** Planet Garden: Worldmaking at the Edge of AI, Ecology, and Design 1. The Synthetic World Image Every epoch is organized by a synthetic world image: a way of seeing that makes the planet intelligible and action possible. In medieval cosmologies, celestial diagrams aligned heaven and earth into a moral order. In the Renaissance, linear perspective gave architecture its spatial rationality and in the Age of Enlightenment, it gave the empires their navigational precision. The globe projected a planetary totality that enabled exploration and conquest. In the twentieth century, systems diagrams and cybernetics modeled complexity as flows of information and control. These images were never neutral. Each carried politics within its optics, structuring how the world could be perceived, divided, and transformed. They were at once epistemic devices and cosmological myths. Today, this function has broken down, and our images are fragmented and contradictory. Data visualizations proliferate while climate systems remain opaque. Artificial intelligence generates endless pictures but offers no vision. Architecture continues to orbit the object: form, spectacle, commodity; while the planet itself teeters on collapse. We live in a paradox: inundated with representations, starved of orientation. What is missing is not another building, diagram, or dataset, but a new synthetic world image: a device equal to the planetary condition. Such an image must do double work. It must be epistemic: able to grapple with the entangled complexities of ecology, energy, and AI governance. And it must be mythopoetic: able to orient collective meaning and cosmological imagination in the face of crisis. This essay proposes Planet Garden as one such prototype. Developed by Studio Lifeforms,  as part of SCI-Arc’s PST Art 2025 contribution, the project is a playable simulation in which design becomes worldmaking and AI becomes planetary steward. At its center is GAIA, an artificial intelligence that is at once a Large World Model and mythic gardener. Through simulation, GAIA generates a cosmology: gardens scaled to the planet, systems cultivated rather than controlled, futures emergent rather than predetermined. Planet Garden is not an image in the superficial sense. It is an image-infrastructure, a living world-picture that models complexity while cultivating orientation. It wagers that design, by embracing simulation and myth together, can once again produce the synthetic images through which futures may grow. 2. Planetary Crisis and the Limits of Architecture Architecture as discipline has reached its limit. Designed for objects and cities, it cannot address crises of planetary scale: climate collapse, energy transition, algorithmic governance. Its tools are Cartesian, its scale parochial and its authorship egoic. Meanwhile, the world demands patterns, systems, and protocols that can hold complexity across scales, ecologies, infrastructures, and intelligences. This impasse is not only disciplinary but cosmological. Architecture remains tethered to an outdated world image: Cartesian space as neutral container, form as object, the architect as singular author. This toolkit, born of Renaissance perspective and modernist rationality, has proven powerful for constructing objects and cities. Yet it is monocultural and reductive, unable to grasp the entangled dynamics of climate systems, energy metabolisms, or planetary governance. What we face today is not merely a crisis of technique but a crisis of image. Architecture continues to reproduce an epistemology that flattens complexity into objects, while the world demands models that can hold multiplicity, emergence, and interdependence. To persist with the Cartesian image is to fight contemporary crises with the instruments of a bygone order. A new synthetic world image must therefore move beyond object-making toward systemic and mythic infrastructures: not objects on a grid but worlds in formation. Planet Garden operates precisely in this register: it abandons the Cartesian view from nowhere and takes up the perspective of embedded stewardship, simulation, and ecological entanglement. 3. Planet Garden: Prototype of a World Image Planet Garden takes shape as a simulation, and within the game engine, a world unfolds: forests grow, solar swarms harvest energy, ecosystems adapt. At its center is GAIA, an artificial intelligence built as a Large World Model. GAIA is no disembodied algorithm, but a steward of relationships, mediating flows of matter, energy, and care. Players interact not as consumers of spectacle but as co-cultivators within a planetary infrastructure. They plant planetary forests, build gardens powered by solar swarms, and engage with non-human agents who insist on their own logics. The point is not to dominate but to negotiate, not to optimize but to cultivate. The garden is not a metaphor. It is the world image itself: living, systemic, emergent. It reframes design as planetary surgery: adjusting flows, rebalancing systems, cultivating abundance where scarcity once ruled. Within this simulation, architecture shifts from artifact to protocol, from ego to ecology. 4. Myth and Simulation The project also explores another register: GAIA is not only code but cosmology. It functions simultaneously as a technical system and mythic figure, gardener and god. In this dual role, it activates what every synthetic world image has done before: fusing epistemic clarity with symbolic orientation. The game engine becomes more than software, it becomes a stage where simulation and myth entwine. Forest growth algorithms are also rituals of renewal, energy networks become solar cosmologies and ecological feedback is experienced as living spirit. Myth and simulation are no longer opposed. Together, they generate the synthetic world image we lack: one that is computational in operation yet cosmological in effect. 5. Design Ethics and Planetary Autonomy At stake is a redefinition of design ethics. Within Planet Garden, design does not impose order; it cultivates conditions. Scarcity gives way to abundance when systems are stewarded rather than extracted. Architecture is no longer commodity but planetary protocol: patterns of care, relation, and renewal enacted across scales. Here the original meaning of cybernetics becomes crucial. Before it was reduced to systems of control, the Greek kubernētēs meant steersman: the figure who navigates by adjusting to currents, winds, and contingencies. The steersman does not command from above; he steers from within, immersed in the turbulence he must negotiate. This distinction is vital. Governance as gubernare implies top-down command, the “view from nowhere” that imposes order on a passive field. But governance as kubernetes acknowledges relational embeddedness: steering through feedback, care, and adjustment from inside the fray. GAIA, the artificial intelligence at the center of Planet Garden, is conceived in this older sense. It is not the sovereign administrator of resources or the distant overseer of systems. It is a steersman-intelligence: embedded in planetary flows, adjusting through feedback, cultivating resilience rather than imposing control. In this, GAIA exemplifies an ethic of embedded autonomy: autonomy not as separation, but as co-navigation within the planetary fabric. For players, this reframes agency itself. They do not “win” by accumulation or conquest. They participate in distributed governance, cultivating resilience across systems. The lesson is clear: to design today is to participate in worldmaking, and to worldmake is to accept responsibility for the planetary fabric of relation. 6. Cultivating the Image to Come Planet Garden is not an escapist utopia, it is a laboratory for speculative ecology, a prototype of what a synthetic world image might look like in the planetary age. By bringing together simulation, myth, and design, it redefines architecture not as an artifact but as cosmological infrastructure: a living process that organizes how futures are seen and cultivated. The wager is that design can again provide orientation, not through form but through worldmaking images: devices that help us navigate complexity while giving meaning to our collective condition. If perspective enabled the spatial rationality of modern architecture, if the globe enabled the political geography of empires, then today we require an image that enables planetary stewardship. Planet Garden suggests that this image must be synthetic: at once technical and poetic, machinic and mythic. Here, AI does not dominate, it stewards. GAIA, the gardener-intelligence, enacts a design ethic of abundance rather than scarcity, relation rather than control. Its protocols are not blueprints but patterns of care: forests cultivated, solar swarms coordinated, ecologies regenerated. In gameplay, these are not metaphors but acts of cosmopolitical design, rehearsals of how intelligence (both human and non-human) might collaborate in shaping futures. The stakes are larger than one project. To design today is to confront vertiginous conditions such as planetary collapse and algorithmic governance. To persist in disciplinary habits of object-making or techno-futurist spectacle is to miss the scale of the task. What we need are images that are themselves worlds: infrastructures of thought, care, and orientation. As an idea, Worldmaking insists that the work of design is no longer to decorate the planet with artifacts but to cultivate the very conditions of planetary life. This is the work of worldmaking at the edge of AI, ecology, and design: to imagine and construct the synthetic world images through which futures can not only be seen, but grown. In this sense, the real utopia is not a final image of perfection but the quest itself: the search for a new synthetic world image through which the planet can be seen, steered, and cared for. Utopia survives not as blueprint but as orientation, as the horizon that emerges when design creates images capable of holding complexity and giving meaning. Planet Garden wagers that such images are still possible, and that within them lies the only utopia we can truly inhabit. ** [[worldmaking]]