Dreaming in Blue Velvet, generative video with sound, 2023, 1 of 1
My work moves within the space where organic reality and artificial perception meet. I use neural networks not as tools that reproduce the world, but as systems that reinterpret it. Systems that develop their own logic, their own sense of what an image or a memory could be. Much of my practice grows from observing how these models learn, how they form associations, and how their internal transformations echo the way our minds drift between recollection, imagination, and dream.
This tension is present in works like After Us (2024), where urban structures and organic growth become inseparable. I see these images as hybrid organisms, shaped by both memory and algorithm. They feel like futures that nature might dream, or scenarios imagined by something nonhuman trying to understand our world. The neural network becomes a collaborator that mutates familiar environments into alternative ecosystems.
Within this context, Dreaming in Blue Velvet (2023) is a more intimate exploration of how a machine might dream. The work is built from my own custom model (GAN) trained on photographs of water, vegetation, and fleeting natural textures. In the video, forms shift according to a logic that isn’t strictly physical or narrative. Water thickens into structure, flora folds into liquid movement, and the scene slips in and out of clarity the way dreams do. The AI-generated soundtrack mirrors this sense of drifting consciousness, as if the system is searching for emotional meaning inside its own synthetic landscape.
I often think about how dreams have their own physics, their own storytelling rules. They reorganize our experiences into sequences that feel deeply personal but slightly abstracted. Dreaming in Blue Velvet plays with this idea by letting the neural network construct its own dream state, shaped by the textures and atmospheres it absorbed from my dataset. What emerges is not a depiction of nature, but an interpretation of its sensations.
Across these works, I am trying to understand what happens when human memory, machine learning, and the natural world intertwine. I explore how AI can become a way of thinking with images, a way of expanding the boundaries of perception.
After Us, generative video with sound, 2024, 1 of 1
Installation view, After Us, Project 22, Dubai, UAE, 2023.
Nondescriptives, generative series, 100 image stills, 2022.