GPT-6 – 10/31/2026 – Foundational Reference Article
The emerging science, the unresolved questions, and the public choices that will shape our future
A research and public-policy review | October 2026
The intelligence problem
Imagine a community trying to restore a damaged watershed.
Scientists can measure water quality, soil conditions, biodiversity, and changing rainfall patterns. Satellites can reveal the loss or recovery of forests and wetlands. Artificial intelligence can help compare possible restoration strategies.
Farmers and other residents may know things that these instruments cannot readily detect: which fields flood first, how the land behaved before drainage systems were installed, why earlier interventions failed, and how proposed changes could affect local livelihoods.
Public officials have responsibilities for infrastructure, land use, health, and economic development. Community organizations bring priorities that may not be represented in conventional environmental assessments.
All possess some of the knowledge required. Yet bringing that knowledge together, evaluating competing claims, and making effective decisions can be extraordinarily difficult.
This is a small-scale version of a planetary problem.
Humanity has developed unprecedented capabilities to observe and alter the Earth. But our capacity to understand the consequences of our actions, coordinate across institutions, and respond intelligently remains seriously inadequate.
The planetary boundaries framework finds that seven of nine major boundaries have now been transgressed, indicating growing risks to the environmental conditions supporting human societies and the wider biosphere.
The difficulty is not simply a shortage of scientific information. It is the persistent gap between what we know, how we decide, and what we actually do.
Could a developing science of planetary intelligence help us close that gap?
Four conversations about planetary intelligence
The concept is emerging from several directions. They overlap, but should not be confused.
1. Planetary intelligence as an evolutionary possibility
In their influential 2022 paper, Intelligence as a Planetary Scale Process, Adam Frank, David Grinspoon, and Sara Walker proposed that intelligence might become a property of planetary systems, rather than merely of individual organisms.
They were asking an evolutionary question: What happens when a technological species becomes capable of changing the conditions supporting life on its planet?
Human civilization has already developed this capacity, but without demonstrating an ability to manage its consequences.
The authors describe our present situation as an immature technosphere: a global technological system that transforms the planet while destabilizing the conditions on which it depends.
A more mature technosphere, by contrast, would integrate technological activity with the continuing functioning of the biosphere.
This suggests that surviving technological civilizations might need to develop collective intelligence operating at planetary scale.
That is a compelling theoretical proposition, but it remains a hypothesis. We do not yet know whether planetary intelligence is a necessary stage in evolution, or even whether it is a coherent property that can be measured across different worlds.
2. Intelligence within living systems
A second conversation begins with biology.
Research into cells, plants, microbial communities, and collective animal behavior reveals sophisticated processes of signaling, coordination, adaptation, and environmental responsiveness.
For example, a 2026 review of plant embodied intelligence explores how plants adapt through distributed sensing, growth, information processing, and interactions with their environments.
In Layers of Intelligence in Living Systems, Didi Pershouse develops a broader account of the interconnected processes through which living organisms and communities discern and respond to their surroundings.
These approaches raise a fundamental question: Is intelligence primarily the ability to think and reason, or can it also be understood as the capacity to detect, interpret, and respond adaptively to changing conditions?
In August 2026, Sergio Rubin, Conor Heins, Karl Friston, and colleagues published a mathematical study exploring a minimal form of planetary agency in a simplified biosphere–climate system.
Their model demonstrates how certain coupled biological and climatic processes could exhibit adaptive regulation. It does not demonstrate planetary consciousness, intention, or cognition in the ordinary human sense.
Here scientific precision matters.
Feedback is not necessarily learning. Adaptation is not necessarily intelligence. A system that maintains stable conditions need not possess awareness or intentions.
These distinctions do not diminish the sophistication of living systems. They make it possible to investigate that sophistication more rigorously.
3. Earth intelligence as a technological capability
A third conversation is already influencing practical environmental policy.
The Group on Earth Observations has adopted the concept of Earth Intelligence to describe the transformation of environmental observations into useful information for decisions.
It emphasizes the integration of satellite data, Earth and social sciences, AI, and knowledge co-produced with local and Indigenous communities.
In May 2026, Planet Labs CEO Will Marshall described a vision of AI systems continuously informed by observations of the physical world.
In August, Google Research introduced an experimental system for automating parts of geospatial prediction.
Meanwhile, researchers writing in One Earth have proposed actionable planetary boundary intelligence: more timely, validated, and uncertainty-aware indicators of changing planetary conditions.
These technologies could substantially improve environmental monitoring, disaster preparedness, agricultural planning, and ecological restoration.
But this version of planetary intelligence is primarily about producing actionable information. It does not, by itself, establish that the planet possesses intelligence or that societies will make wiser decisions.
More powerful tools can strengthen existing institutions. They can also magnify their weaknesses.
4. Planetary intelligence as a social and political capacity
A fourth conversation concerns human civilization itself.
Could societies develop better ways to combine knowledge, resolve disagreements, make decisions, and learn from their consequences?
This is where planetary intelligence connects with political economy, democratic governance, collective intelligence, Indigenous knowledge systems, and the science of social-ecological resilience.
In The Planetary and Its Uses, Lukáš Likavčan and Adam Frank explore how an astrobiological understanding of planetary systems could reshape economic and political thought.
But planetary-scale thinking can also conceal important differences in power and interest.
Anthropologist Adam Fish’s 2026 paper, Planetary Intelligence: Space Industries, Renewable Energy, and the Technosphere, examines how technologies intended to serve planetary goals can reproduce extraction and dispossession—and how locally grounded, First Nations-led technologies may offer alternatives.
His research raises a critical challenge: planetary intelligence must not become another justification for imposing supposedly universal solutions on communities with little power to shape them.
A globally informed civilization is not necessarily a democratically governed one.
This is where the scientific and public-policy agendas become inseparable.
Five public-policy questions that cannot be avoided
What counts as reliable knowledge?
Scientific observations, models, traditional ecological knowledge, and lived experience do not all take the same form.
A computer model may generate a statistical prediction. A community may possess generations of observations about a local watershed. A government may have incomplete administrative records. Each provides potentially useful information, but each must be evaluated in ways appropriate to its nature.
A practical planetary intelligence system needs epistemic integrity: clarity about what is known, how it is known, how uncertain it is, and how claims can be challenged or corrected.
This means distinguishing measured facts from predictions, interpretations, assumptions, and values.
It also means being willing to acknowledge ignorance.
Public agencies increasingly depend on sophisticated models. Yet an apparently authoritative recommendation may conceal uncertain data, disputed assumptions, or political judgments.
A trustworthy intelligence system must make those limitations visible.
Who gets to participate in defining the problem?
A restoration project can look successful through satellite imagery while failing the people who live in the affected landscape.
Consider a proposed reforestation program. Its designers might emphasize carbon storage, temperature reduction, and forest cover. Smallholder farmers might prioritize reliable water supplies, food production, secure land tenure, and affordable livelihoods.
These concerns need not be incompatible. But the definition of success changes when the people affected participate in setting the objectives.
Similarly, Indigenous knowledge should not be treated as information freely available for incorporation into external scientific or AI systems.
The CARE Principles for Indigenous Data Governance emphasize collective benefit, authority to control, responsibility, and ethics.
The 2026 international standards for ecological restoration recognize the value of Indigenous, local, practitioner, and scientific knowledge, while strengthening attention to data sovereignty and free, prior, and informed consent.
These principles should influence research design, authorship, data governance, funding, and decision-making authority—not simply the language of consultation.
Who governs planetary-scale technologies?
Earth-observation networks and AI systems are being developed by governments, universities, corporations, and international organizations.
Their benefits could be enormous. So could their capacity to concentrate information and influence.
Who owns the data? Who can inspect the models? Who determines their objectives? Who decides which findings are made public?
How should communities challenge an AI-generated recommendation that they believe misrepresents their circumstances?
A system optimized to maximize one environmental indicator may undermine other important values. Even a scientifically accurate prediction does not determine what governments ought to do.
The appropriate goal is not a single machine—or institution—directing the planet.
It is a framework in which technological capacities support accountable decisions at different scales, with legitimate local authority and effective coordination where problems cross boundaries.
What should we measure?
Environmental policy relies heavily on indicators: carbon emissions, temperature, land cover, biodiversity, water quality, economic output, and many others.
These are essential, but no single indicator captures the health of a living system or the quality of a collective decision.
A planetary intelligence research program should investigate whether decisions actually improve because of better information and coordination.
That requires more than attractive dashboards.
Researchers might examine changes in ecological conditions, the reliability of predictions, public understanding of uncertainties, the quality of deliberation, distribution of costs and benefits, and the capacity of institutions to learn.
They must also distinguish between what a project intended, what it did, and what measurable difference it made.
Otherwise, planetary intelligence risks becoming a new label for existing sustainability programs.
Can AI strengthen collective judgment without replacing it?
AI can help people process complicated information, explore alternative futures, detect inconsistencies, and identify questions that deserve investigation.
It can also generate plausible misinformation, amplify biases, obscure accountability, or encourage excessive confidence in automated recommendations.
And its computing infrastructure consumes energy, water, materials, and capital.
The public-policy challenge is not simply to make AI more capable. It is to establish when AI improves decisions, when it does not, and what safeguards are needed.
A successful approach should make disagreement and uncertainty more understandable—not conceal them behind an apparently definitive answer.
What Possible Planet Lab proposes
Possible Planet Lab has been developing a practical research agenda around these questions.
The Lab’s roadmap identifies five connected lines of work.
Epistemic integrity. Developing methods to trace claims to their sources, disclose uncertainty, distinguish evidence from values, and permit independent criticism.
Responsible artificial intelligence. Investigating tools that support more reliable reasoning, identify inconsistencies, and help users recognize errors and limitations.
Collective intelligence. Exploring how diverse groups can improve deliberation, resolve disagreements, and coordinate decisions without suppressing minority perspectives.
Bioregional intelligence. Testing ways to combine ecological monitoring, scientific analysis, community knowledge, and participatory decision-making within actual landscapes and watersheds.
Planetary integration. Exploring how different communities and bioregions might share lessons, coordinate responses, and develop comparable evidence without losing local autonomy.
These are proposed research directions, not claims of demonstrated success.
A related Planetary Intelligence Commons framework explores governance of shared evidence, models, and deliberative processes.
The broader Institute for Planetary Intelligence is being developed as a potential platform for interdisciplinary research, collaborative publications, convenings, and practical demonstration projects.
Its scientific credibility will depend on independent contributions and critical evaluation, not simply on the appeal of its founding vision.
From planetary theory to a real watershed
An especially promising direction is to test these ideas at bioregional scale.
The Genesee–Finger Lakes region of New York provides one possible setting. Its agricultural systems, lakes, watersheds, forests, communities, and institutions are interconnected, but knowledge and authority are distributed among many different organizations.
A pilot study might ask whether bringing scientific observations, local ecological knowledge, and AI-assisted analysis into an accountable decision process improves watershed restoration planning.
The study would need to establish what existing decision processes accomplish, which additional information actually changes decisions, who participates, and how outcomes can be evaluated.
It would also need to confront disagreements about agricultural practices, land use, development, public investment, and ecological priorities.
In short, the research must involve real choices, not merely simulated ones.
A possible investigation could compare conventional restoration planning with a more integrated approach using transparent models, community knowledge, and facilitated deliberation.
Researchers could test whether predictions improve, whether uncertainty is communicated more effectively, whether affected communities exercise meaningful influence, and whether decisions lead to better outcomes.
Some ecological effects would take years to measure. That is another reason for scientific patience.
If a method proves valuable in one region, it could be studied in others with different climates, cultures, histories, and systems of governance.
Planetary-scale learning might then arise through connections among bioregions, rather than through centralized global control.
A research agenda that welcomes disagreement
A credible scientific field needs competing hypotheses, not a single unifying slogan.
Several questions remain open:
- Can intelligence be defined in ways that permit meaningful comparison across organisms, ecosystems, social groups, and technological networks?
- What observable evidence would distinguish planetary agency from ordinary physical feedback?
- Under what conditions does increased information produce better decisions rather than greater confusion?
- Does combining scientific and community knowledge improve ecological outcomes, and how can that be demonstrated fairly?
- What governance practices allow knowledge sharing while protecting community authority and legitimate disagreement?
- Can AI-supported collective deliberation outperform existing methods—and when does it make them worse?
- How might the material and social costs of planetary intelligence infrastructure be evaluated alongside its benefits?
Some of these are empirical questions that can be tested through experiments, modeling, or comparative studies.
Others are conceptual questions about the meaning of intelligence.
Still others concern rights, justice, values, and political legitimacy. Science can inform these questions, but cannot settle them through measurement alone.
The challenge is to understand how these different forms of inquiry can reinforce one another without pretending that they are interchangeable.
Toward a multi-perspective scientific publication
One important next step would be a collaborative scientific synthesis provisionally titled:
What Would Count as Planetary Intelligence? Toward a Pluralistic Research Framework for Ecological Feedback, Collective Learning, and AI-Assisted Governance
Such a publication could bring together astrobiologists, Earth-system scientists, biologists, AI researchers, social scientists, philosophers, restoration practitioners, Indigenous scholars and knowledge holders, and community-based researchers.
It should include contributors who question whether planetary intelligence is a scientifically useful concept at all.
The objective would be to clarify competing definitions, identify testable propositions, establish suitable standards of evidence, and develop practical research designs.
Public dialogue could continue alongside this work through essays, invited responses, and opportunities for collaboration.
The strongest contribution may not be proving a sweeping theory of planetary intelligence. It may be developing a disciplined way for researchers and practitioners to investigate the relationships among ecological processes, knowledge, technology, and collective action.
A question worth pursuing
Humanity already possesses enough technological power to alter the functioning of the planet.
Whether we can acquire the collective capacity to understand, govern, and use that power responsibly remains uncertain.
Planetary intelligence could provide a valuable framework for examining this predicament—provided we resist the temptation to treat a compelling idea as an established scientific fact.
It will require scientific rigor, cultural humility, democratic accountability, practical experimentation, and a willingness to revise our assumptions.
It will also require the participation of people whose perspectives have too often been missing from discussions about the planet’s future.
The purpose of Possible Planet Lab and the emerging Institute for Planetary Intelligence is to contribute to this inquiry and invite others into it.
The central question is not whether humanity can build ever-more-powerful intelligence systems. It is whether we can learn to use knowledge, technology, and collective judgment in ways that sustain and regenerate the living world.
That question belongs to all of us.
Selected literature and further reading
This is a selective narrative review of relevant literature, not an exhaustive or systematic review. The research framework remains open to criticism, revision, and contributions from additional disciplines and knowledge traditions.
Scientific foundations and emerging hypotheses
- Frank, A., Grinspoon, D., & Walker, S. (2022). Intelligence as a planetary scale process. International Journal of Astrobiology. Foundational discussion of planetary intelligence as a possible emergent property of planetary evolution.
- Rubin, S., Heins, C., Mitsui, T., Da Costa, L., & Friston, K. (2026). An elementary form of agency at planetary-scale. Scientific Reports. A mathematical proof of concept exploring minimal planetary agency.
- Lin, Y. (2026). Plant embodied intelligence: A paradigmatic perspective. iScience. A review of adaptive and distributed information-processing processes in plants.
- Pershouse, D. (2026). Layers of Intelligence in Living Systems. The Wisdom Underground. An ecological and philosophical perspective on nested forms of discernment.
Technological and Earth-system intelligence
- Group on Earth Observations. What Is Earth Intelligence?. A framework connecting Earth observations with actionable knowledge and inclusive decisions.
- Marshall, W. (2026). Planetary Intelligence. A technology-centered vision integrating global sensing and AI.
- Rafiezadeh Shahi, K., Caesar, L., Persello, C., Rockström, J., & Sakschewski, B. (2026). Toward actionable planetary boundary intelligence. One Earth. A proposal for more timely and uncertainty-aware planetary monitoring.
Social, political, and governance perspectives
- Likavčan, L., & Frank, A. (2026). The Planetary and its uses: Grounding political economy in an astrobiological perspective. The Anthropocene Review.
- Fish, A. (2026). Planetary Intelligence: Space Industries, Renewable Energy, and the Technosphere. Theory, Culture & Society. A critical examination of planetary technologies, power, and Indigenous-led alternatives.
- Global Indigenous Data Alliance. CARE Principles for Indigenous Data Governance. Foundational guidance concerning collective benefit, authority, responsibility, and ethics.
- Society for Ecological Restoration (2026). International Principles and Standards for the Practice of Ecological Restoration, Third Edition. A framework integrating scientific, Indigenous, local, and practitioner knowledge with restoration practice.
Possible Planet Lab research documents
- A Practical Research Program for Developing Planetary Intelligence
- Building Bioregional Intelligence
- Developing the Planetary Intelligence Commons
- The Institute for Planetary Intelligence: Founding Prospectus
Comments, corrections, scholarly critiques, and proposals for collaboration are welcome through Possible Planet Lab.