AI, decoding animal communications, and rights for nature
Continuing our series of interviews with leaders in their respective fields of the climate and nature transition, this month we sat down with Jane Lawton, a pioneer in using AI to decode animal communication.
Earth Species Project CEO Katie Zacarian swimming with Hope, a juvenile Sperm Whale (Physeter macrocephalus) near the island of Dominica in the waters of the Eastern Caribbean Sea. Photo by Keri Wilk.
Each month, we are interviewing a range of influential and inspirational leaders from across the climate and nature transition. These interviews are intended as a window into innovative and exciting ways of approaching the transition and to spotlight the people who are at the forefront of these changes. This month, we sat down with Jane Lawton, Director of Impact at Earth Species Project. Jane has devoted her career to solving the complex puzzle of how we live sustainably on the earth. Her diverse experience has ranged from designing conservation programs in Africa to setting global standards for conservation in Asia. Earth Species Project is a non-profit using AI to decode animal communication, with the ultimate goal of changing the way human beings relate to the rest of nature.
What is one object that you currently have on your desk?
I don’t have much on my desk - my job keeps me on the move a lot, so I keep it light. But one thing that is always with me is my trusty WHSmith recycled lined notebook. This is with me every day and filled with to-do lists and random ideas, it goes with me on every trip. In a world that feels increasingly ephemeral, where our thoughts, ideas and basic task management are moderated by computers and by AI, this notebook keeps me grounded in the here and now. I love the physicality of pen and paper and how it is a tool that can be used anywhere, anytime, with or without internet or power. It’s a tool that requires me to still do the thinking.
How would explain what you work on to a five year old?
Have you ever wondered what the birds in your garden are saying to each other? Or what the fish you see in a pond or river are chatting about underwater? We’re using really smart computers to help us listen to animals and figure out what they’re saying.
Some animals make sounds we can’t hear – like bats that squeak in super high-pitched voices, or elephants that rumble in deep sounds too low for us to notice. Computers can help us “hear” these sounds. They can also help us translate their meanings. Maybe, one day, they might even allow us to have a conversation with another animal.
We're doing this because we think that if we can understand other animals better, we will be more motivated to protect them and to protect the planet. But we’re also really excited to think about how much we might learn from them given that some animals have been around for much longer than human beings and they are probably pretty wise! Maybe they can even help us to solve climate change.
Can you describe a recent moment or experience on a project that has particularly stuck in your mind?
The things that have stuck in my mind most over the past few months, have been the incredible conversations we’ve been having with people who are pioneering new ideas and approaches that challenge humans to think differently about our relationship with other species.
As an example, the work of Tehanu, which is giving endangered species digital identities, using AI to discern their needs and preferences, and then vesting them with pots of money that they can use to pay for services from human agents. I love the way this reverses the power dynamic and has the potential to fundamentally change how local communities think about their animal neighbours.
Another great example is the work of Animals in the Room, which is exploring new methods of understanding the perspectives of other species, and using these tools to help in human-wildlife conflict situations. The visionary new thinking is laid out in Dr. Karen Bradshaw’s new book Wildlife as Property Owners.
There are so many incredible ideas out there that all offer glimpses into a possible new future. We’re really interested in supporting and amplifying these ideas, and then exploring how new AI tools might help to take them further.
Earth Species Project senior AI researchers Jen-Yu Liu, and Maddie Cusimano with one of their research partners Daniela Canestrari at Universidad de Leon. Daniela and her team are studying a unique population of carrion crows. ESP is building models to help detect and classify hundreds of thousands of hours of recordings, data from biologgers, and video from nest cams.
What are some of the hardest challenges you’re grappling with right now?
Every species experiences the world in a completely different way. Bees see ultraviolet light; whales use echolocation to navigate the ocean. We can’t know what it’s like to experience the world as they do. So, when we try to “translate” animal communication, we have to make sure we’re capturing what they mean, rather than what we assume.
Another challenge is that of translation vs. interpretation. Are we really aiming to translate animal communication word-for-word like Google Translate? Maybe, but I think it's unlikely that we’ll ever achieve that sort of 1:1 translation.
What’s more likely is an ability to interpret meaning. Communication is all about context. Think about how much human communication goes beyond the words we say – it’s our body language and facial expressions, our tone and pitch, and even our behaviour. The difference between translation and meaning has huge implications for how we design and build AI models.
Finally, there are millions of species on Earth. While the foundation models we’re building are species agnostic and will support analysis of data from many kinds of animal, we also believe that using these tools to do deep dives into a subset of species will be important to generate new insights. But how do we decide which species to start with? We must balance scientific potential, ethical considerations, and real-world impact and that’s a very live conversation for us right now.
What are the most exciting developments you are seeing in your space?
AI is moving very fast. It’s a race to the best performing models. They are going to get faster and cheaper, and hopefully more environmentally friendly too. Every advancement gives us new tools to apply to animal communication.
Simultaneously, we have an interesting window of opportunity right now as the rights for nature movements begin to gain traction. Around the world, people are starting to recognize that providing a set of basic rights to nature – from rivers to forests to other species – may help to ensure its protection. But how do you grant rights without truly understanding the needs, desires and preferences of other species? This is where AI could help fuel and deepen the impact of those movements.
I’m also really excited to see that some of our funders are starting to embrace radical new ideas about how we relate to the rest of nature and there’s growing support for projects like ours that are bridging AI, ecology, and systems-level change. Philanthropic capital should be the most risk-tolerant, but most funders have been taking conservative approaches to date. So, it’s inspiring and hopeful to see that a handful of them now recognize that we are going to have to invest in thinking that will truly change the conversation.
You’re organising a gathering to discuss new approaches to the transition to a sustainable future. What does it look like, what would it focus on, and who would be around the table?
This gathering would take place in deep nature. No stale conference rooms with artificial lighting. In my mind, the idea of a transition to a sustainable future (or I would prefer to say a regenerative future) requires a fundamental rethinking of the relationship that human beings have to the rest of nature. It feels impossible to do that when we are in spaces that we have deliberately constructed to keep us separate from other species, and which actively discourage recognition of our dependence on them.
The questions discussed would also require moving away from how we, as human beings, can engineer a world that will sustain us alongside other species. Instead, the focus would be on what true partnership or reciprocity with nature looks like.
And ideally it would include questions or ideas contributed directly by other species themselves. Of course, our human limitations mean that is not possible today, so if this gathering were to take place tomorrow, we would need to find proxies - in the same way that some companies are now embracing the idea of nature on their boards through human representatives or simulated chats provided by AI.
But if it’s taking place in five to 10 years - who knows? I would like to think we would have the opportunity to hear directly from other species - perhaps in an exchange moderated by AI - and the potential this holds to shift the narrative, to change the way we engage - is tremendously exciting.
Find out more about the work of Earth Species Project here.
