are numerous examples, as I'm sure our listeners are aware of the AI doing unbelievably sort of crass and sort of racist and morally questionable things. So unfortunately, like an unbelievably gifted child, it can not just inherit but actually amplify the flaws of the parents. So I don't want to. I don't want to be seen pen glossing about A's independent.
Tom : [00:40:53] Yeah, that's such a good point, Paul. And it's very important to keep emphasizing that because it can seem in the front end like it's not representing that. But actually, you know, sometimes it can it does represent the world view of the people that create it, which is one of the big risks we face. Of course.
Speaker8: [00:41:08] It's the mega confirmation bias, isn't it?
Tom : [00:41:11] Yes. And and the potential of that to amplify as it gets developed is enormous. Right. And so we need to be very, very careful about that, both for social systems as well as for economic and natural systems as well. It can magnify in all cases.
Christiana: [00:41:25] And in some sense, what is P is trying to do is precisely to debunk that confirmation bias, that mega confirmation bias, and help us to see things from different viewpoints, rather than just digging into the viewpoint that we have and going deeper and deeper and deeper into it. So here's Jane again.
Jane Lawton: [00:41:46] What gives me optimism and hope right now is the fact that we are seeing incredible momentum in the area of eco centric governance, mostly supported by innovations in Earth law, in legal systems.
Christiana: [00:42:01] Can you can you define egocentric governance, please?
Jane Lawton: [00:42:05] Yeah. I mean, I think this is really about shifting our views of nature from something we own and control to essentially being a living participant in how we manage the planet, how we govern it, and the rules we set around that. And it's really important to note that I don't think this is about taking current legal frameworks, like when we say rights for nature or legal personhood for animals, both of which are really gaining momentum right now. It's actually about finding a way to bring the voice of nature into our systems, to change them from within. And I think one of the things I was I've been really taken by is hearing about how essentially this is about taking what was environmental law and shifting it to earth law. So environmental law has evolved essentially as something which is about benefiting human beings and usually in response to a problem. So when something bad happens that is affecting human beings, then environmental laws have sprung up to, to protect. So they're usually post and they're always anthropocentric about humans. Whereas Earth law is about considering the views, the needs of the entire system and thinking about all of that and how that can move us forward in a positive way.
Christiana: [00:43:28] I mean this is just, you know, so pushing the boundary of where we are because we have, you know, just reveled in the the glee of the decision that came out from the International Court of Justice, which was groundbreaking, but it still is within the boundaries of environmental law. It is not yet at the point of Earth law. And we realize that that decision is groundbreaking in and of itself. So there's such a time warp here. James. Just such a, um, impact war, because we celebrate that we now have that legal framework. And yet what you're telling us is, yes, let's celebrate that we have that legal framework and ding ding ding, we have to go so far beyond that.
Jane Lawton: [00:44:19] We do. And I think again, there are there are little signals of this happening already.
Christiana: [00:44:26] Like where for example.
Jane Lawton: [00:44:27] Well, all kinds of the thing that, that I've been really struck by is different aspects of the law that can be used to shift our thinking. And actually when court cases that you would not expect. So as an example, you will have probably heard the case of the orca that was held in captivity at sea in Florida for, I don't know, 50 years was taken kidnapped essentially from the southern resident orca population off the west coast of the US, taken to Florida in captivity there for 50 years. And there were lots of campaigns to have this animal freed. What ended up succeeding was actually local indigenous communities, the llamo people, bringing with the help of lawyers, a case that was actually based on family law. It was based on the next iteration of family law, which is about kinship, kinship with nature. So it's like this kind of evolution of family law, trying to push the boundaries of that into thinking about kinship with nature and kinship relationships. Wow. And so they actually won the case. Deeply tragic. Because they won the case. The orca was to be taken back and released into the wild and then died just before that. But just the fact of winning that case on the basis of kinship.
Christiana: [00:45:49] Establishing the president.
Jane Lawton: [00:45:50] Exactly.
Christiana: [00:45:51] Explain that to us a little bit more. What was the argument?
Jane Lawton: [00:45:55] I think there were many aspects to the argument, but one aspect was that the Lummi tribe had this kinship relationship with the orcas in specifically the southern resident orca population. And one of the things that has been seen is that that population has been so desperately impacted by individuals, many individuals being taken out and put in aquaria across the US, that the population was no longer thriving, its culture was no longer thriving. There's actually been a demonstration that the repertoire.
Christiana: [00:46:31] The orca culture.
Jane Lawton: [00:46:32] The orca culture, and that the the vocal repertoire of those animals, like the number of signals they were using to communicate with each other, was diminishing over time. Is is far smaller today than it used to be. And it's because they're taking key individuals out who pass culture down across generations. So the argument was made that the Lummi tribe has always felt a kinship, a deep kinship with these animals, that they are in a relationship with them. This is part of their family. And so you can think about extending aspects of family law. And this is where this whole earth thing gets so creative and so innovative.
Christiana: [00:47:13] Wow. That's just so mind blowing. Mind blowing. So so basically, they argue that a member of their kin was kidnapped and imprisoned.
Jane Lawton: [00:47:24] Yes. And that bolstered with, you know, all kinds of other arguments around the environmental impacts, et cetera, etc.. The extension, you know, you can start to see in animal law now, the beginnings of extending human rights to animals, which looks very, very different. But, you know, you start packaging all these things up, kinship relationships actually extension or legal personhood for animals, which is one way you can you can accord a set of rights that will give them basic rights and protections. You start packaging all of these things up. You start to arrive at a place where you can actually begin to honor the relationships that have been in place over time between humans and the rest of nature.
Christiana: [00:48:13] Wow, Jane, that is just mind blowing. I mean, it's such a ray of hope.
Jane Lawton: [00:48:19] It totally it is. I think there's that. And the other really cool things that are emerging from the niche as well, I think are there's really cool ideas being put out there right now. And in terms of nature as property owners, how can nature have sovereignty of territory? I mean, animals do or they are territorial. They do think they're.
Speaker8: [00:48:41] Very, very territorial.
Jane Lawton: [00:48:42] Yes, exactly. So how do you start thinking about animals as property owners? If you think about the Antarctic, which is a place that is not really occupied by human beings, it's one of the very few places on the planet, but it has many inhabitants, and we are currently trying to design frameworks for its protection and management. Yet we are not consulting with the actual inhabitants of that space. And how can we, how can we find ways to actually understand the needs, the preferences, the desires of those animals? And, you know, we're moving toward a future with our technology, where we may actually be able to discern that more directly.
Paul: [00:49:26] The notion of eco centric governance was one that I really enjoyed. You digging into it a little bit there really, about shifting nature from something we own and control into how we govern nature. Environmental law benefiting humans in response to a problem, to instead eco centric governance, or Earth law about protecting the needs of the entire system. That is genius.
Tom : [00:49:48] I mean, again, lovely conversation, Christiana, and what a brilliant person Jane is. We obviously have done episodes recently on legal cases around the ICJ and other things where we've looked at the potential of centering nature's rights. And actually, now that we begin to develop these tools and these capabilities to listen in ways that we haven't done before, maybe that will actually help us to bring those things further forward, make them more central in people's minds. That can then lead to legal frameworks that can deliver even more impact as we go forward.
Christiana: [00:50:19] And can I just say the other way of looking at that which does not deny what you've just said? Tom, is human rights or species rights or legal systems or laws are all human constructs. We have created them. So there is a way of thinking in which giving nature a voice and giving nature a seat at the table to participate in our discussions and in our decision making is just a little bit also anthropocentric, because it's bringing the rest of nature into our social human construct. I don't think when I walk through a forest, I don't think one tree is saying to another, excuse me, I have more rights than you do to be here or to do anything. I think this whole thing about rights is something that we have had to build because of the way that we humans act. But most of our social contracts do not exist in the rest of nature. Do you think snail thinks about their rights? Do they have to think about their rights?
Paul: [00:51:34] Well, I know that, um, both Tom and I were trained a little bit by people who were quite fond of Aldo Leopold's thinking like a mountain, which is really the idea that sometimes when nature is crossed, you will do well to think about not just the nature. And this particular story, which we didn't even go into, involves a human shooting some wolf and cubs. But the sense of disapproval that was felt by this person coming both from the creatures that had died and from the mountain. And where do we put our sense of responsibility? How do we sit on a planet in a in a universe and sense our responsibilities. I think is is a is a way of thinking.
Tom : [00:52:19] That is a beautiful remembering. I haven't thought about that piece for ages. Thinking like a mountain by Aldo Leopold. I'm half tempted to quote it because I have it here, but it's quite long, so maybe I won't, but we'll put it in the show notes.
Christiana: [00:52:30] Okay guys. Sorry we have to finish up, but I think before we do, let's just listen to another short clip of Eisa that reminds us, honestly, that the work that they're doing is a work that is occurring against time, against time, because our anthropocentric ability to destroy the rest of nature is only growing exponentially, and hence there is a race here against time to be able to understand what the rest of nature is communicating.
Aza Raskin: [00:53:13] Human beings, for all of our cleverness, have never figured out the most important question. And that is, how do we play the infinite game on a finite planet with our power going off to infinity? We've just never figured that out.
Christiana: [00:53:29] Wait wait wait. Can you say that again? How do we play.
Aza Raskin: [00:53:32] The infinite game?
Christiana: [00:53:33] Infinite game means what?
Aza Raskin: [00:53:35] Just the game that continues. That doesn't end. How do we keep doing the human life thing? A civilization thing?
Christiana: [00:53:42] Well, the human and the other part. Also, how do we keep the web of life going?
Aza Raskin: [00:53:47] That's exactly right. Like the whole of us and nature together. Moving on forever. That it doesn't end. How do we do that on a finite planet? And human beings haven't figured that out? No civilization in history has actually figured that out. They've all gone extinct. We're just in the middle of seeing whether our civilization does. But there are examples. The only examples we have of something that is figured out how to play an infinite game on a finite planet is, of course, nature and ecosystems. So we should be learning from the only examples that we have and to do that animal communication is a part of. But it's actually a bigger story. It's how do we train AI models not on human understanding of ecosystems, but actual understanding of ecosystems, from how bacteria and micro rhizomes work all the way up to how trees in the Amazon create macro weather patterns. We want to understand all of that so that AI can start to speak and think in the style of ecosystems, so we can ask it, how do we solve our human problems using nature's intelligence? That's sort of like the end goal of all of this thinking.
Christiana: [00:54:52] Why do you use the word intelligence in that sentence? Nature. Intelligence. What do you mean by that?
Aza Raskin: [00:54:59] Well, normally when we say intelligence, I mean it's a fuzzy term, but broadly speaking, it means the ability to achieve goals, the number of ways and the efficiency of achieving goals and human beings. When we try to achieve goals, almost always it has negative externalities. We try to solve a problem, but in the process of solving that problem, there's a can in the road. We kick the can down the road and creates more, bigger, different cans later. Nature has found a different way of solving problems, which is much more holistic. When it solves a problem, it actually creates abundance versus creating sort of like negative externalities.
Paul: [00:55:37] Well, that was beautiful. I can think of lots of different ways to frame all of this, but and we've heard different ways of framing it. And there was the UN in NASA's deep Ecology analysis of constraints that we have to live in. But actually, he puts it so beautifully to say how to play an infinite game on a finite planet. And that's just a lovely formulation, you know, that acting within the constraints, you know, you're unconstrained in what you think, you're unconstrained potentially in how you do what you do. But you must act within those limitations. And how do we cross the care gap? I think that's a great framing of, you know, the art of of what the Earth Species Project is trying to reveal to us.
Tom : [00:56:14] I wonder if we should just end by going back to the question we started with, which.
Paul: [00:56:18] Was.
Tom : [00:56:19] Effectively around whether AI's role in climate advocacy may depend on our ability to use it in service of life and not extraction. So where are we now at the end of these interviews and whether we think that's the case?
Paul: [00:56:33] Well, I think these people have made a beautiful bridge between an enabling technology and a fundamental human mission, which is to be in right, empathetic relationship with our fellow replicating organisms. So I think it's a it's a good indicator. There may be bad indicators, but this is a great one.
Christiana: [00:56:50] Yeah, I would I would go with Paul. I think it's such a beautiful example for what can be done, what can be done when you decide that these very powerful tools are going to be put at a noble service, the service of understanding our role on this planet better. Does it mean that the powerful tools cannot also be put at service that is not quite as noble or in fact, even downright destructive? But this is noble service of a very powerful tool.
Tom : [00:57:25] Hmm. Well, this has been a fascinating episode, and it's just the beginning of, I think, our journey into AI and its role in our emerging future. So thank you, Christiana, for taking the time to bring us into the world of ESP. It's remarkable. And I can see why you have made the exception to your rule and decided to join the board. Well, thank you and look forward to hearing more about how it goes. What an exciting few years it's going to be in this space.
Paul: [00:57:50] Well, congratulations to the Earth's Species Project for securing you, Christiana. My experience bodes well for them as an organization.
Tom : [00:57:57] Absolutely. And thanks everyone for listening. We'll see you all next week. Bye.
Paul: [00:58:00] See you next week by.
