Can AI enable new ways of knowing – and being?

Targeted enhancement of urban soil biodiversity, enhanced by AI, could support human health, in both outdoor and indoor settings. Xin Sun, Institute of Urban Environment, Chinese Academy of Sciences

The following is a chapter I have contributed to a forthcoming book – AI for Sustainable Development Goals – edited by Professor Lou Yongqi, that will be published later this year.

Is it possible that we pay too much attention to the promises being made for new technologies, and too little attention to the social and ecological purpose we would like them to serve?

At a time when progress towards the Sustainable Development Goals (SDGs) seems to have stalled, what would happen if we stopped thinking about those goals as a fixed destination – like the last stop on a railway line – and started treating sustainability as continuously changing system, comprised of social and ecological relationships, of which we are part?

This chapter explores just such an alternative approach to the SDGs. It asks: if sustainability describes a system whose health (or otherwise) is determined by the quality of its social and ecological relationships, in what ways might AI contribute to the health of those relationships?

Otherwise stated: if we were to decide that the purpose of AI is to help all of life thrive – not just human life – what practical consequences should follow from that?

The text explores those question as if we were planning to intervene in five social-ecological systems:

AI gauges vitality of relationships 
if sustainability is determined by the quality of social and ecological relationships, in what ways might AI contribute to the health of those relationships? Jane Memmott develops indicators of biodiversity health that represent diversity, adaptation, and change.


Reliable data are of course important if we are to understand and reverse the destruction of nature that’s so distressing to us all. And it is good news that more and more data about biodiversity is becoming available thanks to the marvels of satellite imagery, DNA analysis, and other data analysed by AI. But is artificial intelligence enough, on its own, to drive the ecological transition we so desperately need?

My starting point in this text: AI can be a support for transformational change. But a truly just transition will only happen when, in the words of s the Spanish theologian Raymond Pannikar, we “see nature differently, relate to nature differently, and understand our purpose here differently”.

If nature connection were just a matter of data, we’d be in a good position. AI and Machine Learning, for example, are being positioned as a global observation platform that monitors ecosystem health at multiple scales – from the planetary, to the microscopic. We’re feeding in data from satellites in space, to microbial communities surveyed by eDNA. As we connect these experiments together, the tools and connectivity are within our grasp to monitor the vital signs of of the planet in real time – place by place, patch by patch.

Scientific discoveries, too, have made remarkable progress since the 1980s. Gaia theory, systems thinking, and resilience science, have all confirmed the wisdom of the ancients: everything is connected. Whether it’s sub-microscopic viruses, mosses, and mycorrhizae – or trees, rivers and climate systems – the evidence is conclusive: all natural phenomena are connected. Indeed, their very essence is to be in relationship with other living beings -including us. No organism is truly autonomous – even if, like man, it thinks it is!

The most exciting prospect – and the least explored so far – are indicators of biodiversity health that represent diversity, adaptation, and change. The study of living systems tells a consistent story. Whether it’s sub-microscopic viruses, mosses, and mycorrhizae – or trees, rivers and climate systems – the health of an ecosystem lies in the vitality of interactions between its component species. Science has confirmed an ancient wisdom: All natural phenomena are not only connected. Their very essence is to be in relationship with other things -including us.

But there’s a dilemma here. A new dashboard is not the same as a new system. The remarkable achievements of earth observation, and the insights of systems thinking, have not transformed our belief systems. On the contrary: we continue to act as if we are separate from nature. The majority of actions taken towards the SDGs, for example, treat the living world as some kind of machine that we need to manage more effectively. Think, as a comparison, about the intensive care unit in a modern hospital. The tubes and gadgets surrounding a sick patient are technical marvels of observation – but they tell us little about the reasons why the patient got sick in the first place.

Dashboards are attractive in an economy that treats raw materials as commodities. In an industrial system uniformity and standardisation – efficiency and control – are success factors. Biodiversity, of the kind found in healthy nature is pretty much the opposite. Diversity and adaptation are the best indicators of vitality. No matter how massive the datasets and simulations created by AI, computational models cannot comprehend the complexity and interdependence of ecosystems. They will remain just that: models of reality. The bank Credit Suisse, with remarkable candour, has put it best: ”biodiversity is the anti-commodity”.

How, then, might we see nature differently, and relate to nature differently? Could AI and machine learning be repurposed to remind us that we are part of nature, not outside it?

If the experience of the last last 50 years is any guide, we’re probably don’t need AI to issue more messages, concepts, and instructions. Rather, what we need are diverse learning environments in which we experience embodied experiences of connection – connection with each other; connection with place; and above all, connection with the living. We need contexts for nature connection in which the health of the soil, microbes, soil, plants – and the health of people – are experienced as a single story.

The word experience, I believe, is key. The challenge for design-plus-art-plus AI: how to become a medium of attention with ecosystems we have neglected; how to be a medium of connection – so we don’t just look; and how to be a medium of relationship with the living world that can persist through time.

A new grassroots movement, AgTech for Agrecology, convened an international grassroots innovation assembly to consider an agreed position on new technologies

One of the goals of natural farming (also called agroecological farming) is to lower inputs of all kinds such as fertilizers and pest controls.

Indesirable inputs in agroecology also include over-complex and non-essential technologies. Food sovereignty – the right of communities to define and govern their food system – encompasses technological sovereignty, too: food-producing communities insist that they should decide what the technology is used in their food systems – or not.

The innovation priority, for these natural farmers, has always been on improvements to existing technologies – and lower tech solutions that can be repaired, reused, shared, or re-purposed.

But lower tech in agroecology does not mean no-tech, and n 2023 a new grassroots movement, AgTech for Agrecology, convened an international grassroots innovation assembly to consider an agreed position on new technologies. The Assembly was uniquely representative because the majority of its officers, as well as members, are small scale farmers themselves. As such, they emphasise the health of whole systems, and seek an equitable balance between ecological, social and economic aspects of farming within the wider food system

Based on that shared understanding about priorities, the Assembly confirmed that many grassroots farmers are already using novel digital tools. These range from smartphone apps, and virtual fencing, to diverse forms of data analysis. Deliberating on on the latest wave of agtech start-ups, and AI, they determined that three application areas were relevant to their movement: Monitoring, Evaluation and Learning (MEL) systems and frameworks; new market pathways for the products of small scale farmers; and platforms to facilitate the peer-to-exchange of knowledge on agroecology .

For Daniel Moss, co-director of the Agroecology Fund, a AI and digital tools can help farmers monitor field conditions in real time, understand soil quality, plan their planting—and connect directly with consumers. “As such, digitalization can strengthen farmers’ understanding of the ecosystems in which they work, their connections with other farmers, their relationships with consumers, even their ability to access native seeds”

How might AI be used, in practice, by small scale farmers? There is particular interest, among some researchers, In so-called Agentic AI. In mainstream healthcare, agentic AIs are being deployed as a personal health assistant that continuously monitor a patient’s data from wearable devices, adjusts treatment plans based on real-time health indicators, and even predicts potential health issues before they become serious.

Could Agentic AI help far for the soil, and their crops?

The answer depends on how such new technologies are used, and by whom. For a new report called Remote Control and Peasant Intelligence – On Automating Decisions, Suppressing Knowledges and Transforming Ways of Knowing, new tools such as AI have positive potential if they accelerate the shift towards decentralised, community-based knowledge-sharing among small scale farmers. They can be useful if they support different types of knowledge, as well as social and ecological relationships.

Strengthening food security is at heart a distributed social and ecological process. It involves people and communities learning by doing, together, in diverse ways appropriate to their culturally unique geographies and temporalities. In the words of Chris Smaje, it involves “a messy politics of relationship, nuance, context, complexity and co-learning”.

AI transforming community management in the online world. It provides sophisticated tools for optimising engagement, personalising interactions on social networks, and more.There is clear potential to repurpose these techniques in real-world communities

One of more damaging myths to emerge during the modern age is the notion that ‘distance is dead’ – and that ‘geography is obsolete’. Any time, anywhere connectivity was a beguiling offer – and we reconfigured the economy accordingly. But a global economy inconstrby ained biophysical realities turned out to be a chimera, and our attention is turning once again to the local. With that focus, a novel innovation question has arisen: can AI be used as a relocalisation engine that compresses space and time? Can AI help us connect overlooked people, assets and places within a bioregional context?

There is good news on this front. Changing the word “faster” to “closer”, it transpires, is not so resource-intensive as ‘killing distance’ was. Many of the ingredients needed for localisation already exist. Ninety five per cent of the economic activity we relied on before globalisation is still there: the small farm, the corner shop, the doctor, the builder, the carer. Indeed, a big majority of the world’s population already meets needs its daily life needs locally. Around the world, hundreds of millions of small and medium-sized companies meet daily life needs within a radius of fifty kilometers of their base.

The relocalisaton movement has up-ended the assumption that the future is all about cities – and a new generation of designers is exploring the consequences. In Design Harvests, for example, Tongji University has already led twelve years of urban-rural design research on Chongming Island near Shanghai. In Design Harvests 2.0, on the learning landscape of Zhangyan, numerous new opportunities for jobs and livelihoods have been explored that combine social, ecological, business, and technical innovation.

The diverse ways in which poor people meet their daily needs are usually described as impoverished, or backward. But during 40-odd years as a guest in what used to be called the. developing world, I came to a startling conclusion: living sustainably is how people survive when they don’t have access to the high entropy support systems of the industrial world. But these local economies can always be enhanced. Finding ways to join the dots together, in ways that add value without adding material and energy intensity, is a key opportunity.

In the mainstream AI space, two application areas have potential to be repurposed for relocalisation: recruitment, and community management.

A common feature in all these positive experiences is the presence of a skilled person, based in the community, who identifies opportunities, connects local actors, and develops projects – all on a continuous basis. In Europe, we call these individuals “village hosts”. Europe needs hundreds, maybe thousands, of village hosts – but, right now, but we only have a few. Their profiles tend to be atypical, so using AI to identify these unusual individuals could be a game-change in rural development. 
https://www.ifad.org/en/w/opinions/4-ways-ifad-is-using-ai-to-transform-rural-development

AI is transforming community management in the online world. It provides sophisticated tools for optimising engagement, personalising interactions on social networks, and more. AI can be integrated into every stage of the social network communication process, making each phase more effective thanks to its intervention.There is clear potential to repurpose these techniques in real-world communities.

A lot of the work people already do locally is unpaid. It doesn’t register as GDP because it involves care. Care is the essential activity people have always undertaken to raise and educate their families, cultivate their land, and support each other in times of difficulty. Mutual aid, in different forms, occurs throughout human history. Whether it’s cooking, cleaning, doing laundry, ironing, consoling, caring for others, listening, value arises from relationships. not from things.

These relationships among individuals, groups, networks, and cultures, make up what Cormac Russell calls associational life. It’s how billions of people with low cash incomes meet daily life needs outside the money economy through traditional networks of reciprocity and gifts. They survive, and often prosper, within social systems based on kinship, sharing, and myriad ways to share resources.

For biologists, the health of an ecosystem lies in the vitality of interactions between its component species. This lesson applies equally to a locality. The process enabling diverse stakeholders to work together is a key success factor. The key processes of design for local are:

Scanning and Mapping Identify opportunities for local provisioning. Search for neglected value – overlooked people, places practices. Discover what is. Explore “what if?” Keep track of where resources come from, identify leakages in the local economy. Explore ways to plug them using local skills and resources

Curating and Convening Local cannot successfully be addressed without the engagement of all the actors concerned. A variety of different stakeholders – formal and informal, big and small – need to to work together. The question – and it is also a design question – is how Designing the process by which groups work together is just as important as deciding what needs to be done, if not more so

The journey back to local is a demanding one in terms of knowledge. Social and ecological contexts are complex. The closer you get to a local situation, the greater diversity of ways to learn about it, to know it. This is a positive: In nature, diversity is healthier than order, control & uniformity. But it does mean that there will be no universal “solution” to fall back on in your social design. Healthy local systems are small scale, and there is no universal best-practice we can apply.

A new kind of infrastructure is needed – social infrastructure – to help us grapple with these new questions. Platform cooperatives, for example, are proving to be effective ways for different actors and stakeholders who need to work together.. Platforms can also provide fair compensation for services provision among the people who make those services valuable.

Numerous regional institutions are also available to ease our transition.A surprising number of these can be repurposed for today’s transition. There are more public libraries in the U.S.A. (120,000) than there are McDonalds. Regional and specialty museums are looking to redefine their roles.Thousands of post offices and local shops already act as place-based meeting points; we can use them, too, as hubs for learning networks.

AI has an important support role to play as the supporting infrastructure needed for these new social relationships to flourish. The re-emergence of gift exchange can be made possible by electronic networks. Mobile devices, and the internet of things, make it easier for local groups to share equipment and space, or manage trust in decentralised ways; technology can help us transition from an economy of transactions, to an economy of relationships. Technology can help reinvent cooperative.

AI empowers indigenous knowledge 
in Decentralising Digital (DEDI). DEDI explores the possible roles that mesh networks, the Internet of Things, voice-enabled Internet, machine learning, and artificial intelligence might play in enhancing ecological agriculture


Around the world, we need to develop rural areas in ways that help their local ecosystems thrive. A key ingredient in this work is fine-grained knowledge of people who work intimately with the land. Can their knowledge and practices be enhanced by AI?

Whilst ‘systems thinking’ is a relative novelty for most of us in the industrialised world, sensitivity to ecological contexts is quite natural to small-scale farmers, shepherds and other pastoralists who have lived on the land for generations. Small-scale herd-owners, for example, practice a form of agriculture production on arduous drylands that dates back some 6,000 years. The lands they work on are home to approximately 80 percent of the world’s biodiversity. Over millenia, these indigenous peoples have had a richer understanding – – and therefore more empathy – with with biosphere – than those who practice ‘production agriculture’.

These indigenous ways of knowing and being in the world tended to be marginalised by Western research and innovation But this changing. Knowledge practices that combine Indigenous knowledge systems, and Western scientific approaches, are now appearing – and the use of AI is easing this transition.

The concept of cognitive justice is based on the recognition of the plurality of knowledge and expresses the right of the different forms of knowledge to co-exist. Indian scholar Shiv Visvanathan coined the term cognitive justice in his 1997 book “A Carnival for Science” in which he documents how different knowledges are connected with different livelihoods and lifestyles and should therefore be treated equally. Since that book, the idea has spread that academic or scientific knowledge is not superior to the knowledge of Indigenous peoples and farmers for whom the protection of life is their guiding principle.

Sometimes described as Everyday Experts, peasants or small-scale traditional farmers are responsible most of the world’s agricultural production.With a world population between 1,300 and 1,600 million, their farming knowledge and practices are the product of over 10,000 years of tradition and experimentation. When it come to whole systems thinking – and being – we have a lot to learn from indigenous and nature dependent communities for whom reciprocity, interdependence and harmony with the rest of nature have been their lived practice for centuries.

Language is a key obstacle to mutual learning among diverse knowledges. The languages of millions of people are marginalized – especially online. As reported by Billy Perrigo in Time magazine, although large language models like ChatGPT—work well in languages like English, where text and audio data is abundant online, they work much less well in languages like Kannada. “Even though Kannada is spoken by millions of people” Perrigo explains, “it is scarce on the internet. Wikipedia has 6 million articles in English, for example, but only 30,000 in Kannada”

But experiments are underway to close this gap.

India, a country where mobile data is among the cheapest in the world, and where even poor rural villagers to have access to both a smartphone and a bank account, is proving to be a fertile testing ground. As explained to me by Charvi Shrimali, their Communications and Storytelling Lead, a start-up called Karya is working with an Indian healthcare NGO to harvest speech data about tuberculosis—a mostly curable and preventable disease that still kills around 200,000 Indians every year. Voice recordings, collected in ten different dialects of Kannada, help train an AI speech model to understand local people’s questions about tuberculosis. Chatbots respond, in the caller’s dialect, with information aimed at reducing the spread of the disease.

But a question arises: do chatbots, even when communicating in a minority language, actually empower the holders of indigenous knowledge? Is the intimate ,embodied, fine-grained knowledge of everyday experts served well by a machine that basically answers questions within a predetermined script?

Another project in India is pushing the boundaries further. Decentralising Digital (DEDI) explored roles that mesh networks, the Internet of Things, voice enabled Internet, machine learning, and artificial intelligence, might play in enhancing daily life practices of small-scale farmers. The work explored ways to enhance farmers’ ability to understand the health of their soil, and their care for biodiversity. New ways to connect and collaborate with with adjacent organisations were a second priority. Download the book:

Indigenous knowledges are not recipes, as if for a cake. They cannot be extracted, printed on method cards, and applied around the word at will. Indigenous knowledge is situated, place-specific, relationa It operates at multiple scales – from microbiome to bioregion .It is also shaped by multiple timescales – from geological time, to the bacterial time of fermentation.

Eco Jurisprudence Monitor, an interactive online platform, compiles ecological jurisprudence initiatives globally as well as related resources for researchers, lawyers, policymakers, and activists

Laws — and the institutions that impose them — are what people mean by the ‘hard-wiring’ that locks us into damaging relationships with living systems. Until recently, only humans have rights in most of the world’s legal systems. Laws were based on the Enlightenment notion that the universe is a repository of dead resources which we can exploit as and when we like — for the exclusive benefit of our own species. In contradiction to ecological principles of wholeness, relationship, and interconnection, legal definitions of property tend to perpetuate the division up of land into discrete parcels. Nature’s inherent diversity is at odds, too, with free trade treaties that support large-scale monoculture projects; these, as we know, destroy biodiversity .

In The Great Work, published in 1999, Thomas Berry called for a new jurisprudence to re-define the relationship between the human community and the Earth community in which it lives. “We need a legal system that governs the relationship between humans and the natural world as a totality, not as a collection of parts and which respects equally the rights of the natural world to exist and thrive” argued Berry.

Changes to the legal status of living systems and property rights are now emerging in a wide variety of legal systems around the world. The remarkable scale and range of these initiatives is recorded by the Eco Jurisprudence Monitor. This interactive online platform, compiles ecological jurisprudence initiatives globally as well as related resources for researchers, lawyers, policymakers, and activists.

Some if these initiatives emerge from unexpected quarters. The More Than Human Rights Project (MOTH) , for example, is based in a major US law school at New York University. MOTH works with a range of scientists – from mycologists to marine biologists to botanists – as well as Indigenous communities and leaders – to disseminate crucial knowledge of the more than human world. It focusses on integrating this knowledge into legal and nonlegal actions and advocacy to protect ecosystems and communities.“We deploy a mycelial mode of thinking” the MOTH website explains. “We connect different parts of the more-than-human rights field, and bolster individual actors’ work, while cohering a larger community of practice and knowledge – much like mycelial networks are often the foundational builders of rich and complex ecosystems.”.

At the scale of the nation state, radical legal expressions of a new world view are beginning to emerge.In 2010, when Bolivia hosted a World People’s Conference on Climate Change and Rights of Mother Earth, it was attended by 30,000 people from 100 countries. One outcome, a Universal Declaration on Rights of Mother Earth, was presented to the UN. And a Global Alliance for the Rights of Nature has been created with an initial 60 member organisations from around the world.

A shift away from seeing Earth solely in terms of ‘resources’ to be exploited for our own use is beginning to appear in international law and governance at a global level, too. A large number of universities are involved in the the Earth System Governance Project, for example, which was launched in 2009. This multidisciplinary network of scholars and practitioners is forging new connections between the social and natural sciences in exploring new models of environmental governance.

For the political ecologist Mihnea Tănăsescu, earth laws can do more than prevent harm to nature. They can also be “infrastructures for reciprocity” . In his book Ecocene Politics, Tănăsescu explains how to cultivate a relational ethics of reciprocity, cooperation, and care for living beings. “We must learn to renovate our legacy forms of political economy and culture, and develop the infrastructures and practices to support mutualism”.
https://david-bollier.simplecast.com/episodes/mihnea-tanasescu-on-the-need-for-ecocene-politics

A transformational new concept of the law along these lines is outlined in The Ecology of Law: Toward a Legal System in Tune with Nature and Community. Its authors, Fritjof Capra and Ugo Mattei, explain that the solution is not just to pass pass new and better laws. “We need reconceptualize law itself, and shift from a paradigm that sees the world as a machine, to a systemic, ecological paradigm that sees the world as a network of interdependencies”.

So where and how can AI support this complex transformation?

The MOTH Project is working with AI ethicists and other experts in technology law and regulation to pursue opportunities and protections for the more-than-human world at the intersection of emerging technologies, data collection, and more than human rights. In a test case project in Ecuador, MOTH experts exploring ways to create a licensing framework in which the fungal datasets collected in an Indigenous territorywould be owned by the forest itself, as a holder of rights, together with the Indigenous community. “Our work is intended to serve as a prototype to guide similar efforts in other areas to develop data ownership and licensing models” say MOTH. “ Respecting the needs and interests of the more-than-human world must be central to the deployment of modern technology”

This book explores a timely question: can Artificial Intelligence be deployed in ways that help us reach the Sustainable Development Goals? As our contribution to the conversation, we focused in this chapter on the purpose of AI development. If sustainability describes a system whose health (or otherwise) is determined by the quality of its social and ecological relationships, we asked, in what ways might AI contribute to the health of those relationships? Among an almost infinite diversity potential ways to intervene in a living system, we selected five zones as examples of what can – and is – being done with AI: nature connection; natural farming; social infrastructure; cognitive justice; and Earth Law. Our hope is that that this attention to social and ecological purpose, in real-world contexts, will add new insights and energy to the AI innovation landscape.

4,500 words