Plants shape the visual language of gardens, streets and interiors. We recognise places through leaves, blossoms, rhythms of growth and seasonal change. In this part of my 2024 research into AI and visual culture, I consider how future links between computation, cultivation and breeding might change that language. The subject is both what we might grow and how we might learn to see living environments.

AI could help researchers compare genetic information, images and growing conditions to identify promising questions and breeding candidates. It would support biological research rather than turn an image into a viable plant. A proposed trait would still need to be developed and tested. Colour, form, nutritional value and resilience would not necessarily improve together; choosing among them would remain part of the work.

Red berries with unusual surface textures · Concept illustration
Red berries with unusual surface textures · Concept illustration

Historical varieties offer another direction. Seed collections, botanical records and genetic evidence could help recover knowledge about plants that have fallen out of cultivation. AI might make those connections easier to explore. Identifying a relationship is different from recreating a lost variety, however. I see an opportunity to reconnect landscapes and food cultures with their histories, without presenting a speculative reconstruction as an exact return.

In controlled growing environments, systems could combine observations of plants with light, temperature, water and nutrient management. AI might help adjust conditions or identify a problem that deserves closer attention. The goal would be healthier growth and more considered resource use. Whether a particular farm achieves that would depend on its crops, equipment, energy supply and operating decisions.

Fantastical trees with multicoloured rounded crowns · Concept illustration
Fantastical trees with multicoloured rounded crowns · Concept illustration

Breeding for future climates could bring researchers, growers and designers into a closer conversation. Models might help select candidates for testing under heat, drought or salinity. Success in one setting would not guarantee success elsewhere. Keeping diverse varieties and local knowledge in view could be as important as pursuing a single high-performing plant.

For visual culture, I imagine landscapes composed through time as well as space. Different flowering periods, leaf textures and growth patterns could create changing arrangements, while cultivation tools help people care for them. New ornamental varieties might expand this vocabulary. Their value would depend on how they live in a place, including their relationship with other organisms, not only how unusual they look.

The same approach could extend beyond growing to sorting, storage and distribution, helping teams test ways to reduce losses. Yet a new variety or a controlled environment would not automatically deliver ecological balance. Decisions about water, energy, propagation and introduction into a landscape would need evidence. Design could make those relationships visible instead of hiding them behind a green aesthetic.

The future I imagine is a richer conversation with living form. AI could help us draw on botanical knowledge and explore possibilities, while growth continues to unfold beyond the image on a screen. The most interesting result might be neither a perfect plant nor a spectacular novelty, but environments whose beauty includes diversity, change and the care needed to sustain them.