The ecological crisis looks like a problem of technology and policy, but at root it is a philosophical one: where do we draw the boundary between "human" and "nature"? Four thinkers redraw that line—Leopold expands the moral community to the land; Næss expands the small self into an ecological Self; Laozi demotes human mastery to "assisting the self-becoming of all things"; Buddhism uses dependent origination to dissolve the isolated "I". The first two enlarge the circle, the last two dissolve the boundary—but both point to one place: the self set apart from all things is itself the origin of the destruction.
Aldo Leopold
Western · Land Ethic
A Sand County Almanac (1949); American ecologist, father of modern conservation; 1887–1948
Core Thesis · Original Text
A thing is right when it tends to preserve the integrity, stability, and beauty of the biotic community. It is wrong when it tends otherwise.
— A Sand County Almanac, "The Land Ethic"
Thesis: Ethics should extend from the interpersonal to "the land"—the biotic community of soils, waters, plants and animals. The human is not a conqueror of the land but a plain member of it.
Background & Key Insight
Leopold began as a "resources for human use" manager who once urged shooting wolves to boost the deer harvest. Only after he shot a wolf and watched the "fierce green fire" die in its eyes—and saw that without predators the deer exploded and stripped the mountains bare—did he learn to "think like a mountain." Against the utilitarian conservation of his day (nature = a resource to develop), his key insight: the history of ethics is the history of the "community" boundary widening; the next step is to admit the land.
Cross-disciplinary cross-reference
"Biotic community," "integrity, stability" are not poetic flourishes but built directly on 1940s ecology—food chains, energy flow, the biotic pyramid. Making "stability" an ethical criterion anticipates ecology's later "diversity–stability" hypothesis and complex-network resilience: the more redundant the species and links, the better the system withstands shocks.
Contemporary Relevance
"Thinking like a mountain" is a discipline against single-objective optimization: don't fixate on a local short-term metric (kill wolves = more deer); watch the system's second-order effects. This is especially dangerous when AI optimizes a single reward function—reward hacking is the algorithmic version of "killing the wolves": the metric climbs, the system collapses.
In one line: the history of ethical progress is the history of the circle of "us" steadily widening.
If AI becomes some kind of "new member," should it be admitted to the moral community, or remain forever a tool?
Arne Næss
Western · Deep Ecology
"The Shallow and the Deep, Long-Range Ecology Movements" (1973); Norwegian philosopher, founder of deep ecology; 1912–2009
Core Thesis · Original Text
The flourishing of human and nonhuman life on Earth has value in itself. This value is independent of the usefulness of the nonhuman world for human purposes.
— Deep Ecology Platform, principle 1
Thesis: Distinguish "shallow" ecology (managing pollution for human benefit) from "deep" ecology (granting all life intrinsic value); and through "Self-realization" expand the isolated small self into an "ecological Self" that identifies with all of life.
Background & Key Insight
Næss was a mountaineer and a Spinoza scholar. In 1973 he charged that mainstream environmentalism was merely "shallow" reform—cleaning up pollution so humans live more comfortably, still anthropocentric. His insight: as the boundary of "self"-identification widens, protecting a forest becomes not altruistic sacrifice but self-care—you defend the forest as you would defend your own body. He drew explicitly on Spinoza's "God, or Nature" and on Buddhism, against anthropocentrism.
Cross-disciplinary cross-reference
The "ecological Self" borrows openly from Buddhist no-self: as the boundary of "I" widens, the self/other opposition dissolves—directly isomorphic with "dependent origination" below. It also echoes cognitive science's "extended mind": the self does not stop at the skin.
Contemporary Relevance
For the "AI super-individual": real sustainability rests not on moral preaching (demanding sacrifice) but on widening identification. When you recognize your team, users, and ecosystem as the "extended self," helping others is helping yourself, and motivation stops fighting itself. Same for parenting: cultivate a child's sense of identification with life first, then talk rules of "being green."
In one line: when the "self" is large enough, protecting nature is no longer sacrifice but self-preservation.
Could extending identification to all things dilute one's responsibility to the concrete person right in front of you?
Daodejing, ch. 25 & ch. 64; founder of Daoism; ca. 6th century BCE
Core Thesis · Original Text
"Humanity follows the Earth, Earth follows Heaven, Heaven follows the Dao, and the Dao follows what-is-so-of-itself (ziran)." — Daodejing, ch. 25
"Assist the self-becoming of the ten thousand things, and dare not act [coercively]." — ch. 64
Thesis: The highest "Dao" takes ziran (self-so-ness) as its model; humans should "assist" the self-unfolding of all things, not impose their will upon it.
Background & Key Insight
Laozi wrote amid the chaos of the Spring and Autumn era, against the Confucian "active doing" (rebuilding order through rites and institutions). Two terms must be clarified: ziran means not "Nature" as an object but "self-so, spontaneous"; wuwei means not doing nothing but not over-acting, not forcing against the grain. Confucians would save the world by adding human order; Laozi by subtracting human contrivance. Key insight: forced intervention often fails through the system's backlash; the best governance "assists" rather than "acts."
Cross-disciplinary cross-reference
"Assisting the self-so of all things" is strikingly isomorphic with self-organization in complex systems: ecosystems, markets, ant colonies generate order spontaneously without central command, and rigid central control often destroys their resilience. This is the modern scientific footnote to Laozi's "doing nothing, yet nothing left undone." Ecological "rewilding"—removing human management so the system heals itself—is almost the practical version of "assisting its self-so."
Contemporary Relevance
For technologists: when designing systems (teams, products, AI agents), use less rigid control and leave more room for self-organization—set the boundary rules, don't micromanage every step. The most skillful management is often the "assisting" that goes with the grain, not the "forcing" that goes against it.
In one line: the most skillful action assists with the grain rather than forces against it.
As AI lets us "force" ever more (precise intervention in nature and society), can Laozi's "assist the self-so" still hold?
Dependent Origination
Eastern · Buddhism / Huayan
Saṃyukta Āgama on dependent origination; Avataṃsaka Sūtra's Net of Indra; pratītyasamutpāda
Core Thesis · Original Text
"This being, that is; this arising, that arises. This not being, that is not; this ceasing, that ceases." — Saṃyukta Āgama, fascicle 12
Thesis: Nothing has independent own-nature; all things arise interdependently through conditions. Human and nature are not two; to wreck the environment is to wreck the very web of conditions on which one's own existence rests.
Background & Key Insight
Dependent origination is the Buddha's core teaching: nothing has svabhāva (independent, fixed essence); all co-arise. The Huayan school pushed this to its limit in the "Net of Indra": a cosmic net with a jewel at each knot, every jewel mirroring all the others without end—one is all, all is one. Modern Engaged Buddhism (Thich Nhat Hanh) calls this "interbeing": in a single sheet of paper there is the cloud, the sunshine, the logger. Its target is exactly the illusion of an isolated self that drives grasping and exploitation.
Cross-disciplinary cross-reference
The Net of Indra is strikingly isomorphic with ecology's "food web / material cycles": a disturbance at any node propagates through the network to move the whole (trophic cascades, the butterfly effect). This is no mere analogy—ecology just is a science of "this being, that is." It also connects directly to the "ecological Self" above: no own-nature means no isolable "I" left to be guarded.
Contemporary Relevance
For systems thinkers, "interbeing" is an anti-reductionist tool: to analyze any node (a user, a line of code, a species) is to read its web of conditions, not to optimize it in isolation. Parenting: teaching a child that "a whole forest lives in one sheet of paper" plants an ecological intuition far deeper than preaching "save paper."
In one line: there is no isolated "I" to protect, because the "I" was always the whole net.
If all things inter-are, is individual responsibility diluted—or magnified to infinity?
All four redraw the same boundary. Leopold and Næss enlarge the circle of "us"; Laozi and Buddhism dissolve the boundary of "I"—but both return the human from master of nature to member within the net. For the "AI super-individual," this is a hard discipline: before optimizing any node, ask what kind of net it moves. True wisdom lies not in controlling the part but in guarding the integrity, stability, and resilience of the whole.
Going Deeper
Leopold admits "the land" into the moral community; Næss and Buddhism dissolve the boundary of "self." Which is more radical: "enlarging the circle" or "abolishing the circle"?
"Abolishing" is more radical but harder to operationalize. Enlarging the circle (Leopold) keeps the distinction between "I" and "the admitted," easing legislation and trade-offs; dissolving the boundary (Buddhist no-self, Næss's larger Self) uproots the anthropocentric pivot, yet its "non-discrimination" struggles to guide "which to save first." The pragmatic path: let "interbeing" change the cognitive ground tone, then use "enlarging the community" to build workable institutions.
Laozi's "assist the self-so" and Næss's "Self-realization" both oppose human domination of nature. How do they differ in practice?
Opposite routes, complementary. Laozi works by subtraction—removing surplus human intervention so the system self-organizes (rewilding, leaving space); Næss by addition—widening the range of identification so defending nature becomes defending oneself. One is "do less," the other "identify wider." The first governs the system, the second reforms motivation; together they leave the system room and give the human an inner reason.
If all things "inter-are" with no own-nature, yet ecological action must set priorities (which species first?), how can a non-discriminating ontology guide discriminating action?
Buddhism never denies distinctions at the "conventional truth" level. The "ultimate truth" of no own-nature is meant to dissolve grasping and arrogance; the conventional level still chooses by conditions and weight—protecting a keystone species precisely because it moves more of the web. Non-discrimination dissolves the conceit of "human above all," not all judgment. It changes motive and field of view; priorities still follow the network's actual dependencies.
Counting AI and data centers as part of the "web of conditions"—reasonable extension, or abuse?
It depends whether you extend "relation" or "value." Saying data centers consume water and power and are embedded in material cycles, hence belong to the web, is reasonable and necessary (their carbon footprint truly moves the ecosystem). But saying an algorithm has "intrinsic value" the way a wolf or forest does demands caution: smuggling the descriptive "interdependence" into a normative "moral status" oversteps. Extend relation freely; conferring value needs a separate argument.
Western environmental philosophy enters via "value/rights," the East via "ontology/identification." Which should AI governance learn from?
Both are indispensable. The Western "rights/value" frame is computable and can be written into law and alignment targets—the skeleton of governance—but it tends to stall at the surface of "what rights to grant." The Eastern "interbeing / assist the self-so" supplies the cognitive ground tone: it forces you to ask what whole net each optimization moves, and counsels against rigid control, leaving room for self-organization. The skeleton sets boundaries; the ground tone sets direction—use Eastern systems intuition to calibrate Western rule-engineering.