Exploring Designers’ Perceptions and Practices of Collaborating with Generative AI as a Co-creative Agent in a Multi-stakeholder Design Process: Take the Domain of Avatar Design as an Example

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Design with multi-stakeholder participation is a particular branch of participatory design, in which stakeholders do not directly participate in the co-design stage with design professionals but mainly contribute by making requirements and providing feedback to designers.

“In the field of design research, design with multi-stakeholder participation is a particular branch of participatory design, in which these stakeholders wouldn't directly participate in the co-design stage together with the design professionals [59].They mainly contribute to the design process by making requirements and providing feedback to the designers [51].”

• cites [59] 59. Michael J Muller and Sarah Kuhn. 1993. Participatory design. Commun. ACM 36, 6 (1993), 24-28. sourceNot checked — no full text of the source
• cites [51] 51. Victoria McArthur. 2014. E is for Everyone: The Role of Stakeholders in Partici- patory Design and Game Studies. Journal ISSN 2368 (2014), 6103. sourceDisputed

  Split into 2 atoms — 0 confirmed, 1 partially supported, 1 disputed, 0 misattributed, 0 unverifiable. Expand the claim to see per-atom verdicts.

  Checked by a reasoning model — may be wrong.

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

In many real-world design scenarios, multi-stakeholder opinions are highly crucial to design outcomes, since stakeholders usually work as end-users who manage the deployment of the finished product.

“In many real-world design scenarios, multistakeholder opinions are highly crucial to the design outcomes, since they usually work in the role of end-users who manage the deployment of the finished product [51].”

• cites [51] 51. Victoria McArthur. 2014. E is for Everyone: The Role of Stakeholders in Partici- patory Design and Game Studies. Journal ISSN 2368 (2014), 6103. sourcePartly supported

  The Introduction says participatory design includes stakeholders so their needs are met in the design and deployment of the finished product. That supports the importance of stakeholder input, but it does not say stakeholders usually serve as end-users who manage deployment. (LLM, conf=0.87)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing).

Fine-tuning models of AI drawing algorithms, when provided with a larger training set, can better capture artistic styles.

“Additionally, fine-tuning models of AI drawing algorithms, when provided with a larger training set, can better capture artistic styles [65].”

• cites [65] 65. Nataniel Ruiz, Yuanzhen Li, Varun Jampani, Yael Pritch, Michael Rubinstein, and Kfir Aberman. [n. d.]. DreamBooth: Fine Tuning Text-to-Image Diffusion Models for Subject-Driven Generation. arXiv:2208.12242 [cs] http://arxiv.org/abs/2208. 12242 sourcePartly supported

  Section 4.4 (page unavailable) says the fine-tuned model can generate renditions of a subject in a famous painter’s style. Figure 20 (page unavailable) says more input images can improve reconstruction of a subject’s identity, but it does not link a larger training set to better capture of artistic styles. (LLM, conf=0.87)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing).

Discussions around generative AI have become an interdisciplinary issue deeply intertwining in the socio-cultural context of human society.

“However, discussions around generative AI could never be constrained within the subject matter of purely technical innovation; it has become an interdisciplinary issue deeply intertwining in the socio-cultural context of human society nowadays [2,12].”

• cites [2] 2. Giuseppe Amato, Malte Behrmann, Frédéric Bimbot, Baptiste Caramiaux, Fabrizio Falchi, Ander Garcia, Joost Geurts, Jaume Gibert, Guillaume Gravier, Hadmut Holken, et al. 2019. AI in the media and creative industries. arXiv preprint arXiv:1905.04175 (2019). sourcePartly supported

  The Challenges section calls for a transdisciplinary approach spanning art, technology and gaming, and section 3 says AI in media and creative industries raises societal and ethical challenges, not just technical ones. The paper also discusses generative approaches in its Creation section, but it does not establish the claim about discussions of generative AI across society generally; its stated scope is media and creative industries. (LLM, conf=0.82)

  Experimental — accuracy not yet measured
• cites [12] 12. Baptiste Caramiaux and Sarah Fdili Alaoui. [n. d.]. "Explorers of Unknown Planets": Practices and Politics of Artificial Intelligence in Visual Arts. 6 ([n. d.]), 1-24. Issue CSCW2. https://doi.org/10.1145/3555578 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing).

The contemporary art field is generally embracing the novel opportunities brought by computational experiments and algorithmic randomness within current generative AI systems.

“With its nature of being avant-garde to the limits, the contemporary art field is generally embracing the novel opportunities brought by the computational experiments and algorithmic randomness within the current generative AI systems [49,65].”

• cites [49] 49. Lev Manovich. [n. d.]. Defining AI Arts: Three Proposals. ([n. d.]).Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

• cites [65] 65. Nataniel Ruiz, Yuanzhen Li, Varun Jampani, Yael Pritch, Michael Rubinstein, and Kfir Aberman. [n. d.]. DreamBooth: Fine Tuning Text-to-Image Diffusion Models for Subject-Driven Generation. arXiv:2208.12242 [cs] http://arxiv.org/abs/2208. 12242 sourceCould not verify

  Section 4.4, “Applications” (page not provided), shows that DreamBooth can generate novel artistic renditions. The cited paper does not address whether the contemporary art field generally embraces generative AI, computational experimentation, or algorithmic randomness; searches for those ideas found no supporting passage. (LLM, conf=0.77)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing).

Wu et al. described the role of designers in China's e-market as an executant of tasks without much creative input.

“Wu et al. also described the role of designers in China's e-market as an executant of tasks without much creative input [73], resonating with our findings.”

• cites [73] 73. Xiuxiu Wu, Kin Wai Michael Siu, Jörn Bühring, and Caterina Villani. [n. d.]. The Relationship between Creative Self-Efficacy, Achievement Motivation, and Job Burnout among Designers in China's e-Market. 11, 11 ([n. d.]), 509. https: //doi.org/10.3390/socsci11110509 sourceSupported

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

AI image generators are built by scraping millions of photos from private databases or the open web, and are trained to detect patterns and correlations to create new images in a similar style.

“AI-images generators are built by scraping millions of photos from private databases or the open web.The tools are then trained to detect patterns and correlations in the images to create new ones in a similar style.The reason why AI systems like Midjourney can approximate such a high level of detail is that they have been trained on images, some of which are copyrighted works by real artists [2,3”

• cites [2] 2. Giuseppe Amato, Malte Behrmann, Frédéric Bimbot, Baptiste Caramiaux, Fabrizio Falchi, Ander Garcia, Joost Geurts, Jaume Gibert, Guillaume Gravier, Hadmut Holken, et al. 2019. AI in the media and creative industries. arXiv preprint arXiv:1905.04175 (2019). sourcePartly supported

  The Creation section describes generative models trained on images and models that learn artistic style and content to produce new works. It does not say that image generators scrape millions of photos from private databases or the open web, or that their training images include copyrighted artists’ works. The Diffusion and Consumption section discusses large training sets, but not those claimed sources of images. (LLM, conf=0.88)

  Experimental — accuracy not yet measured
• cites [3] 3. Nantheera Anantrasirichai and David Bull. [n. d.]. Artificial intelligence in the creative industries: a review. 55, 1 ([n. d.]), 589-656. https://doi.org/10.1007/ s10462-021-10039-7Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Probably not a substantial claim: may not pass S3 (load-bearing).

Prevailing AI drawing algorithms Stable Diffusion has already augmented the Low-Rank Adaptation of Large Language Models (LoRA) of DreamBooth model to finetune text-to-image results, including art style in the output images.

“Prevailing AI drawing algorithms Stable Diffusion has already augmented the Low-Rank Adaptation of Large Language Models (LoRA) [36] of DreamBooth model to finetune text-to-image results [65], including art style in the output images.”

• cites [36] 36. Edward J. Hu, Yelong Shen, Phillip Wallis, Zeyuan Allen-Zhu, Yuanzhi Li, Shean Wang, Lu Wang, and Weizhu Chen. [n. d.]. LoRA: Low-Rank Adaptation of Large Language Models. arXiv:2106.09685 [cs] http://arxiv.org/abs/2106.09685 sourceSupported
• cites [65] 65. Nataniel Ruiz, Yuanzhen Li, Varun Jampani, Yael Pritch, Michael Rubinstein, and Kfir Aberman. [n. d.]. DreamBooth: Fine Tuning Text-to-Image Diffusion Models for Subject-Driven Generation. arXiv:2208.12242 [cs] http://arxiv.org/abs/2208. 12242 sourceSupported

Probably not a substantial claim: may not pass S3 (load-bearing).

Prior HCI literature has investigated design methodologies or strategies facilitating HAI interaction and dynamics in co-creative practices.

“Previous HCI literature has also investigated in design methodologies or strategies facilitating HAI interaction and dynamics in co-creative practices [32,33].”

• cites [32] 32. Chao Guo, Yong Dou, Tianxiang Bai, Xingyuan Dai, Chunfa Wang, and Yi Wen. [n. d.]. ArtVerse: A Paradigm for Parallel Human-Machine Collaborative Painting Creation in Metaverses. 53, 4 ([n. d.]), 2200-2208. https://doi.org/10.1109/TSMC. 2022.3230406 sourceNot checked — no full text of the source
• cites [33] 33. Matthew Guzdial, Nicholas Liao, and Mark Riedl. 2018. Co-creative level design via machine learning. arXiv preprint arXiv:1809.09420 (2018). sourceSupported

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

In the field of creative production, technological capabilities have constrained creative productivity and informed the ways in which creative practices are perceived and consumed by social stakeholders.

“In the field of creative production, it is an evolving relationship that technological capabilities have constrained creative productivity and informed the ways in which these creative practices are perceived and consumed by groups of social stakeholders who are actively participating as diverse agencies in this socio-technical innovation landscape [67].”

• cites [67] 67. James She (Ed.). [n. d.]. Introduction to the Special Issue on Big Data, Machine Learning, and AI Technologies for Art and Design. 15, 2 ([n. d.]), 1-3. https: //doi.org/10.1145/3338002 sourcePartly supported

  The introduction says technological capabilities have constrained creative products and shaped how audiences perceive and interpret art and design. It does not say they constrained creative productivity, nor does it describe consumption by social stakeholders participating as diverse agencies. (LLM, conf=0.84)

  Experimental — accuracy not yet measured

Probably not a substantial claim: may not pass S3 (load-bearing).

Prior works have revealed the importance of having a distinctive art style for modern artists like painters.

“Prior works have revealed the importance of having a distinctive art style for modern artists like painters [72].”

• cites [72] 72. Triyono Widodo. [n. d.]. Exploration and Experimentation in Modern Art: Paint- ing in a Personal Style. ([n. d.]). https://doi.org/10.18502/kss.v4i12.7622 sourceSupported

May combine two claims.

Audiences often subjectively did not accept generative AI to produce the manuscripts they requested, but they were often unable to recognize them.

“Often, the audiences still subjectively did not accept generative AI to produce the manuscripts they requested, but they were often unable to recognize them”

Probably not a substantial claim: may not pass S3 (load-bearing). The quote only partly states this claim.

Training models for a specific designer's personal style, called washing, is considered by designers to be a completely wrong way to use generative AI tools.

“there is also the personal behavior of training models for a specific designer's personal style, called”

Probably not a substantial claim: may not pass S3 (load-bearing).

The designers thought that this was due to AI's fatal lack of logical linkage in the design workflow which human designers always had.

“The designers thought that this was due to AI's fatal lack of logical linkage in the design workflow which human designers always had.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Five of the designers pointed out that it was more unacceptable to train models on the style of a specific designer than to train models on a wide range of designers' works.

“Five of the designers pointed out that it was more unacceptable to train models on the style of a specific designer than to train models on a wide range of designers' works.”

May combine two claims.

Audiences treat artworks produced with the assistance of AI with biases, often associating it with unethical practices in the training stage and diminished artistic value.

“Our findings suggest that audiences treat artworks produced with the assistance of AI with biases, often associating it with unethical practices in the training stage and diminished artistic value.”

• AssumptionRests on an assumption: Designer accounts can adequately portray the perspectives of capital investors and audiences.

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims. The quote only partly states this claim.

According to P16, the designers thought this kind of thinking was a kind of subconscious refraction: if you could see the traces of learning, it was unacceptable; if you couldn't see it, you defaulted to not caring about it.

“According to P16, the designers thought this kind of thinking was a kind of subconscious refraction,”

Probably not a substantial claim: may not pass S3 (load-bearing).

The designers usually gather inspiration from various sources and multiple media and use them as references.

“The designers usually gather inspiration from various sources and multiple media and use them as references.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

P11 said that if asked to create something with a traditional Chinese style, AI often produces a mix of different styles like Japanese style or Korean style, which can potentially lead to trouble if used inappropriately.

“if I ask AI to create something with a traditional Chinese style, it often produces a mix of different styles like Japanese style or Korean style, which can potentially lead to trouble if used inappropriately.(P11)”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

The emergence of generative AI has made the positions of designers in the industry become fewer, and those who are willing to pursue a career as original designers have become scarcer.

“the emergence of generative AI has made the positions of designers in the industry become fewer, and those who are willing to pursue a career as original designers have become scarcer.”

Classified as probably not a substantial claim. May combine two claims.

Designers hope that AI can free them from tedious work and give them inspiration in their current workflow so that they can put more energy into creative and innovative parts.

“However, designers still hope that AI can free them from tedious work and give them inspiration in their current workflow so that they can put more energy into creative and innovative parts.”

Probably not a substantial claim: may not pass S3 (load-bearing).

P14 admitted that if there is such a tool as generative AI that can significantly improve the efficiency of design while with more respect to the labor of the designers themselves, designers are more than willing to use it.

“As P14 admitted that if there is such a tool as generative AI that can significantly improve the efficiency of design, while with more respect to the labor of the designers themselves, designers are more than willing to use it.”

May combine two claims.

Audiences subjectively did not accept generative AI to produce the manuscripts they requested, but they were often unable to recognize them with the naked eye.

“the audiences still subjectively did not accept generative AI to produce the manuscripts they requested, but they were often unable to recognize them with the naked eye.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Capital investors considered the targeted use of generative AI to be limited to only certain links of the general design process.

“the capital investors were also considering the targeted use of them to be limited within only certain links of the general design process.”

Probably not a substantial claim: may not pass S1 (asserted), S3 (load-bearing). May combine two claims.

Some designers argue that the current aesthetic of mainstream generative AI is one of excessive sophistication, which may stimulate human aesthetics in the future toward the pursuit of handmade texture and flawed art.

“there are also designers who argue from the opposite direction, that the current aesthetic of mainstream generative AI is one of”

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based). May combine two claims.

The high efficiency and low cost of generative AI had attracted a large amount of capital to join and was irreversible.

“Although the high efficiency and low cost of generative AI had attracted a large amount of capital to join and was irreversible.”

Probably not a substantial claim: may not pass S3 (load-bearing).

P5 stressed that the emergence of generative AI has suppressed the growth of young designers who were new to the design industry.

“P5 stressed that the emergence of generative AI has also suppressed the growth of young designers who were new to the design industry”

The quote only partly states this claim.

Designers tended to let generative AI take some work of the less important avatar design, which ought not to be the main characters in the game, to avoid affecting audience perceptions.

“they tended to let generative AI take some work of the less important avatar design, which ought not to be the main characters in the game.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

According to seven interviewees, from the industry's point of view, generative AI at the current stage is able to fulfill the needs of most low-end manuscripts of design, successfully replacing some bottom-level designers, which puts them out of business and creates a situation of supply exceeding demand in the whole market.

“according to seven of the interviewees, from the industry's point of view, generative AI on the current stage is able to fulfill the needs of most low-end manuscripts of design, successfully replacing some of the bottom-level designers, which really puts them out of business and creates a situation of supply exceeding demand in the whole market.”

Probably not a substantial claim: may not pass S1 (asserted), S3 (load-bearing).

P21 stated that today the grand influence made by capital on the promotion of certain popular aesthetic standards and creative landscape could not be ignored.

“As P21 stated faithfully, today the grand influence made by capital on the promotion of certain popular aesthetic standards and creative landscape could not be ignored anyhow.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Some designers (n=5) believed that if the generative AI companies were willing to pay for the training rights of the designer's original works, the two sides could probably reach an agreement on how to deal with generative AI ethics one day.

“Some designers (n=5) believed that if the generative AI companies were willing to pay for the”

Probably not a substantial claim: may not pass S1 (asserted), S3 (load-bearing). May combine two claims.

They argued that the process should be reversed, and that it should be the case that the AI company paid the painter who provided the original artwork, and that the generated images were paid for to be used or belonged to all of humanity according to almost all the designers' common wishes.

“They argued that the process should be reversed, and that it should be the case that the AI company paid the painter who provided the original artwork, and that the generated images were paid for to be used or belonged to all of humanity according to almost all the designers' common wishes.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

According to the majority of designers (n=19), using human creators' published works without permission is very disrespectful to the rights and interests of the creators themselves, and that is why there is a wave of opposition to generative AI on the Internet, which can be regarded as emotional catharsis and backlash responding to the violent exploitation of creativity done by huge AI companies.

“According to the majority of designers (n=19), this is very disrespectful to the rights and interests of the creators themselves, and that's why there is a wave of opposition to generative AI on the Internet for long, which can be regarded as a kind of emotional catharsis and backlash responding to the violent exploitation of creativity done by those huge AI companies.”

• AssumptionRests on an assumption: Most generative AI companies and related departments directly use human creators' published works as resources in their databases without permission.

Probably not a substantial claim: may not pass S4 (based).

Not all designers possess a vast collection of artworks for training.

“Yet, not all designers possess a vast collection of artworks for training.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Designers note that audiences might not always differentiate between AI-generated and human-made art.

“While designers often note, consistent with previous findings [62], that audiences might not always differentiate between AI-generated and human-made art, they remain apprehensive about public knowledge of such collaborations for fear of damaging their reputation.”

• cites [62] 62. Martin Ragot, Nicolas Martin, and Salomé Cojean. [n. d.]. AI-generated vs. Human Artworks. A Perception Bias Towards Artificial Intelligence?. In Extended Abstracts of the 2020 CHI Conference on Human Factors in Computing Systems (Honolulu HI USA, 2020-04-25). ACM, 1-10. https://doi.org/10.1145/3334480. 3382892 sourceNot checked — no full text of the source

May combine two claims.

Some designers were subjectively willing to accept the convenience brought by generative AI, but considering objective market acceptance, they were still unwilling to disclose that they were using generative AI.

“Some designers were subjectively willing to accept the convenience brought by generative AI, but considering the objective market acceptance, they still weren't willing to”

Probably not a substantial claim: may not pass S3 (load-bearing).

The logical flow of human designers usually requires years of training and practice.

“the logical flow of human designers usually requires years of training and practice.”

May combine two claims.

Current AI as a co-creative agent fails to align with the logic flow of human designers, which leads to obstacles in their collaborations.

“Our study compares the difference between human designers' and AI's workflow in avatar design and reveals that current AI as a co-creative agent fails to align with the logic flow of human designers, which leads to obstacles in their collaborations.”

Probably not a substantial claim: may not pass S4 (based). May combine two claims.

Capital investors are the measurement of the quality of artworks, so designers must follow their requests in the creation process instead of for purely artistic purposes.

“On the other hand, capital investors are the measurement of the quality of artworks, thus designers must follow their requests in the creation process instead of for purely artistic purposes.”

• AssumptionRests on an assumption: Designer accounts can adequately portray the perspectives of capital investors and audiences.

Probably not a substantial claim: may not pass S3 (load-bearing).

Many game companies are starting to recruit AIGC designers who are aimed to get prepared to collaborate with AI fluently from the beginning of their jobs.

“there are a lot of game companies starting to recruit AIGC (Artificial Intelligence Generated Content) designers, who should be aimed to get prepared to collaborate with AI fluently since the beginning of their jobs.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Some designers (n=7) pointed out that due to the low-cost attribute of generative AI systems, it is especially favored by most capital investors today.

“some designers (n=7) further pointed out that due to the low-cost attribute of generative AI systems, it is especially favored by most capital investors today.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

A number of interviewees (n=10) had concerns that if the power of capital makes the commercial style of generative AI widely used in various design domains and even everywhere in daily life, then audiences will soon get used to it.

“a number of our interviewees (n=10) had similar concerns as follows.If the power of capital makes the commercial style of generative AI widely used in various design domains and even everywhere in our daily life, then the audiences will soon get used to it.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Designers believe the composition of generated images can be the key to embedding emotional expressiveness in avatar design artworks.

“The composition of the generated images, which the designers believe can be the key to embedding emotional expressiveness in the avatar design artworks, should also be an available input dimension.”

May combine two claims.

Most of the designers (n=15) felt that being learned by AI was a very annoying and unacceptable thing because the designers believed that forming their own style required years of aesthetic accumulation and practice on paper.

“Most of the designers (n=15) felt that being learned by AI was a very annoying and unacceptable thing because the designers believed that”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Unlike human designers, the existing generative AI agents are not capable of utilizing sources of multiple forms as reference; rather, they can only integrate visual information, which might lose meaning without the knowledge of cultural contexts.

“Unlike human designers, the existing generative AI agents are not capable of utilizing sources of multiple forms as reference.Rather, it can only integrate visual information, which might lose meaning without the knowledge of cultural contexts.”

May combine two claims.

There are two main groups of designers approaching the ethical concerns of generative AI systems: those who believed there were ethical concerns to be reconsidered and coped with (n=12), and those who did not care at all (n=9).

“currently there are two main groups of designers who were approaching the ethical concerns of generative AI systems from two varied directionsthose who believed there were ethical concerns to be reconsidered and coped with (n=12), and those who did not care at all (n=9).”

Probably not a substantial claim: may not pass S3 (load-bearing).

Several designers (n=8) suggested that paying for membership in generative AI models was equivalent to users paying for the right to use these AI-generated images commercially.

“several designers (n=8) suggested that paying for membership in generative AI models was equivalent to users paying for the right to use these AI-generated images commercially.”

Probably not a substantial claim: may not pass S3 (load-bearing).

P13 pointed out that even if there were moral problems, the creator community or the masses couldn't fight against the capital investors due to the extreme power disparity.

“As P13 pointed out that, even if there were moral problems, the creator community or the masses couldn't fight against the capital investors due to the extreme power disparity.”

May combine two claims.

Designers did not want to play with the randomness of AI drawing because it hindered their execution of investors' tasks, further reducing their freedom to collaborate with AI.

“Hence, designers didn't want to play with the randomness of AI drawing which hindered their execution of investors' tasks, further reducing the freedom for them to collaborate with AI.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Human designers' designs not only fit the requirements of the client but also fit the cultural expectations and are explainable to the audiences.

“their designs not only fit the requirements of the client but also fit the cultural expectations and are explainable to the audiences.”

May combine two claims.

According to P2 and P14, most audiences' reaction to generative AI works today can be referred to modern people's reaction to the emergence of novel machinery tools during previous industrial revolutions, which can be stated as a fear of being outperformed by machines.

“According to both P2 and P14, how most of their audiences react to generative AI works today could be referred to modern people's reaction to the emergence of novel machinery tools during previous two rounds of industrial revolutions, which could be stated as a fear to be outperformed by machines.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

P6 and P17 raised that there are even logical errors in AI works, such as improper simulation of human muscle structure, which are poisonous to developing a rigorous approach towards avatar design.

“P6 and P17 were concerned to raise that there are even”

Probably not a substantial claim: may not pass S3 (load-bearing).

Avatar design is a complex and time-consuming process.

“However, the creation of an avatar is still recognized to be a complex and time-consuming process by a great number of avatar designers today [21].”

• cites [21] 21. Nicolas Ducheneaut, Ming-Hui Wen, Nicholas Yee, and Greg Wadley. 2009. Body and mind: a study of avatar personalization in three virtual worlds. In Proceedings of the SIGCHI conference on human factors in computing systems. 1151-1160. sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

The interview guide was reviewed and confirmed with potential participants beforehand, after gaining consent forms with their signatures.

“The interview guide was reviewed and confirmed with our potential participants beforehand, after gaining the consent forms with their signatures.”

Classified as probably not a substantial claim.

Three co-authors independently coded the interview transcripts in an inductive manner.

“Three co-authors independently coded the transcripts in an inductive manner.”

Probably not a substantial claim: may not pass S3 (load-bearing).

To resolve differences in coding, the authors convened and engaged in discussions until consensus was reached.

“To resolve any differences in coding, the authors convened and engaged in discussions until consensus was reached.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

UX designers are not trained to integrate ML as material into their design and the collaboration between designers and engineers is usually not design-driven, but rather technology-driven.

“They show that UX designers are not trained to integrate ML as material into their design and that the collaboration between designers and engineers is usually not designdriven, but rather technology-driven [14].”

• cites [14] 14. Eugene Ch'ng (Ed.). [n. d.]. Art by Computing Machinery: Is Machine Art Ac- ceptable in the Artworld? 15, 2 ([n. d.]), 1-17. https://doi.org/10.1145/3326338 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based). May combine two claims.

Generative AI tools are proven to be helpful in solving complicated tasks in specific design domains, thus enhancing the working efficiency of the industry.

“While in various design domains, generative AI (tools/features) are proven to be helpful in solving complicated tasks in specific design domains, thus enhancing the working efficiency of the industry.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

There are two main challenges that augment design complexities in avatar design: the trending of highly customizable avatars and the design pattern with multi-stakeholder participation.

“Drawing experience from previous studies, there are two main challenges suggested by researchers that are always augmenting the design complexities in this certain field -the trending of highly customizable avatars and the design pattern with multi-stakeholder participation.”

Probably not a substantial claim: may not pass S3 (load-bearing).

The study was approved by Duke Kunshan University's Institutional Review Board (IRB).

“This study has been approved by Duke Kunshan University's Institutional Review Board (IRB).”

Probably not a substantial claim: may not pass S3 (load-bearing).

There are a growing number of design research projects that use AI as a resource to critique and speculate on the use of AI technologies in people's daily lives.

“There are also a growing number of design research projects that use AI as a resource to critique and speculate on the use of AI technologies in people's daily lives [15].”

• cites [15] 15. Pei-Sze Chow. [n. d.]. Ghost in the (Hollywood) machine: Emergent applications of artificial intelligence in the film industry. 9 ([n. d.]).Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing).

It is even more difficult to improve the overall aesthetic evolution of human beings.

“In turn, it is even more difficult to improve the overall aesthetic evolution of human beings”

May combine two claims.

Using AI as a co-creative agent may put designers at risk of losing their own signatures in their artworks and may raise moral challenges for style imitation.

“Thus, using AI as a co-creative agent may put the designers at risk of losing their own signatures in their artworks, and even undertake moral challenges for style imitation.”

May combine two claims.

Current AI has a low capability of creating new art styles: it either generates images of random art styles or imitates other popular art styles.

“However, current AI has a low capability of creating new art styles: it either generates images of random art styles or imitates other popular art styles [14].”

• cites [14] 14. Eugene Ch'ng (Ed.). [n. d.]. Art by Computing Machinery: Is Machine Art Ac- ceptable in the Artworld? 15, 2 ([n. d.]), 1-17. https://doi.org/10.1145/3326338 sourceNot checked — no full text of the source

Classified as probably not a substantial claim. May combine two claims.

The identification of challenges in HAI collaboration is anchored in designer narratives, adopting a qualitative interview approach.

“Adopting a qualitative interview approach, our identification of challenges in HAI collaboration is anchored in designer narratives.”

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing). May combine two claims. The quote only partly states this claim.

The concept of style should be considered an important part of designers' work and an intellectual asset.

“the concept of”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Some in HCI research embrace new creative tools to create AI Art, but others are skeptical of the hype from different perspectives, including disturbance of social orders, legality of style appropriation, and potential copyright infringement.

“Specifically in the field of HCI research, some embrace these new creative tools to create various AI Art, but others are skeptical of the hype from different perspectives [17], including the disturbance of social orders [44], the legality of style appropriation [38], and potential copyright infringement [18].”

• cites [17] 17. Kate Crawford. 2021. The atlas of AI: Power, politics, and the planetary costs of artificial intelligence. Yale University Press. sourceNot checked — no full text of the source
• cites [44] 44. Octavio Kulesz. 2018. Culture, platforms and machines: the impact of artificial intelligence on the diversity of cultural expressions. Intergovernmental committee for the protection and promotion of the diversity of cultural expressions 12 (2018).Not checked yet

  Not checked in the free run.

• cites [38] 38. Michael Kearns and Aaron Roth. 2019. The ethical algorithm: The science of socially aware algorithm design. Oxford University Press. sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing).

Earlier studies indicate that audiences generally evaluate AI-created art on par with human-made art.

“This view contrasts with earlier studies indicating that audiences generally evaluate AI-created art on par with human-made art [62].”

• cites [62] 62. Martin Ragot, Nicolas Martin, and Salomé Cojean. [n. d.]. AI-generated vs. Human Artworks. A Perception Bias Towards Artificial Intelligence?. In Extended Abstracts of the 2020 CHI Conference on Human Factors in Computing Systems (Honolulu HI USA, 2020-04-25). ACM, 1-10. https://doi.org/10.1145/3334480. 3382892 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Prior works had revealed how culture factors might influence the understanding of visual language and users' trust in avatar design, but their influence on the design process remained unclear.

“Although prior works had revealed how culture factors might influence the understanding of visual language [53] and users' trust in avatar design [1,53], their influence on design process remained unclear.”

• cites [53] 53. Melissa McMullen. 2016. Intercultural design competence: A guide for graphic designers working across cultural boundaries. The International Journal of Visual Design 10, 3 (2016), 19-30. sourceNot checked — no full text of the source
• cites [1] 1. Hussain M. Aljaroodi, Marc T. P. Adam, Timm Teubner, and Raymond Chiong. 2023. Understanding the Importance of Cultural Appropriateness for User Inter- face Design: An Avatar Study. ACM Trans. Comput.-Hum. Interact. 29, 6, Article 52 (jan 2023), 27 pages. https://doi.org/10.1145/3517138 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing). May combine two claims.

AIs learning and imitating styles in a short period of time is dissolving the value of human labor, which in contemporary society is often equated with the value of existence, thus making many designers feel threatened professionally and many human beings feel confused about their existence.

“AIs learning and imitating styles in a short period of time is dissolving the value of human labor, which in contemporary society is often equated with the value of existence, thus making many designers feel threatened professionally, and many human beings feel confused about their existence.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

The problem of manifesting division rather than cooperation between the creative powers of human and AI is accentuated within the realm of creative industries in modern society, with additional economic significance attached.

“Notably, the problem of manifesting division rather than cooperation between the creative powers of human and AI mentioned above is even more accentuated within the realm of creative industries in the modern society, with additional economic significance attached [27,61].”

• cites [27] 27. Philip Galanter et al. 2020. Towards ethical relationships with machines that make art. Artnodes 26 (2020), 1-9. sourceNot checked — no full text of the source
• cites [61] 61. Matteo Pasquinelli. 2020. How a Machine Learns and Fails-A Grammar of Error for Artificial Intelligence. Spheres.Not checked yet

  Not checked in the free run.

Classified as probably not a substantial claim. May combine two claims.

The study primarily engaged with designers, portraying other stakeholder perspectives based on designer accounts.

“We have primarily engaged with designers, portraying other stakeholder perspectives based on designer accounts.”

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing). May combine two claims.

Encouraging AIGC artists to develop their unique artistic style with AI algorithms can help make AI-generated artworks a new genre of art and regain their art reputations.

“This can help them make AI-generated artworks a new genre of art as mentioned [12], and thus regain their art reputations.”

• cites [12] 12. Baptiste Caramiaux and Sarah Fdili Alaoui. [n. d.]. "Explorers of Unknown Planets": Practices and Politics of Artificial Intelligence in Visual Arts. 6 ([n. d.]), 1-24. Issue CSCW2. https://doi.org/10.1145/3555578 sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

Although the study is centered specifically on 2D static cartoon avatar design, the findings might still offer general insights regarding designer interactions with AI across broader domains.

“While our study is centered specifically on 2D static cartoon avatar design, we posit that our findings might still offer general insights regarding designer interactions with AI across broader domains.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Artists who have shared their work online may have unknowingly trained their algorithmic rivals for the creative battle.

“As a result, artists who have shared their work online may have unknowingly trained their algorithmic rivals for the creative battle [76].”

• cites [76] 76. Martin Zeilinger. 2021. Tactical entanglements: AI art, creative agency, and the limits of intellectual property. meson press.Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing).

The unauthorized learning of styles can be seen as a form of theft of intellectual property.

“Moreover, the unauthorized learning of styles can be seen as a form of theft of intellectual property.”

Probably not a substantial claim: may not pass S1 (asserted), S3 (load-bearing). May combine two claims.

Designers will no longer have the motivation to create higher-order aesthetics, and eventually the overall aesthetics of human beings will be degraded.

“designers will thus no longer have the motivation to create higher-order aesthetics, and eventually the overall aesthetics of human beings will be degraded from now on.”

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing).

Concerns about AI emulating existing artistic styles highlight a fundamental challenge in art and creativity.

“Concerns about AI emulating existing artistic styles highlight a fundamental challenge in the world of art and creativity.”

Probably not a substantial claim: may not pass S3 (load-bearing).

The creative industries have witnessed the advent of generative AI programs of various kinds due to the rapid development of multilayered machine learning architectures.

“In recent times, due to the rapid development of multilayered machine learning (ML) architectures such as Deep Learning (DL) and its increasing application in the field of artificial intelligence (AI) technology [52], the creative industries of modern human society have borne witness to the advent of generative AI programs of various kinds.”

• cites [52] 52. Jon McCormack, Toby Gifford, and Patrick Hutchings. 2019. Autonomy, au- thenticity, authorship and intention in computer generated art. In International conference on computational intelligence in music, sound, art and design (part of EvoStar). Springer, 35-50. sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing).

The availability of tools for training, keeping, and using personal LoRA models on local machines can decrease the concern of artists.

“This can decrease the concern of the artists.”

Probably not a substantial claim: may not pass S3 (load-bearing).

The recently released DALL•E 3 has integrated image generative AI with ChatGPT, a Large-Language Model (LLM).

“The recently released DALL•E 3 9 has integrated image generative AI with ChatGPT, a Large-Language Model (LLM).”

• cites [9] 9. Emanuela Bran, Cosima Rughiniş, Gheorghe Nadoleanu, and Michael G Flaherty. 2023. The Emerging Social Status of Generative AI: Vocabularies of AI Com- petence in Public Discourse. In 2023 24th International Conference on Control Systems and Computer Science (CSCS). IEEE, 391-398. sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

These detailed mistakes might not have been easily noticeable by people without art or design expertise but were detrimental to the final design outcomes.

“These detailed mistakes might not have been easily noticeable by people without art or design expertise but were detrimental to the final design outcomes.”

Probably not a substantial claim: may not pass S2 (checkable), S4 (based). May combine two claims.

The importance of having a personal art style is critical to contemporary human designers as it differentiates them from AI, which can hardly produce unique and impressive styles.

“The importance of having a personal art style is critical to contemporary human designers as it differentiates them from AI which can hardly produce unique and impressive styles.”

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing). May combine two claims.

Generative AI will challenge or update the human definition of arts or aesthetics.

“Therefore, generative AI will challenge or update the human definition of arts or aesthetics.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

This productive strategy would not particularly affect audience perceptions during play and thus was not likely to be exposed.

“This productive strategy would not particularly affect the audience's perceptions during their play and thus was not likely to be exposed.”

Probably not a substantial claim: may not pass S3 (load-bearing).

The force of capital tends to widely affect the general public, further shaping people's outlook on aesthetics.

“The force of capital tends to widely affect people -the general public, further shaping people's outlook on aesthetics.”

Probably not a substantial claim: may not pass S2 (checkable), S4 (based). May combine two claims.

The personal art style is not only a signature of the designer, but a property that can be more valuable than his/her artworks.

“The personal art style is not only a signature of the designer, but a property that can be more valuable than his/her artworks.”

Probably not a substantial claim: may not pass S3 (load-bearing).

The concept of creation has changed from manual labor to intellectual labor, which had long been the sole right of human beings.

“it's basically the concept of creation had now changed from manual labor to intellectual labor, which had long been the sole right of human beings.”

Classified as probably not a substantial claim.

The autonomy of generative AI agents is consistently growing and intervening in the design process.

“As the autonomy of such generative AI agents is consistently growing and intervening the design process, more viewpoints from diverse social stakeholders getting involved in the design process should be reconsidered and reinterpreted by the novel HAI creative community [55].”

• cites [55] 55. Eitan Mendelowitz et al. 2020. Intelligent Environments and Public Art. Artnodes 26 (2020), 1-9. sourceNot checked — no full text of the source

May combine two claims.

DALL•E 3 can relieve users from prompt engineering by handling the brainstorming and refinement of prompts to ChatGPT, and generate images that stick to user descriptions.

“It can relieve users from prompt engineering by handling the brainstorming and refinement of prompts to ChatGPT, and generate images that stick to user descriptions.”

• cites [9] 9. Emanuela Bran, Cosima Rughiniş, Gheorghe Nadoleanu, and Michael G Flaherty. 2023. The Emerging Social Status of Generative AI: Vocabularies of AI Com- petence in Public Discourse. In 2023 24th International Conference on Control Systems and Computer Science (CSCS). IEEE, 391-398. sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing).

The main body of involved designers was from the Chinese avatar design industry.

“The main body of involved designers was from Chinese avatar design industry.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Generative AI lacks a logical workflow to support its work.

“This is because generative AI lacks a logical workflow to support its work.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Generative AI systems are considered a pioneering attempt to boost a novel form of creativity that diverges from established human paths.

“With the capability of performing high-level creative tasks in diverse domains, for instance, visual arts [4], music composition and performance [23,26], creative writing [47], these generative systems are considered a pioneering attempt to boost a novel form of creativity that is diverging from the established path human beings have carved out and then followed each day [20], if examined through”

• cites [20] 20. Marcus Peter Francis Du Sautoy. 2019. The Creativity Code: How AI is learning to write, paint and think. (No Title) (2019). sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing).

The study has not moved to the phase of practical implementation and testing of the outlined design implications.

“While we have outlined potential design implications based on our findings, we have not moved to the phase of practical implementation and testing.”

Probably not a substantial claim: may not pass S3 (load-bearing).

The idea of designers who didn't think or care could be concluded as an attitude of helplessness to the current scenarios.

“The idea of designers who didn't think or care could be concluded as an attitude of helplessness to the current scenarios.”

Probably not a substantial claim: may not pass S3 (load-bearing).

There are available tools for training, keeping, and using personal LoRA models on local machines on Hugging Face.

“There are also available tools for training, keeping, and using personal LoRA models on local machines on Hugging Face 8 .”

• cites [8] 8. Margaret A Boden. 2014. Creativity and artificial intelligence: A contradiction in terms. The philosophy of creativity: New essays (2014), 224-46. sourceNot checked — no full text of the source