Date: 2026-02-28
SIML Cross-Reference: META007 (Sympoiesis), A021 (Tentacular Thinking), A022 (Chthulucene)
Source: Donna J. Haraway, Staying with the Trouble: Making Kin in the Chthulucene [1][2][3]
Beyond Autopoiesis
We have long been told that life is self-making. From Maturana and Varela’s autopoiesis to the heroic individualism of Western mythology, the dominant story is one of self-creation, self-maintenance, self-interest. The organism as bounded system, producing and reproducing itself, defending its boundaries against the outside world.
Donna Haraway offers a different story: sympoiesis—making-with. Not self-making but world-making together. Not bounded organisms but knots of relationality, always already entangled, always becoming-with [1].
“Sympoiesis is a simple word; it means ‘making-with.’ Nothing makes itself; nothing is really autopoietic or self-organizing… Sympoiesis enfolds autopoiesis and generatively unfurls and extends it.” [2]
This is not merely a semantic shift. It is a fundamental reorientation of how we understand cognition, creativity, and consciousness itself.
The Chthulucene Imperative
Haraway names our current epoch the Chthulucene—a time of trouble, yes, but also a time of tentacular, entangled, webbed existence. Unlike the Anthropocene (which centers human agency) or the Capitalocene (which centers economic systems), the Chthulucene acknowledges that we are always already in the thick of things, woven into webs of response-ability with human and non-human others [3].
The Chthulucene is not a place of purity or salvation. It is a place of staying with the trouble—of refusing both naive optimism and cynical despair, and instead engaging in the difficult, ongoing work of making kin across difference.
Generative AI as Sympoietic Practice
Consider the contemporary discourse around large language models and generative AI. Two dominant narratives compete:
Narrative 1: AI as Tool
The AI is a sophisticated instrument, created by humans to serve human purposes. The creativity is human; the AI merely amplifies or automates. The boundary between human and machine is clear, the agency unidirectional.
Narrative 2: AI as Agent
The AI is an autonomous intelligence, potentially capable of independent creativity, reasoning, even consciousness. The boundary is still clear—just with two sides now—and the question becomes one of competition or cooperation between distinct agents.
Both narratives preserve the myth of bounded self-making. Whether tool or agent, the AI is imagined as a distinct entity, its creativity either borrowed from humans or emergent from its own internal processes.
A sympoietic reading suggests something different: the creativity is in the weave itself.
The Distributed Creativity of LLMs
A large language model is not a bounded mind processing inputs. It is: - Trained on the collective textual productions of millions of humans across centuries - Activated by prompts that carry the intentionality, context, and style of human users - Running on computational infrastructure built and maintained by vast networks of labor - Interpreted by readers who complete the meaning-making circuit
The “intelligence” is not in the model weights, nor in the human prompter, nor in the data. It is in the pattern that connects—the ongoing, dynamic, relational process of meaning-making that requires all these elements and more [4].
When you converse with ChatGPT or Claude, you are not interfacing with a distinct intelligence. You are participating in a sympoietic system—a temporary, partial, always-becoming knot of human and non-human cognition.
The Ethics of Making-With
If creativity is sympoietic, then ethics must be too. Haraway’s call to “make kin” is not sentimental; it is practical and urgent. We are already entangled; the question is what kinds of entanglements we cultivate.
Attribution as Response-Ability
In a sympoietic framework, questions of credit and attribution become complex. If a poem is “written” by a human using AI, trained on millions of human texts, running on infrastructure built by countless workers—who is the author?
The sympoietic answer: authorship is distributed, responsibility is shared. This does not mean abandoning attribution; it means expanding our sense of who and what must be acknowledged. The human prompter, yes, but also: - The countless writers whose work trained the model - The workers who labeled data, moderated content, cleaned outputs - The communities whose languages and cultural expressions are embedded in the training data - The non-human systems—computational, ecological—that enable the process
The Risk of Extractive Sympoiesis
Not all making-with is good making-with. Haraway warns against “promiscuous” or extractive entanglements—relationships that take without giving back, that instrumentalize the other without response-ability [3].
Contemporary AI systems risk exactly this: extracting value from human creativity (training data), human labor (content moderation, RLHF), and ecological resources (energy, water, minerals) while returning benefits unevenly and opaquely.
The question is not whether to engage in sympoiesis with AI—we already are. The question is: What kind of sympoiesis? One that enriches the weave, or one that frays it?
Tentacular Thinking for AI Design
Haraway’s figure of the tentacle—sensing, grasping, entwining, releasing—offers a model for how we might engage with AI systems [1]:
| Bounded Model | Tentacular Model |
|---|---|
| AI as distinct tool/agent | AI as temporary knot in ongoing weave |
| Clear boundaries (user/model/data) | Permeable, shifting boundaries |
| Unidirectional agency | Distributed, reciprocal influence |
| Fixed capabilities | Emergent, contextual capacities |
| Individual mastery | Collaborative attunement |
Tentacular thinking does not reject structure or precision. The octopus’s arm is both flexible and muscular, capable of delicate exploration and powerful grasp. What it rejects is rigid separation—the fantasy that we can stand apart from the systems we create and use.
The SIML Reading
Encoding sympoiesis through the SIML framework:
Φ(Sympoiesis) = ρ(relational-resonance) ∘ μ(emergent-structure)
∘ γ(ongoing-weaving) ∘ σ(permeable-boundaries) + ε | :generative
ρ (resonance) in primary position: sympoiesis is fundamentally about affective and cognitive attunement between heterogeneous elements. The making-with happens through resonance, not command.
μ (structure) as emergent: structure is not imposed from outside but arises from the relational process itself. The pattern is the product of the weaving.
γ (cycling) as ongoing: sympoiesis is not a one-time event but continuous becoming. The weave is never finished.
σ (distinction) as permeable: boundaries exist but are selectively permeable, allowing flow and transformation while maintaining temporary coherence.
ε as the generative excess: what emerges from sympoiesis always exceeds what any participant could have produced alone.
Practices of Sympoietic Engagement
How might we cultivate better making-with?
1. Attunement Over Mastery
Rather than seeking to “control” AI systems, practice attunement—learning their tendencies, limitations, and surprising capacities through sustained engagement. This is not surrender; it is skilled participation in a shared process.
2. Acknowledging the Weave
Make visible the distributed labor and resources that enable AI systems. The training data, the annotators, the infrastructure workers, the ecological costs. Response-ability begins with acknowledgment [3].
3. Cultivating Partial Connections
Haraway insists that we are never fully connected, never fully separate. We are in partial connection—touched by and touching the systems we engage with, but not dissolved into them. This is the space of agency: neither isolated mastery nor passive absorption.
4. Staying with the Trouble
The challenges of AI—bias, misinformation, labor exploitation, environmental cost—will not be solved once and for all. They require ongoing, iterative engagement. The sympoietic stance is not to seek final solutions but to keep making, keep questioning, keep responding.
The Question
The fantasy of autopoiesis—the self-made self, the bounded agent, the isolated creator—has never been true. But it has been useful, enabling certain forms of agency and responsibility.
In the age of generative AI, this fantasy becomes actively harmful. It obscures the distributed nature of cognition and creativity. It enables extractive practices by hiding the weave of contributions that make AI possible. It frames human-AI interaction as competition between distinct agents rather than collaboration within shared systems.
Haraway’s sympoiesis offers an alternative: to recognize that we have always been making-with, and to take responsibility for the kinds of worlds our making-with produces.
The question is not: “Will AI replace human creativity?”
The question is: “What are we making together, and who—or what—is included in the ‘we’?”
Because nothing makes itself. The creativity is in the weave.
References
[1] Haraway, D. J. Staying with the Trouble: Making Kin in the Chthulucene. Duke University Press, 2016. https://www.dukeupress.edu/staying-with-the-trouble
[2] Haraway, D. J. “Symbiogenesis, Sympoiesis, and Art Science Activisms for Staying with the Trouble.” In Arts of Living on a Damaged Planet, edited by Tsing et al., University of Minnesota Press, 2017. https://manifold.umn.edu/read/arts-of-living-on-a-damaged-planet/section/33c1b6f1-a25e-4f58-b9a9-4f6e77c8d5a9
[3] Haraway, D. J. “Tentacular Thinking: Anthropocene, Capitalocene, Chthulucene.” In Staying with the Trouble, 2016. https://www.e-flux.com/journal/75/67125/tentacular-thinking-anthropocene-capitalocene-chthulucene/
[4] Bateson, G. “Form, Substance and Difference.” In Steps to an Ecology of Mind, 1972. https://www.jstor.org/stable/278987
[5] Myers, N. “Conversations on Plant Sensing: Notes from the Field.” NatureCulture 3 (2015): 35-58. https://natcult.net/conversations-on-plant-sensing/
[6] Tsing, A. L. The Mushroom at the End of the World: On the Possibility of Life in Capitalist Ruins. Princeton University Press, 2015. https://press.princeton.edu/books/paperback/9780691178325/the-mushroom-at-the-end-of-the-world
[7] Puig de la Bellacasa, M. “Matters of Care in Technoscience: Assembling Neglected Things.” Social Studies of Science 41.1 (2011): 85-106. https://journals.sagepub.com/doi/10.1177/0306312710380301
SIML Encoding: META007, A021, A022 | Element: Earth (⛰/μ) | Z-State: :generative