Transforming Abstract Drawings into Generated Architectonic Constructions
Abstract:
With the rise of generative tools, a sudden confusion has emerged regarding the fundamental understanding of what Art & Design Artefacts might be and could become in the near future. In these turbulent times, it has become evident that some creatives, eager to utilise these generative tools, are beginning to create Things that cannot be accurately categorised within the known realm of Art & Design Artefacts. A new sub-category of Artefacts has come into existence: non-physical objects that are dematerialized yet appear real, forming a distorted spatial realm. These Artefacts, non-existent in plain reality, are partly human - partly fictional, echoing the sounds of an imaginary Piranesian playground.
The generated output initially appears as simple sensory impressions, lacking the original authorship of a human creator. Subsequently, a personal touch —automatic drawings— is incorporated and used as a synthesising guide. Through predefined repeating instructions, the AI image generator transforms the abstract drawings into ensemble sets of images. A carefully selected body emerges, a curated slice from the whole, shining through these abstract, nonrepresentational experiments.
Initially, {Human} drawings and {Artificial} generations seem unrelated. Drawings are human, and generations are artificial. Yet, when we merge the two, we find ourselves in contradiction. What seemed distinct and separate has suddenly fused into a chaotic, yet occasionally pleasant, entanglement of forms. The output {generated data} reveals instances of artistic expression, which at irregular intervals transform the abstract input into more elaborate organic three-dimensional structures. An unforeseen hint of architecture, where architecture itself emerges in a generated image.
To further narrow the area of latent space in which the generations occur, a personalized ranking is created of generations produced with slightly different prompts (variation in material), prioritizing the most peculiar results. The parameters and settings that were intuitively discovered to achieve these unique results are then modelled into a virtual machine: a series of sequentially operating notes through which the transformations materialize. The resulting script forms the base for a more effective and streamlined workflow, while still retaining a desired degree of chaotic behaviour {hallucination}.
(1) Compositional Affects in Generated Images: An Examination
(2) Morphological Transformers: Automatic Drawing into Synthesised Visual Representations
(3) Metamorphic Models: Curated from a Sampling Set of Personalized Generations
(4) From Metamorphic Model towards Implementation: A Case Study
(5) Reflecting on Spatial Form-Finding with AI: Insights and a Vision for the Future
[6] Future Worlds
Keywords: Multidimensional, Latent Spaces, Compositional Affects, Prompt Playing, Looping Problems, Purifiers, deep learning models, generative process, diffusion, distortion, compression, biases in datasets, compilations, fine-tuning, transformers, infinite possibilities, curation, originality.
Ⓜ Marcel Moonen 01.06.2024, Transforming Abstract Drawings into Generated Architectonic Constructions [EDUCATIONAL PURPOSE ONLY] Triple-A Society, M. Production