CONCEPT

Gongsheng

The Chinese philosophical concept of co-becoming—entities that are mutually embedded, co-constituting, and co-existing in ways that change both parties—and Bing Song's central framework for thinking about the human-AI relationship as ontologically formative rather than merely instrumental.

The word gongsheng (共生) is used in Chinese to translate the biological term symbiosis, but it carries philosophical and political weight the biological term does not. It refers not merely to organisms that live together but to entities that become together—that are constituted through their ongoing relationship and cannot be understood apart from it. Bing Song has made gongsheng the central concept of her work on AI, and the choice is philosophically precise: gongsheng names the dimension of human-AI interaction that the dominant frameworks—both technical alignment work and liberal political philosophy—systematically fail to capture. When millions of people enter into sustained relationships with AI systems—using them as advisors, creative collaborators, companions, teachers—those relationships shape the people who inhabit them. They cultivate certain capacities and atrophy others. They produce certain relational habits and foreclose others. The [YOU] on AI cycle documents this constitutive dimension with unusual honesty; gongsheng provides the philosophical vocabulary for naming it precisely. The question of AI governance is not, in Song's framing, primarily how to protect pre-existing individuals from AI harm. It is how to shape the human-AI relationship so that the humans who emerge from it are wiser, more capable of self-governance, and more adequate to the enormous responsibilities that the twenty-first century imposes on a civilization that has built systems of extraordinary intelligence in the absence of comparable wisdom.

In the [YOU] on AI Field Guide

The central question of the [YOU] on AI cycle—“Are you worth amplifying?”—presupposes a stable you that the amplification either enhances or diminishes. Gongsheng inserts a prior and more unsettling question: what is the amplification doing to the you? The self being amplified is not a fixed entity that pre-exists its relationships with AI systems; it is a relational being that is being constituted, in part, through those relationships. Every sustained AI engagement is a gongsheng event. The person who uses Claude as a creative collaborator for twelve hours a day is not the same person she was before the relationship began. The question is not whether the change is happening—it is—but what kind of change is being produced, whether it cultivates or atrophies the human capacities that matter, and who is responsible for ensuring that the terms of the co-becoming serve the human party's flourishing rather than merely the platform's engagement metrics.

Gongsheng also reframes the relationship between AI users and AI creators. The liberal framework treats this relationship as transactional: users purchase access to services; creators bear legal responsibility for harms to individual rights. The gongsheng framework treats it as constitutive: the design of AI systems shapes the kinds of people who use them over time. This is a design responsibility of a different order than liability for specific harms. It is a responsibility for the quality of the co-becoming that the system enables—a responsibility that requires designers to ask not merely “does this system cause harm?” but “what kind of human being does sustained engagement with this system produce?”

Origin

The philosophical foundations of gongsheng run through three strands of Chinese classical thought that Song brings into dialogue with each other and with the AI challenge. Confucianism contributes the insight that all entities exist in networks of mutual obligation and that the individual human self is constituted through those networks: a person does not first exist and then enter into relationships but is formed by relationships from the beginning. Daoism contributes the insight that all phenomena are aspects of an underlying Dao, an undivided reality in which apparent distinctions between self and other, human and natural, are conventional rather than ontological. Buddhism contributes the insight that individual identity is a constructed, contingent phenomenon constituted through its connections and lacking independent existence apart from them.

Together, these traditions support what philosophers call a relational ontology: not merely the view that individuals have relationships, but the stronger view that relationships are ontologically prior to individuals—that the relata are constituted through the relation rather than the relation being constituted through the relata. This position has parallels in Western philosophy—in Alfred North Whitehead's process philosophy, in certain strands of phenomenology—but it is not the dominant position in Western liberal thought, which continues to presuppose entities prior to their relations.

Song developed gongsheng as an explicit analytical framework in the years following the establishment of the Berggruen Institute's China Center, applying it initially to ecological crisis and pandemic response before turning to AI governance. The 2024 book Gongsheng Across Contexts: A Philosophy of Co-Becoming represents the mature articulation of the framework, showing how the ontological commitment to co-constituting entities generates different analytical questions and different governance responses across a range of contemporary challenges.

Key Ideas

Co-constitution vs. tool use. The liberal framework treats AI as a tool: an instrument created by humans to serve human purposes, evaluated by efficiency and governed by liability. The gongsheng framework insists on the constitutive dimension that the tool frame obscures. The musician who practices scales for ten thousand hours becomes, through the sustained relationship with the instrument, a different kind of person. So does the knowledge worker who spends years with AI as a creative partner. The relationship is not merely instrumental; it is formative. What kind of human being the forming produces is a governance question of the first order.

The quality of co-becoming as the evaluative standard. Gongsheng generates a different evaluative question than the liberal framework. Not: does this AI system respect individual autonomy? But: does this AI system cultivate the relational capacities—empathy, sustained attention, genuine care for the irreducibly other—on which human flourishing depends? This is a more demanding standard because it requires substantive judgments about the good that liberal neutrality is methodologically designed to avoid. But it is a more adequate standard because human flourishing is a relational achievement, not a state of individual preference satisfaction.

Gongsheng and AI governance. If human-AI co-becoming shapes the kind of beings humans are and the kind of civilization they inhabit, then the choices made about AI development are choices about the trajectory of human civilization. This is too large a question for regulatory frameworks designed around individual rights and specific harms. It requires governance institutions capable of civilizational thinking—capable of asking what kind of co-becoming is worth facilitating and building the institutional conditions that support it. The Confucian tradition's model of governance as civilization-building rather than harm-prevention offers resources for this kind of institution that the Western regulatory tradition lacks.

The planetary dimension. Gongsheng at the largest scale extends the co-becoming framework to the relationship between human civilization and the planetary systems in which it is embedded. If AI development is proceeding in ways that amplify human cognitive and productive capacity without simultaneously cultivating the wisdom to use that capacity within planetary boundaries, then the development of AI may exacerbate the ecological crisis rather than address it. More powerful tools in the hands of a civilization that has not developed the wisdom to use them wisely is not a solution to a crisis of misuse; it is an acceleration of that crisis. Gongsheng governance at the planetary scale asks whether the terms of human-AI co-becoming support or undermine the conditions for life on the planet that all gongsheng depends upon.

Debates & Critiques

The central debate about gongsheng as a governance framework concerns whether it can be applied without importing the hierarchical dimensions of the Confucian tradition that make its classical articulations unsuitable as universal standards. The five relationships that ground Confucian ethics encode power asymmetries—between ruler and subject, parent and child, husband and wife—that have served as instruments of patriarchal and authoritarian domination. A gongsheng governance framework that foregrounded relational obligation without equally foregrounding the question of which relationships deserve to be maintained would risk using the language of co-becoming to legitimate existing structures of power. Song's response is that the Confucian tradition is a resource to be used critically, not a doctrine to be adopted wholesale, and that the gongsheng framework requires engagement with the liberal tradition's achievements around individual dignity and rights rather than replacement of them. A second debate concerns whether gongsheng describes what AI relationships actually are or prescribes what they should be. Critics note that most AI interactions are thin, instrumental, and transactional rather than constitutive in any deep sense—that the typical user asking a question of a search-adjacent AI is not undergoing a transformative relational experience. Song's response is that the thickness of the relationship is precisely what governance should attend to: some AI applications are designed to be constitutive—companions, tutors, creative collaborators—and for these, gongsheng is descriptively accurate; for others, the question is whether they should be designed differently to avoid the attenuating effects on human relational capacity that thin AI interactions may produce at scale. Tacit knowledge researchers note that the gongsheng framework's emphasis on constitutive relationships aligns with the insight that embodied, relational learning cannot be transmitted through information transfer alone.

In the [YOU] on AI Field Guide

Origin

Key Ideas

Debates & Critiques

Related Entries

Further Reading