Appendix F — Digital Art Categories

Comprehensive Guide on Segments of the Digital Art Market

This document provides a self-contained reference for our final set of digital artwork categories. It is intended for anyone who was not part of the original discussions yet needs to understand exactly how and why we arrived at these categories, and how we classify artworks (or artists) within them.

We started with three different “sets” (or approaches) to categorizing digital art. Each set came from different research sources or classification tasks (some from surveys, some from an internal best-effort assignment). Our goal was to synthesize them into a final “source of truth” that:

  1. Accurately reflects the breadth of digital artwork styles and methods,
  2. Accounts for subdomains like AI art, generative vs. non-generative, or physical-digital hybrids,
  3. Simplifies some redundancy yet preserves enough detail to distinguish key subgroups (e.g. glitch art, pixel art).

By the end of our process, we aimed to produce a classification scheme that anyone could pick up and apply consistently.

However, some widely referenced research identifies Digital Fine Arts as the largest subsegment, whereas our data shows a different picture.1

Taking these factors into account, we find the frequently cited 25% figure for Digital Fine Arts’ share of the broader digital art and collectibles market to be both plausible and, in our view, the most reliable current estimate.

Why We Created These Categories

We started with three different “sets” (or approaches) to categorizing digital art. Each set came from different research sources or classification tasks (some from surveys, some from an internal best-effort assignment). Our goal was to synthesize them into a final “source of truth” that:

  1. Accurately reflects the breadth of digital artwork styles and methods,
  2. Accounts for subdomains like AI art, generative vs. non-generative, or physical-digital hybrids,
  3. Simplifies some redundancy yet preserves enough detail to distinguish key subgroups (e.g. glitch art, pixel art).

By the end of our process, we aimed to produce a classification scheme that anyone could pick up and apply consistently.

G Final Categories: Overview

We arrived at ten (10) high-level categories:

  1. Digital Fine Art
  2. Generative Art
  3. AI Generative Art
  4. Photography
  5. Video/Animation
  6. Glitch Art
  7. Pixel Art
  8. Experimental/Interactive
  9. Collectibles
  10. Blockchain-Based / Conceptual

In addition, we allow subcategories whenever we see more specific styles (like gaming, music, 3D, phygital, pixel, etc.) that fit under the main category. Below, we define each category in detail, discuss why we included it, and give examples from the data.

H Category Definitions and Rationale

1. Digital Fine Art

Description

Digital Fine Art covers artworks created with a “traditional” artistic skillset or mindset, but executed or exhibited digitally. This includes digital painting, illustration, or even scanned physical works if the focus remains on painting, sketching, or other standard “fine art” processes. We also place “phygital” items here if they revolve around gallery-worthy artistry (like a sculptor bridging physical sculpture with a digital NFT).

Examples - Beeple: He is well-known for extremely detailed digital painting and compositing. - Matt Kane: Creates vibrant, painterly digital pieces. - Damien Hirst: Bridges physical paintings with NFT representations (subcategory: “Phygital”).

Why It’s Separate

We wanted a dedicated space for artistry that strongly resembles the “canvas-based” or “hand-drawn” tradition, rather than pure code-based or purely collectible projects. This helps us spotlight artworks where human painting/drawing remains central to the creation process.

2. Generative Art

Description

Generative Art includes non-AI algorithmic, parametric, or code-based artworks. It could use fractals, randomization in code, or procedural geometry, but does not explicitly rely on machine learning or neural networks.

Examples - Dmitri Cherniak (creator of “Ringers”): Algorithmic lines and shapes. - Casey Reas: Pioneer in Processing-based art. - Rich Lord: Abstract, code-driven geometry.

Why It’s Separate

We found that in many classification attempts, “generative” was intermingled with AI or with conceptual approaches. We decided to isolate “pure” generative code from AI-based approaches, given many collectors or data sets treat them differently.

3. AI Generative Art

Description

AI Generative Art is specifically reliant on machine learning or neural networks to produce visual output (e.g., DALL-E, stable diffusion, custom ML models). If the mention of “AI” is in the detail or the project specifically markets itself as AI-driven, it goes here.

Examples - Botto: An AI “autonomous artist” using ML to generate art. - Claire Silver: AI-assisted generative visuals. - Refik Anadol: Known for large-scale AI immersive experiences.

Why It’s Separate

Surveys and data show a distinction between purely algorithmic vs. AI-based. Collectors often track them differently. So we keep “AI Generative Art” top-level.

4. Photography

Description

Photography refers to NFT collections or artists primarily using real-world photographic processes. If it is purely photographic, it fits here. If AI or generative coding modifies it extensively, we often place it under AI Generative or Generative if that is the project’s main focus.

Examples - Justin Aversano: “Twin Flames” series is photography-based. - Driftershoots: Famous cityscape and rooftop photography NFTs.

Why It’s Separate

Photography is a large and established sub-market within digital collectibles. Setting it apart from painting/illustration or generative code is practical for clarity.

5. Video/Animation

Description

Video/Animation includes short-form video pieces, moving illustrations, or 3D animations rendered via software. We also put 3D-based digital sculptures/animation here. If the 3D style is purely “fine art” painting in 3D, we might instead choose Digital Fine Art, but typically 3D software or animation loops land in this category.

Examples - Everkender: 3D-based digital sculptures/animations (subcategory “3D”). - DeeKay: Known for fun, stylized animated NFTs. - Kinetic Visuals: 3D animations.

Why It’s Separate

Motion-based digital art can differ significantly from still paintings or generative code. By isolating “Video/Animation,” we highlight time-based or 3D software-based mediums.

6. Glitch Art

Description

Glitch Art focuses on digital “errors,” data corruption, or intentionally broken images. It typically has a distinct aesthetic that embraces chaos or distortions.

Examples - Bard Ionson: Glitch visuals with abstract data corruption. - Robness: Glitch plus conceptual statements. - Kim Asendorf: Combines glitch manipulations with generative processes.

Why It’s Separate

Glitch is a recognized style, historically important in digital art. We do not bury it inside “fine art” unless the detail strongly indicates painting or drawing skill is dominant.

7. Pixel Art

Description

Pixel Art means low-resolution, retro-style imagery reminiscent of early video games or 8-bit/16-bit aesthetics. If the detail says “retro pixel” or references old-school gaming sprites, we place them here.

Examples - CryptoPunks or derivatives, if not purely collectible. - Pixel Reverse, Retro Gridworks.

Why It’s Separate

Pixel Art is a specific, easily recognized style. We want to track it distinctly rather than bury it under “Digital Fine Art.”

8. Experimental/Interactive

Description

Experimental or Interactive Art often involves new media, heavy interactivity, or conceptual installations that do not cleanly fit into generative or painting. If it’s extremely conceptual or performative but not obviously on-chain conceptual, it goes here.

Examples - Emi Kusano: Experimental NFT experiences. - Thomas Webb: Tech-driven interactive projects.

Why It’s Separate

This category captures ephemeral, user-involved, or truly experimental pieces that do not revolve around painting or code-based generative but rather user interaction or performance.

9. Collectibles

Description

Collectibles covers PFP (profile picture) projects, music NFTs focusing on collectible aspects, gaming assets, domain names, or brand collaborations. If it is “character-based” or “large set” driven, we see it as a collectible.

Subcategories - Music: If the detail specifically says “music collectibles.” - Gaming: If it’s in-game skins, virtual worlds, or assets. - PFP: If it’s a large set with unique traits, e.g. Bored Ape Yacht Club. - General: For everything else collectible not specifically music or gaming.

10. Blockchain-Based / Conceptual

Description

Blockchain-Based Art prioritizes heavy conceptual ideas about crypto, decentralization, or network-native functions. Often includes performance, data-visual NFTs, or multi-chain experiments.

Examples - Pak: Blockchain-centric conceptual projects. - Terra0: Projects focused on DAOs and ecosystems. - Moxara: Experiments around multi-chain narratives.

Why It’s Separate

It combines innovation with heavy use of decentralized tech that does not prioritize traditional art techniques.

Subcategories

Beyond these ten categories, we often assign subcategories to highlight more specific aspects. For example:

  • ‘Music’ or ‘Gaming’ under Collectibles
  • ‘Phygital’ under Digital Fine Art (if bridging physical and digital)
  • ‘3D’ under Video/Animation
  • ‘Illustration’ or ‘Painting’ under Digital Fine Art if the detail clarifies the style

We do not split them further unless the detail strongly justifies it.

I Decisions and Motivations

  1. Why separate AI from Generative?
    • Collector data often isolates machine learning or neural nets as a separate phenomenon. We saw real distinctions in multiple research sources (e.g., 30 percent focusing specifically on AI-based approaches).
  2. Why keep Glitch or Pixel distinct?
    • These are historically recognized styles. We preserve them if the detail strongly suggests a glitch or pixel aesthetic. If it also says “fine art skill,” we might place it in Digital Fine Art and note ‘Glitch’ or ‘Pixel’ as a subcategory.
  3. Physical-Digital Art as Digital Fine Art
    • We consider the nature of “phygital.” If it’s an artist bridging real and digital, we treat it as Digital Fine Art with the subcategory “Phygital,” especially if galleries are involved.
  4. Subcategory for Collectibles
    • We recognized many subtypes: music, gaming, brand collabs, or just PFP. Consolidating them all under “Collectibles” helps unify the classification.
  5. Handling Overlaps
    • If an artist is “AI + Photography,” we typically choose AI Generative Art with subcategory “Photography” if the code truly drives the final piece.

J Illustrative Examples

  • Beeple was originally labeled “Digital Painting” in one set but had some animations in others. We resolved to place him in ‘Digital Fine Art’ (subcat ‘Painting’) because his work is widely recognized as painting/illustration-centric.

  • Deafbeef was “Generative Audio Art,” so we classified them as ‘Generative Art’ with subcategory ‘Music’ because we see code-based, non-AI processes merging with audio.

  • Damien Hirst had “Physical-Digital” bridging. We placed him in ‘Digital Fine Art’ subcategory ‘Phygital’ due to the synergy with galleries and painting tradition.

K Classification Logic Summary

  • We read each row, focusing on the detail column over the label if there’s a conflict.
  • If the detail says ‘AI,’ we pick ‘AI Generative Art’ as the category.
  • If it’s clearly painting, illustration, or a physical piece minted as fine art, we pick ‘Digital Fine Art.’
  • If it references large sets, brand collabs, or PFP, we choose ‘Collectibles,’ adding subcategories if it’s music or gaming.
  • If it’s conceptual on-chain, we choose ‘Blockchain-Based / Conceptual.’
  • Pixel or glitch aesthetics stand alone, unless we see strong reason to merge with Fine Art.
  • 3D animations go under ‘Video/Animation’ with subcategory ‘3D.’
  • We do not subdivide further unless details specifically justify it.

L Conclusion

Our final categories ensure that:

  • The entire range of digital artistry from classic painting, photography, and glitch, up to purely generative or AI code, is captured.
  • Major market segments like PFP or conceptual on-chain art remain distinguishable.
  • We can preserve smaller styles (glitch, pixel, experimental) that matter to art historians or collectors.

  1. Although widely referenced industry research identifies Digital Fine Arts as the largest subsegment, our data indicates that Digital Collectibles dominates among the artists in our analysis. On closer inspection, we discovered this $1.6 billion valuation was disproportionately driven by just two collections—most notably CryptoPunks, which exceeds $1 billion and makes up roughly 65% of the category’s market cap. Additionally, our classification choices for Generative Art may have caused certain borderline projects to be categorized differently than in other published reports.↩︎