# created ```dataview LIST WITHOUT ID file.link FROM -"windsurf" WHERE file.cday = date(this.file.name) AND !startswith(file.folder, "windsurf") SORT file.cday ASC ``` # modified ```dataview LIST WITHOUT ID file.link FROM -"windsurf" WHERE file.mday = date(this.file.name) AND !startswith(file.folder, "windsurf") SORT file.mday ASC ``` --- # [[diary and thoughts]] #thought #diary --- # Slides for [[the environmental aesthetics of generative AI]] 5 Jun 2025 #### Work in Progress - June 2025 --- ### Slide 2 ## Introduction: [[The Challenge]] of AI Art * **[[Traditional Art]] (Human-made):** * Can take months or years. * Direct result of individual labor, attention, effort. * Involves human intention. * **Generative AI Output (e.g., Midjourney, ChatGPT, Sora):** * Produced in seconds. * Result of computational processes. * Lacks human-like intention (at least in the same way). --- ### Slide 3 ## Two Options for Appreciating AI Art 1. AI art is **not** the sort of thing that can be aesthetically appreciated. * *[[This seems]] too dismissive.* 2. We appreciate AI art in a **different way** than we do [[traditional art]]. * *This paper develops this option.* *But what way is this?* --- ### Slide 4 ## Drawing from [[Environmental Aesthetics]] * This presentation argues for an analogy: * Appreciating the **[[natural environment]]** (Carlson's model) * Appreciating **generative AI outputs**. * **Core Idea:** Just as nature isn't just a collection of pretty objects, AI art isn't just a collection of pixels or words. --- ### Slide 5 ## Appreciating Nature: Carlson's [[Natural Environmental Model]] > First, that, as in our appreciation of works of art, we must appreciate nature as what it in fact is, that is, as natural and as an environment. Second, it recommends that we must appreciate nature in light of our knowledge of what it is, that is, in light of knowledge provided by the natural sciences, especially the environmental sciences such as geology, biology, and ecology. > – (Carlson 2000, p. 6) --- ### Slide 6 ## The [[Natural Environmental Model]] Explained * **Appreciate nature for what it is:** * As natural (not artificial). * As an environment (a system of interconnected elements and processes). * **Role of Knowledge:** * Appreciation is informed by [[scientific knowledge]] (geology, biology, ecology). * This knowledge reveals the underlying processes and unity. * Don't appreciate a forest as if it's a painting (static, isolated). --- ### Slide 7 ## Example: Beachy Head * **Uninformed Observer:** * Admires its dramatic height, striking white color (surface features). * **Informed Observer (Carlson's Model):** * Appreciates what the cliff *is*: * A stratified chalk formation. * Shaped by geological uplift and marine erosion. * Part of a broader shoreline ecosystem. * This scientific cognizance reveals unity among disparate features. --- ### Slide 8 ## Unity in Nature: The Environment of Creation > natural objects possess [...] an [[organic unity]] with their environments of creation: such objects are a part of and have developed out of the elements of their environments by means of [[the forces]] at work within those environments. Thus the environments of creation are aesthetically relevant to natural objects. > – (Carlson 2000, p. 44) * Aesthetic character emerges from understanding [[the object]] as part of a larger, dynamic process. --- ### Slide 9 ## Spinoza: [[Natura Naturans]] & Natura Naturata * Carlson's "unity" is complemented by Spinoza's distinction: * **Natura Naturans ("naturing Nature"):** * Active, productive aspect of nature. * The ongoing generative processes and laws. * **Natura Naturata ("natured Nature"):** * Passive, product aspect. * The concrete world of forms and events resulting from those laws. * **Key:** Not two worlds, but two perspectives on one unified Nature. Process and product are inseparable. --- ### Slide 10 ## Unity, Natura, Appreciation * Informed aesthetic appreciation of nature involves understanding the relationship between *natura naturans* and *natura naturata*. * **Examples:** * Geological forces (*naturans*) → Coastal cliff formations (*naturata*). * Photosynthesis, genetic adaptation (*naturans*) → Flower structures (*naturata*). * We appreciate the product in light of the process. --- ### Slide 11 ## Appreciating Machina: AI as a Generative Environment * **The Analogy:** Generative AI systems can be understood as artificial environments with their own generative dynamics. * **Principle:** Appreciate AI outputs "for what they in fact are" – products of a generative computational process. --- ### Slide 12 ## Machina Naturans & Machina Naturata * Extending Spinoza's terms to AI: * **Machina Naturans ("naturing Machine"):** * The AI's generative processes: algorithms, neural network architecture, training methodologies. * The active, artificial generative process. * **Machina Naturata ("natured Machine"):** * The AI's outputs: images, texts, videos. * The resulting artifacts. * To appreciate AI *machina naturata*, one must apprehend it as the result of an underlying generative unity (*machina naturans*). --- ### Slide 13 ## Latent Space as Cultural 'Soil' * Generative AI (LLMs, image generators) uses **latent spaces**: * High-dimensional structures encoding patterns from vast cultural corpora (text, images). * A compressed archive of training data. * **Metaphor:** * Training data (human culture) ≈ rich compost. * Latent space ≈ fertile soil formed by decomposition. * The AI "digests" cultural artifacts into abstract representational forms – a "multi-dimensional archive of culture." --- ### Slide 14 ## From Generative Substrate to Emergent Artifact * **Generation Process:** * A prompt acts like a seed in fertile soil. * The AI's *machina naturans* (autonomous generative mechanism) develops the output, drawing on latent "nutrients." * **Autonomy:** * The process unfolds via the model's internal learned rules, not direct human design of every detail. * Resembles nature's *natura naturans*. * Appreciating an AI artifact involves seeing it as an "emergent cultural flower." --- ### Slide 15 ## Appreciating an AI-Generated Landscape * Imagine an AI landscape resembling Impressionism. * **Misguided Appreciation (if assumed human-painted):** * Focus on non-existent artist's intentions or skills. * **Informed Appreciation (knowing AI origin):** * Marvel at how *machina naturans* synthesized the image from learned patterns. * Wonder how the model evoked specific styles or realism. * Appreciation includes the generative process itself. --- ### Slide 16 ## Participatory Appreciation: The Prompter as Gardener * Carlson acknowledges direct participation in natural processes (e.g., gardening) as a mode of aesthetic appreciation. * **Analogy:** * The AI **Prompter** is to *machina naturans*... * ...as a **Gardener** is to *natura naturans*. * This is a practice-based, participatory aesthetic engagement. --- ### Slide 17 ## Guiding a Partly Autonomous Process * **Gardener:** Provides conditions, guides; nature executes flower formation. * **Prompter:** Inputs prompt, adjusts parameters; AI (*machina naturans*) produces output. * **Shared Characteristics:** * Skillful guidance, not absolute control. * Interaction with a system having "dynamic recalcitrance." * Iterative refinement and adaptation. * Creativity as partnership/stewardship. --- ### Slide 18 ## The Prompter's Situated Knowledge * Hands-on engagement develops a tacit understanding of the AI model's: * "Personality" * Tendencies, strengths, limitations. * Analogous to a gardener's practical knowledge of soil, weather, plant responses. * Allows appreciation of subtleties in AI outputs. * Aesthetic satisfaction comes from the interactive process and its culmination. --- ### Slide 19 ## Aesthetic Profundity for the Informed Spectator * What about those who observe AI art without prompting it? * Analogous to a hiker admiring a wildflower they didn't plant. * **Key (Carlson):** Appreciate the artifact for "what it in fact is" – recognizing it as a product of an AI generative process. > This model suggests that in aesthetic appreciation of anything... appreciation must be centered on and driven by the real nature of the object of appreciation itself... dependence on and guidance by means of knowledge, scientific or otherwise, that is relevant given the nature of the thing in question. > – (Allen Carlson) --- ### Slide 20 ## Knowledge as a Bridge to Appreciation * Knowledge of natural history enhances landscape appreciation. * Similarly, for AI art, knowledge of its generative origin attunes the spectator: * Perceives convergence of influences orchestrated by an algorithm. * Recognizes emergent quirks vs. human artistic signatures. * Sees *machina naturata* (output) in relation to *machina naturans* (process/dataset). * Prevents category errors (e.g., praising non-existent brushwork). * The knowledge need not be deep technical expertise, but a general understanding of origin. --- ### Slide 21 ## Imagination and Aesthetic Profundity * Informed imagination allows the spectator to: * Conceive of the artwork as a machinic creation. * Imagine the vast data and iterative processes behind the output. * Can yield a sense of **aesthetic sublime**: * The artwork points beyond itself to the complexity of its creation. * Parallels profundity in nature: connecting the perceiver to a grand creative force (natural processes vs. cultural/computational convergence). * The artwork encapsulates a "cultural and computational journey." --- ### Slide 22 ## Conclusion: Key Arguments * **Analogy:** Environmental aesthetics (Carlson) ↔ AI art aesthetics. * **Framework:** *Machina naturans* (AI generative process) & *machina naturata* (AI product). * **Core Principle:** Appreciate AI outputs in relation to their generative context, as the kind of things they truly are. * **Modes of Appreciation:** * **Participatory:** The prompter as "digital gardener." * **Contemplative:** The informed spectator understanding generative origins. --- ### Slide 23 ## Virtues of this Framework * **Coherence:** Unifies thinking about AI aesthetics. * **Explanatory Power:** * Explains exploratory, discovery-laden use of AI. * Addresses the "lack of human intention" by focusing on AI's dynamics. * Sidesteps "who is the artist?" – art emerges from human-machine interaction. * **Fosters a Stance:** Respectful fascination with AI's complexity and the interplay of input and autonomous development. * Suggests a new kind of beauty born from code and data, resonating with ideals of unity and creative power. --- ### Slide 24 ## Quoting Chris Olah (Co-founder Anthropic) To emphasize the 'grown' nature of AI: > I think one useful way to think about neural networks is that we don't program and we don't make them. We kind of, we grow them...we have these neural network architectures that we design and we have these loss objectives that we create. And the neural network architecture, it's kind of like a scaffold that the circuits grow on...it's this almost biological...entity or organism that we're studying. > – Chris Olah ---