Pith. sign in

REVIEW 3 major objections 5 minor 70 references

Unlimited Editions: Documenting Human Style in AI Art Generation

T0 review · 3 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read HCI's real task with AI art is to document human style lineage, not to detect fakes, this paper argues.

desk verdict A clear, well-argued position paper that reframes HCI's role in AI art from detection to documentation, but its proposed technical system would document visual provenance, not the 'creative struggle' it defines as style. read the letter →

arxiv 2507.19497 v1 pith:7YOQY4L5 submitted 2025-06-05 cs.HC cs.AIcs.CYcs.IR

Alex Leitch , Celia Chen This is my paper · ORCID
classification cs.HCcs.AIcs.CYcs.IR
keywords AIartgenerationstylisticlineageprovenancedocumentationartisticvaluegenerativemetadatastandardshuman-computerinteractionstyletransfer
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper argues that HCI research on AI art has aimed at the wrong target: detection, authenticity, and generation quality. The authors claim that artistic value derives not from visual appearance alone but from the finite, struggle-filled human process that produced a style. Because every human artist dies, their oeuvre is limited, and that scarcity is what lets art carry cultural and market value. AI systems that endlessly reproduce a style flatten away this provenance. The paper therefore proposes that HCI's job is to build documentation systems that trace the human origins and evolution of style inside generated images.

What carries the argument

The central mechanism is a proposed provenance stack: at the training-data level, attribution graphs that record which artists' works influenced which model weights and how strongly; at the generation level, a per-channel, per-pixel attribution layer, stored in XMP metadata, that records what percentage of each pixel comes from each piece of the trained model. The paper defines two anchors for this machinery: the catalog raisonné (a comprehensive annotated listing of all known artworks by an artist) as the historical model for documentation, and a comparison with prior metadata-comparison work that uses embedded metadata to discover image changes. Reversibility in style transfer is the technical property that would make such attribution auditable.

What would settle it

Generate an image with a publicly known diffusion model and its training set, then attempt to compute the proposed per-pixel percentage attributions from the model's internals. If the decomposition cannot be computed or is non-additive (e.g., contributions change when other training samples are removed), the proposed documentation system cannot be built as specified.

Watch

Extended reading notes

Core claim

On the paper's own account, artistic style is not merely a visual surface but the residue of creative struggle: artists wrestle with influences, technical constraints, and a limited lifespan to produce a unique way of seeing. Because each artist's output is finite, the paper argues, art functions as a store of value only through systems of authentication and provenance such as the catalog raisonné. AI image generators ingest and reproduce style without this provenance, flattening singular human achievement into endlessly reproducible patterns. The discovery the paper stakes out is that the meaningful HCI problem is not detecting AI output but documenting stylistic lineage—building attribution graphs at the training-data level and embedding per-pixel provenance in the generated image's metadata.

Load-bearing premise

The proposal depends on the assumption that a trained generative model can be reverse-engineered to compute, for each pixel of an output image, the exact percentage contributed by each piece of training data; the paper gives no evidence or prototype that this decomposition is technically achievable.

Editorial extensions

If this is right

  • HCI evaluation of AI art systems should be re-centered on 'documentation fidelity'—how well a system records the human decisions and influences behind an output—rather than on generation quality or detection accuracy.
  • An image metadata standard that embeds attribution at training-data and generation levels would create an unbroken chain from original artists to final AI-assisted work.
  • Restorable and aesthetic-guided style transfer techniques already point toward reversibility, which is what would make style attribution auditable.
  • Because living artists' styles are the most valuable and most vulnerable targets, the documentation proposal most directly protects them, not already-canonized artists.
  • The finite nature of human creative production—lifespan, physical capacity, experience—should be treated as the foundation of artistic value, not as a limitation to be overcome.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The proposed per-pixel attribution layer presupposes that an output image can be decomposed into additive contributions from individual training samples; this may hold for linear or attention-based models but is unverified for diffusion models, where generation is iterative and non-linear.
  • The same infrastructure could enable a measurable royalty market: if style contribution is quantified as percentages, payments and consent could be computed and enforced automatically, which the paper only gestures at.
  • A concrete next experiment would build a small controllable generative model, embed per-pixel attribution metadata, and test whether displaying provenance changes viewers' valuation of AI-assisted images.
  • If the argument is right, detection research is not merely incomplete but conceptually misguided, because knowing that an image is AI-generated does nothing to preserve the human value that documentation would capture.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 5 minor

Summary. Unlimited Editions: Documenting Human Style in AI Art Generation argues that HCI research on AI art has focused too narrowly on detection, authenticity, and generation, and that the field should instead develop systems to document the human origins and stylistic lineage of generated images. The paper grounds this argument in art-historical scholarship (Impressionism, commercial lithography, catalogues raisonnés) and in contemporary studies of AI art perception and human-AI collaboration. Section 4 sketches a technical framework: metadata embedding via EXIF/XMP, a JSON schema for stylistic provenance, and a speculative per-pixel attribution layer intended to show which parts of a trained model influenced each output pixel. The paper proposes new evaluation metrics centered on documentation fidelity rather than output quality. The central claim is that preserving and tracing finite human creative development is a more productive HCI agenda than perfecting reproduction or detecting fakery.

Significance. If this reframing is accepted, it would redirect HCI research from detection and generation toward documentation, provenance, and attribution—an area that arguably receives less attention than it deserves. The paper's historical synthesis and its engagement with recent empirical work on AI art provide a coherent foundation for this programmatic argument. It also offers concrete, albeit preliminary, artifacts such as the JSON metadata schema (Section 4.2), which could seed prototyping and discussion. The paper makes no machine-checked claims or empirical evaluations, and its contribution is conceptual rather than technical. The main strength is the clarity of its normative reframing; the main weakness, as detailed below, is the gap between the paper's own definition of style as creative struggle and the technical proposal's operationalization of style as visual features and training-data provenance.

major comments (3)
  1. [§4.2–4.3 vs. §3] The paper defines artistic style as "the resolution of creative struggle" and "unique ways of seeing" (Section 3), but the proposed documentation system in Sections 4.2 and 4.3 operationalizes style as visual features: color palette arrays, brush technique lists, and per-pixel attribution percentages showing "which percentage of each pixel comes from each piece of the trained model." These capture statistical influence from training images, not the human decisions, constraints, or biographical finiteness that the paper argues constitute style. Even the prompt-logging component in Section 4.1 records textual inputs, not the deeper creative development the authors emphasize. The proposed system would therefore document surface aesthetics and training-data provenance, but not the "human style" the paper claims HCI should preserve, leaving the central research agenda unsupported by the proposed technical direction.
  2. [§4.3] The claim that "it seems possible to therefore encode a per-channel per-pixel set of predictive information that allows users to see which percentage of each pixel comes from each piece of the trained model" is asserted without a technical argument, a concrete method, a citation to existing work, or a prototype. The analogy to AlphaGo does not establish feasibility to attribute output pixels back to specific training images in generative models. Because this per-pixel attribution is the core of the proposed technical intervention, the paper should either present a concrete approach, cite relevant interpretability or attribution methods, or explicitly label this as an open research challenge requiring substantial new work.
  3. [§3 (Bach faucet and scarcity)] The paper asserts that scarcity is "the current mechanism dictating whether or not an object is art" and that infinite generation devalues artistic output, but this economic premise is not empirically or theoretically defended. It does not engage with counterexamples such as conceptual art, digital art, open-source aesthetics, or performance art, where reproducibility or ephemerality do not obviously undermine value. Since this premise motivates the entire "Unlimited Editions" argument and the proposed documentation agenda, the authors should qualify the claim or provide supporting evidence and address potential objections.
minor comments (5)
  1. [§2.1] The phrase "avante-garde" should be spelled "avant-garde."
  2. [§4.1] The subsection title "Current Approaches and Their Limitations" promises a review of limitations, but the text discusses only EXIF2Vec at any length and does not systematically assess existing attribution, provenance, or metadata systems for AI-generated images.
  3. [§4.2] The JSON schema includes a "consent_status" boolean, but the paper does not explain how such consent would be obtained, maintained, or verified in practice; this warrants clarification if the schema is meant as a realistic proposal.
  4. [§4.3] The discussion of XMP claims it cannot store binary data but can store Base64; this is imprecise because XMP can embed Base64-encoded binary, and the proposed "per-channel per-pixel" layer is not described in enough detail to judge its storage or representation.
  5. [Introduction] The title "Unlimited Editions" is evocative but the term is never explicitly defined in the text; defining it early would help readers connect the title to the argument about infinite generation devaluing finite human production.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: this is a position and reframing paper with no fitted quantities, no derivations, and no load-bearing self-citations.

full rationale

The paper makes no formal derivation and contains no fitted parameters that are later renamed as predictions. Its central argument is historical and conceptual, drawing on external citations about Impressionism, art markets, and prior empirical studies of AI art perception, none of which are authored by the present authors. The technical proposal in Sections 4.2 and 4.3 is explicitly speculative, using language such as "it could maintain" and "presumably could store," so it does not claim to derive a documented output from a defined input. Although there is an internal tension between the paper's definition of style as "the resolution of creative struggle" and its proposed pixel-level attribution metadata, that is a coherence or feasibility concern, not a circular reduction: the paper never asserts that recording color palettes and per-pixel percentages is the same thing as documenting creative struggle, nor does it use that equation as a premise to conclude its own thesis. No self-citation chain, uniqueness import, or ansatz-smuggling citation appears. The paper is therefore self-contained as an argumentative essay, and no circularity step can be exhibited with the required textual reduction.

Assumptions & free parameters 1 free parameters · 5 assumptions · 3 invented entities

The paper rests on domain assumptions about the source of artistic value, the flattening of provenance in AI systems, and the technical feasibility of documenting style lineage. It introduces illustrative entities such as the provenance document and per-pixel attribution layers, but none have been implemented or validated. There are no fitted parameters in the argument, aside from the illustrative influence_weight.

free parameters (1)
  • influence_weight = 0.85 (example, not fitted)
    In the illustrative JSON in Section 4.2, a made-up numeric weight is assigned to Monet's influence. It is not derived from data, but serves as a placeholder for a planned metric.
assumptions (5)
  • domain assumption Artistic value is founded on the finite and unrepeatable nature of human creative output.
    Introduction and Section 5 argue that scarcity, not aesthetics, is the basis of art's value.
  • domain assumption AI systems flatten human creative choices into reproducible patterns without preserving provenance.
    Abstract and Section 2.3 characterize current generative systems as losing the human context of style.
  • domain assumption The historical analogy between 19th-century mechanical reproduction and modern AI generation is a valid basis for predicting value dynamics.
    Section 2.1 and 2.2 draw direct parallels between lithography/photography and AI art systems.
  • domain assumption Stylistic influence can be decomposed into documentable elements and attributed through metadata and visual analysis.
    Section 4 proposes attribution graphs, per-pixel layers, and provenance documents as feasible without proof.
  • domain assumption Documenting stylistic lineage would preserve or restore the cultural and market value of art.
    Section 4.5 argues that documentation fidelity should replace output quality as an evaluation metric.
invented entities (3)
  • Stylistic provenance document (JSON schema with influence_weight and style_elements)
    purpose: To record which artists, artworks, and style features influenced a generated image.
    Illustrated in Section 4.2 but not implemented; the influence_weight of 0.85 is an arbitrary placeholder.
  • Per-channel per-pixel attribution layer
    purpose: To visualize and quantify the contribution of each training input to each pixel of an AI-generated image.
    Proposed in Section 4.3 as a possibility; no algorithm or evidence of feasibility is given.
  • Attribution graphs from training data to model weights
    purpose: To show which artists' works influenced specific model capabilities and outputs.
    Proposed in Section 4.1 without a concrete method for tracing influence through model weights.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Unlimited Editions: Documenting Human Style in AI Art Generation." pith.science (2026). https://pith.science/paper/7YOQY4L5

@misc{pith2026250719497,
  author       = {Pith},
  title        = {Pith review of: Unlimited Editions: Documenting Human Style in AI Art Generation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/7YOQY4L5}},
  note         = {Machine review of arXiv:2507.19497}
}
read the original abstract

As AI art generation becomes increasingly sophisticated, HCI research has focused primarily on questions of detection, authenticity, and automation. This paper argues that such approaches fundamentally misunderstand how artistic value emerges from the concerns that drive human image production. Through examination of historical precedents, we demonstrate that artistic style is not only visual appearance but the resolution of creative struggle, as artists wrestle with influence and technical constraints to develop unique ways of seeing. Current AI systems flatten these human choices into reproducible patterns without preserving their provenance. We propose that HCI's role lies not only in perfecting visual output, but in developing means to document the origins and evolution of artistic style as it appears within generated visual traces. This reframing suggests new technical directions for HCI research in generative AI, focused on automatic documentation of stylistic lineage and creative choice rather than simple reproduction of aesthetic effects.

Figures

Figures reproduced from arXiv: 2507.19497 by the authors.

Figure 1
Figure 1. "Thomas Kinkade painter of light image Renoir people at lunch", Image Creator in Bing [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

70 extracted references · 57 canonical work pages

  1. [1]

    Lisa Aaron, Suzanne Abbate, Nicole M Allain, Bridget Almas, Brian Fallon, Diane Gavin, Chinara Gordon, Margarete Jadamec, Anthony Merlino, Lori Pierie, Gregory Solano, and Dacia Wolf. 2024. AI Bias Concerns. (2024), 5–

    work page 2024

  2. [2]

    Scott Alexander. 2024. How Did You Do On The AI Art Turing Test?Astral Codex Ten(2024)

    work page 2024

  3. [3]

    Artnet. n.d.. Price Database. https://www.artnet.com/price-database/ Online database

    work page

  4. [4]

    ArtTactic and Sotheby’s. 2013. The $1m+ Impressionist & Modern Market. https://www.sothebys.com/en/articles/the-1m-impressionist-modern-market

    work page 2013

  5. [5]

    Philip Ball. 2011. Music: How to create the perfect playlist.The Guardian (2011). https://www.theguardian.com/music/2011/jun/27/music-mathematics- fibonacci

    work page 2011

  6. [6]

    Rebecca Barham and Ann Graham. 2015. What is it Worth? A Guide to Art Valuation and Market Resources.Reference & User Services Quarterly54, 4 (2015), 21–25

    work page 2015

  7. [7]

    Rebecca Bedell. 2011. What Is Sentimental Art?American Art25, 3 (2011), 9–12. doi:10.1086/663949

    work page doi:10.1086/663949 2011

  8. [8]

    Lisa Bellaiche, Reza Shahi, Martin H Turpin, Adrian Ragnhildstveit, Sarah Sprock- ett, Nathaniel Barr, Andrew Christensen, and Paul Seli. 2023. Humans versus AI: whether and why we prefer human-created compared to AI-created artwork. Cognitive Research: Principles and Implications8, 1 (2023), 1–22

    work page 2023

Show all 70 references
  1. [9]

    doi:10.2307/jj.20522984.11

    doi

  2. [10]

    Walter Benjamin. 1935. The Work of Art in the Age of Mechanical Reproduction. Illuminations(1935)

    1935

  3. [11]

    1973.The Anxiety of Influence: A Theory of Poetry(2 ed.)

    Harold Bloom. 1973.The Anxiety of Influence: A Theory of Poetry(2 ed.). Oxford University Press

    1973

  4. [12]

    2000.The Art of Impressionism: Painting Technique and the Making of Modernity

    Anthea Callen. 2000.The Art of Impressionism: Painting Technique and the Making of Modernity. Yale University Press

    2000

  5. [13]

    Eva Cetinic and James She. 2022. Understanding and creating art with AI: Review and outlook.ACM Transactions on Multimedia Computing, Communications, and Applications18, 2 (2022), 1–22

    2022

  6. [14]

    Rebecca Chamberlain, Caitlin Mullin, Benoît Scheerlinck, and Johan Wagemans

  7. [15]

    Kate Compton. 2023. Declaring a new term: A Bach Faucet is a situation where a generative system makes an endless supply of some content. https://twitter. com/GalaxyKate/status/1583907942834716672 Tweet posted on Twitter/X

    2023

  8. [16]

    Michael Clapper. 2006. Thomas Kinkade’s Romantic Landscape.American Art 20, 2 (2006), 76–99. doi:10.1086/507501

    2006 doi

  9. [17]

    Blaise Cronin. 2012. Collaboration in Art and in Science: Approaches to Attri- bution, Authorship, and Acknowledgment.Information & Culture47, 1 (2012), 18–37

    2012

  10. [18]

    Kate Crawford and Trevor Paglen. 2021. Excavating AI: The politics of images in machine learning training sets.AI & Society36 (2021), 1105–1116

    2021

  11. [19]

    Shani Don-Yehiya, Leshem Choshen, and Omri Abend. 2023. Human Learning by Model Feedback: The Dynamics of Iterative Prompting with Midjourney. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing (EMNLP 2023). Association for Computational ...

    2023

  12. [20]

    Jia Deng, Wei Dong, Richard Socher, Li-Jia Li, Kai Li, and Li Fei-Fei. 2009. Im- ageNet: A large-scale hierarchical image database. In2009 IEEE Conference on Computer Vision and Pattern Recognition. IEEE, 248–255. doi:10.1109/CVPR.2009. 5206848

    2009 doi

  13. [21]

    Michele Elam. 2022. Signs Taken for Wonders: AI, Art & the Matter of Race. Daedalus151, 2 (2022), 198–217

    2022

  14. [22]

    Death- Effect

    Robert B Ekelund, Rand W Ressler, and John Keith Watson. 2000. The "Death- Effect" in art prices: A demand-side exploration.Journal of Cultural Economics 24, 4 (2000), 283–300

    2000

  15. [23]

    Jake Elwes. 2016. Machine Learning Porn. https://www.jakeelwes.com/project- MLPorn.html

    2016

  16. [24]

    Irene Eleta and Jennifer Golbeck. 2012. A study of multilingual social tagging of art images: cultural bridges and diversity. InProceedings of the ACM 2012 conference on Computer Supported Cooperative Work (CSCW ’12). ACM, 695–704. doi:10.1145/2145204.2145310

    2012

  17. [25]

    Caroline Goldstein. 2023. Artists Land a Win in Class Action Lawsuit Against A.I. Companies.Artnet News(2023). https://news.artnet.com/art-world/artists- vs-stability-ai-lawsuit-moves-ahead-2524849

    2023

  18. [26]

    Pamela Fletcher. 2012. On the Rise of the Commercial Art Gallery in London. In BRANCH: Britain, Representation and Nineteenth-Century History, Dino Franco Felluga (Ed.). Romantic Circles

    2012

  19. [27]

    Google Developers. 2023. Depthmap Metadata. https://developers.google.com/ Last updated 2023-06-23

    2023

  20. [28]

    Caroline Goldstein. 2023. Is A.I. Coming for the Drag Queens Next? A Deep Fake Cabaret at the V&A Exposes the Tech’s Limitations.Artnet News (2023). https://news.artnet.com/art-world/is-a-i-coming-for-the-drag-queens- next-a-deep-fake-cabaret-at-the-va-exposes-the-techs-limita...

    2023

  21. [29]

    Jesse Greenspan. 2013. Never Underestimate the Power of a Paint Tube.Smithso- nian Magazine(2013). https://www.smithsonianmag.com/arts-culture/never- underestimate-the-power-of-a-paint-tube-36637764/

    2013

  22. [30]

    Clement Greenberg. 1939. Avant-Garde and Kitsch.Partisan Review6, 5 (1939), 34–49

    1939

  23. [31]

    Jun Hitsuwari, Yoshiyuki Ueda, Wonjun Yun, and Michio Nomura. 2023. Does human-AI collaboration lead to more creative art? Aesthetic evaluation of human- made and AI-generated haiku poetry.Computers in Human Behavior139 (2023), 107502. doi:10.1016/j.chb.2022.107502

    2023

  24. [32]

    A. S. Hamrah. 2010. A Cottage for Sale: The High Price of Sentimentality.The Baffler2, 1 (2010), 54–64

    2010

  25. [33]

    Haiting Huang, Mathias Zinnen, Shumei Liu, Andreas Maier, and Vincent Christlein. 2024. Scene Classification on Fine Arts with Style Transfer. InProceed- ings of the 6th workshop on the analySis, Understanding and proMotion of heritAge Contents. 18–27. doi:10.1145/3689094.3689468

    2024

  26. [34]

    Ji Won Hong and Norah E Curran. 2019. Does it matter who created it? Under- standing the impact of disclosure and human-involvement in AI-generated art evaluation.Psychology of Aesthetics, Creativity, and the Arts13, 4 (2019), 359–369. doi:10.1037/aca0000250

    2019 doi

  27. [35]

    2015.Dreamscapes of modernity: Sociotech- nical imaginaries and the fabrication of power

    Sheila Jasanoff and Sang-Hyun Kim. 2015.Dreamscapes of modernity: Sociotech- nical imaginaries and the fabrication of power. University of Chicago Press

    2015

  28. [36]

    Waldemar Januszczak. 2015. Rodin. In P. Langan (Executive Producer), Fake or Fortune? BBC One

    2015

  29. [37]

    Christoph Kern, Frederic Gerdon, Ruben L Bach, Florian Keusch, and Frauke Kreuter. 2022. Humans versus machines: Who is perceived to decide fairer? Experimental evidence on attitudes toward automated decision-making.Patterns 3, 10 (2022), 100591. doi:10.1016/j.patter.2022.100591

    2022

  30. [38]

    Patrick Radden Keefe. 2023. How Larry Gagosian Reshaped the Art World.The New Yorker(2023). https://www.newyorker.com/magazine/2023/07/31/larry- gagosian-profile

    2023

  31. [39]

    Eileen Kinsella. 2024. Here Are the 10 Most Expensive Lots Sold at Auction in September 2024.Artnet News(2024). https://news.artnet.com/market/here-are- the-10-most-expensive-lots-sold-at-auction-in-september-2024-2560773

    2024

  32. [40]

    Jong Seok Kim, Minjeong Kim, and Tae Hyun Baek. 2024. Enhancing user expe- rience with a generative AI chatbot.International Journal of Human-Computer Interaction2024, 0 (2024), 1–13. doi:10.1080/10447318.2024.2303860

    2024

  33. [41]

    Nils Köbis and Luca D Mossink. 2021. Artificial intelligence versus Maya Angelou: Experimental evidence that people cannot differentiate AI-generated from human- written poetry.Computers in Human Behavior114 (2021), 106553. doi:10.1016/j. chb.2020.106553

    2021

  34. [42]

    Cornelius Krieghoff. 1850. Habitant Family with Horse and Sleigh

  35. [43]

    Rita Latikka, Johanna Bergdahl, Nina Savela, and Atte Oksanen. 2023. AI as an Artist? A Two-Wave Survey Study on Attitudes Toward Using Artificial Intelligence in Art.Poetics(2023), 101839

    2023

  36. [44]

    Moreno La Quatra, Giuseppe Gallipoli, and Luca Cagliero. 2024. Self-supervised Text Style Transfer Using Cycle-Consistent Adversarial Networks.ACM Transac- tions on Intelligent Systems and Technology15, 5 (2024), 1–38. doi:10.1145/3678179

    2024 doi

  37. [45]

    David Lawrence, David Pines, and Patricia J Barnett. 1984. MUSEUM TOL: THE METROPOLITAN MUSEUM OF ART AUTOMATES ACQUISITIONS.Art Documentation: Journal of the Art Libraries Society of North America3, 3 (1984), 91–92

    1984

  38. [46]

    Devorah Lauter. 2024. Impressionism at 150: Here Are 9 Shows Advancing the Historic Movement.Artnet News(2024). https://news.artnet.com/art-world/ impressionism-at-150-shows-2482318

    2024

  39. [47]

    Yingnan Ma, Chenqiu Zhao, Bingran Huang, Xudong Li, and Anup Basu. 2024. RAST: Restorable Arbitrary Style Transfer.ACM Transactions on Multimedia Computing, Communications and Applications20, 5 (2024), 1–21. doi:10.1145/ 3638770

    2024

  40. [48]

    Vivian Liu and Lydia B Chilton. 2022. Design guidelines for prompt engineer- ing text-to-image generative models. InCHI Conference on Human Factors in Computing Systems

    2022

  41. [49]

    Minneapolis Institute of Art. n.d.. The Lasting Impression of French Impressionism. Retrieved from https://new.artsmia.org/programs/teachers- and-students/teaching-the-arts/five-ideas/the-lasting-impression-of-french- impressionism

  42. [50]

    Peter C Marzio. 1971. Lithography as a Democratic Art: A Reappraisal.Leonardo 4, 1 (1971), 37–48. doi:10.2307/1572229 Unlimited Editions CHI EA ’25, April 26-May 1, 2025, Yokohama, Japan

    1971 doi

  43. [51]

    Jane Morris and Mason L Klein. 2019. The Do’s and Don’ts of Death.Sotheby’s (2019). https://www.sothebys.com/en/articles/the-dos-and-donts-of-death

    2019

  44. [52]

    Christian Morgner. 2014. The Evolution of the Art Fair.Historical Social Research 39, 3 (2014), 318–336

    2014

  45. [53]

    Jonas Oppenlaender. 2022. The creativity of text-to-image generation. InPro- ceedings of the 25th International Academic Mindtrek Conference. ACM, 192–202. doi:10.1145/3569219.3569456

    2022

  46. [54]

    Musée national Picasso-Paris. n.d.. Image rights. https://www.museepicassoparis. fr/en/image-rights

  47. [55]

    Aloys Prinz, Jan Piening, and Thomas Ehrmann. 2015. The success of art galleries: a dynamic model with competition and information effects.Journal of Cultural Economics39, 2 (2015), 153–176

    2015

  48. [56]

    Jonas Oppenlaender, Johanna Silvennoinen, Ville Paananen, and Aku Visuri. 2023. Perceptions and realities of text-to-image generation. InProceedings of the 26th International Academic Mindtrek Conference. 279–288

    2023

  49. [57]

    Molly Rose Sauter. 2022. Seeing Without Looking: Sidewalk Toronto is dead, but its legacy shows there are more ways to violate one’s privacy than collecting one’s image.Real Life Magazine(2022). https://reallifemag.com/seeing-without- looking/

    2022

  50. [58]

    Martin Ragot, Nicolas Martin, and Sébastien Cojean. 2020. AI-generated vs. Human Artworks. A Perception Bias Towards Artificial Intelligence?. InExtended Abstracts of the 2020 CHI Conference on Human Factors in Computing Systems (CHI EA ’20). ACM, 1–10. doi:10.1145/3334480.3382892

    2020

  51. [59]

    Scientific American. 1915. The Invention and Development of Photography.Sci- entific American112, 23 (1915), 530–566. https://www.jstor.org/stable/26022223

    1915

  52. [60]

    Tim Schneider. 2018. The First AI-Generated Portrait Ever Sold at Auction Shatters Expectations, Fetching $432,500—43 Times Its Estimate.Artnet News (2018). https://news.artnet.com/market/first-ever-artificial-intelligence-portrait- painting-sells-at-christies-1379902

    2018

  53. [61]

    John Spencer, J Matthew Huculak, Carrie Higginbotham, Alanna Callaway, and Eric Singhalampong. 2016. Impressionism (Painting). InRoutledge Encyclopedia of Modernism. Routledge. doi:10.4324/9781135000356-REMO28-1

    2016 doi

  54. [62]

    1977.On Photography

    Susan Sontag. 1977.On Photography. Farrar, Straus and Giroux

    1977

  55. [63]

    Bruce Sterling. 2019. The Work of Art in the Age of Artificial Intelligence. WIRED(2019). https://www.wired.com/beyond-the-beyond/2019/09/work-art- age-artificial-intelligence/

    2019

  56. [64]

    Fiona Staples. 2022. Instagram Post on AI Art. https://www.instagram.com/ fionastaples Instagram post

    2022

  57. [65]

    Robert Verhoogt. 2019. Art Reproduction and the Nation: National Perspectives in an International Art Market. InArt Crossing Borders: The Internationalisation of the Art Market in the Age of Nation States, 1750-1914, J Baetens and L Lyna (Eds.). Brill, 300–326

    2019

  58. [66]

    Antonio Torralba and Alexei A Efros. 2011. Unbiased look at dataset bias. In CVPR 2011. IEEE, 1521–1528. doi:10.1109/CVPR.2011.5995347

    2011

  59. [67]

    Ran Yi, Haokun Zhu, Teng Hu, Yu-Kun Lai, and Paul L Rosin. 2024. AesStyler: Aesthetic Guided Universal Style Transfer. InProceedings of the 32nd ACM Inter- national Conference on Multimedia. 9789–9798. doi:10.1145/3664647.3680784

    2024

  60. [68]

    Kai Xu, Feng Liu, Yi Mou, Yue Wu, Jing Zeng, and Mike S Schäfer. 2020. Using Machine Learning to Learn Machines: A Cross-Cultural Study of Users’ Responses to Machine-Generated Artworks.Journal of Broadcasting & Electronic Media64, 4 (2020), 578–598. doi:10.1080/08838151.2020.1835136

    2020

  61. [70]

    2020.AI Art: Machine Visions and Warped Dreams

    Joanna Zylinska. 2020.AI Art: Machine Visions and Warped Dreams. Open Humanities Press. Received 5 December 2024; accepted 30 January 2025

    2020

  62. [2018]

    Psychology of Aesthetics, Creativity, and the Arts12, 2 (2018), 177–192

    Putting the art in artificial: Aesthetic responses to computer-generated art. Psychology of Aesthetics, Creativity, and the Arts12, 2 (2018), 177–192

    2018

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.