Humans versus AI: whether and why we prefer human-created compared to AI-created artwork

“Linear mixed-effects models using maximum likelihood estimation were estimated in R (version 4.2.0; R Core Team, 2022) using the {lme4} package (version 1.1.33; Bates et al., 2015).”

“…cascade of evaluations, implicating more-complex and elaborative processes in judgements of art that seek to understand meaning and communication through the piece (Berlyne, 1960, 1971; Chatterjee & Vartanian, 2016; Cupchik & Berlyn, 1979; Graf & Landwehr, 2015; Leder et al., 2004; Silvia, 2005).”

“Though contention ensues over whether artificial programs are capable of producing “creative” products (in line with the computer science principle known as the Lovelace Objection; Natale & Henrickson, 2022), it is nonetheless critical to understand how audiences perceive such art given that contemporary psychological…”

“…e.g., Margulis et al., 2019; McAuley et al., 2021), the act of deriving meaning from artwork that is often posited in models of aesthetics (e.g., Leder et al., 2004; Pelowski et al., 2017), and that effort and time involved in creation of a product is often a heuristic for quality (Kruger…”

“…the extant literature on the role of labels in subjective evaluations—both in aesthetics and other domains of preferences (e.g., Kirk et al., 2009; McClure et al., 2004; Newman & Bloom, 2012; Plassmann et al., 2008; Turpin et al., 2019)—our findings demonstrate that labels significantly…”

“…in subjective evaluations—both in aesthetics and other domains of preferences (e.g., Kirk et al., 2009; McClure et al., 2004; Newman & Bloom, 2012; Plassmann et al., 2008; Turpin et al., 2019)—our findings demonstrate that labels significantly influence people’s judgements of visual artworks.”

“Before rating the images, participants were asked to complete the 7-item cognitive reflection test (CRT; Frederick, 2005; Toplak et al., 2014).”

“Participants also completed the General Attitudes toward Artificial Intelligence Scale (GAAIS) to assess positive and negative sentiments of AI in general (Schepman & Rodway, 2020).”

“Past work with the CRT has found negative relationships between CRT scores and perceived profundity of randomly generated statements (Pennycook et al., 2015), implying a relationship between cognitive reflection traits and subjective judgements of profundity that could extend to aesthetics.”

“The 20-item GAAIS has been fully validated and found to report a consistent positive subscale (⍺ = 0.88) and negative subscale (⍺ = 0.82), with strong convergent validity with other AI-related measures (Schepman & Rodway, 2022).”

“…that can produce, as purely physical stimuli, high-quality artworks that are indiscernible from human-created artworks (Gangadharbatla, 2022; Johnson, 1997; Mazzone & Elgammal, 2019), the Computer Age in which we live has afforded considerable experimental control over these two conceptions of art.”

“…et al., 2019; McAuley et al., 2021), the act of deriving meaning from artwork that is often posited in models of aesthetics (e.g., Leder et al., 2004; Pelowski et al., 2017), and that effort and time involved in creation of a product is often a heuristic for quality (Kruger et al., 2004).”

“Other personality traits explored in aesthetics and creativity research include empathy and embodied cognitions (Freedberg & Gallese, 2007; Rusu, 2017), and beliefs of creative mindsets (Hass et al., 2016; Karwowski, 2014).”

“(2004), people use effort as a metric to inform their judgements of the quality of a product (e.g., poems, paintings, suits of armor; Kruger et al., 2004).”

“In fact, several recent studies have reported similar anti-AI findings in music (Shank et al., 2022), creative writing (Raj et al., 2023), dance (Darda & Cross, 2023), poetry (Köbis & Mossink, 2020), and non-art texts (Darda et al., 2023).”

“In line with the extant literature on the role of labels in subjective evaluations—both in aesthetics and other domains of preferences (e.g., Kirk et al., 2009; McClure et al., 2004; Newman & Bloom, 2012; Plassmann et al., 2008; Turpin et al., 2019)—our findings demonstrate that labels…”

“…can also assess Consciousness, Extraversion, Agreeableness, and Neuroticism, we chose to implement only the Openness subset given its previous specific involvement in aesthetic encounters (Kaufman, 2013; McCrae, 2007; McCrae & Greenberg, 2014; Silvia et al., 2015), and to reduce study length.”

“…can also assess Consciousness, Extraversion, Agreeableness, and Neuroticism, we chose to implement only the Openness subset given its previous specific involvement in aesthetic encounters (Kaufman, 2013; McCrae, 2007; McCrae & Greenberg, 2014; Silvia et al., 2015), and to reduce study length.”

“…for marketing and product branding, as it is reminiscent of the mass production advertising practices in the early twentieth century where consumers were overall hesitant to purchase goods from multinational conglomerates until narratives were paired with the product (Freeman, 2014; Pulizzi, 2012).”

“These latter criteria require higher cognitive elaboration to determine communicative properties of art, as discussed in Graf and Landwehr (2015) and Sherman and Morrissey (2017).”

“…can also assess Consciousness, Extraversion, Agreeableness, and Neuroticism, we chose to implement only the Openness subset given its previous specific involvement in aesthetic encounters (Kaufman, 2013; McCrae, 2007; McCrae & Greenberg, 2014; Silvia et al., 2015), and to reduce study length.”

“Third, given that aesthetic appraisals are personal and subjective (Chatterjee, 2014; Roseman & Evdokas, 2004), also imperative in their investigation is consideration of individual differences that might reliably predict these judgements.”

“Other emerging models have since provided more nuance to Berlyne’s traditional behaviorist approach (see Berlyne, 1975; Silvia, 2005), including more cognitive principles, and an increased understanding of the processes people utilize to engage with art.”

“…to empathize with other agents, including AI, perhaps explaining AI-art judgements), openness to experience (given its role in aesthetic encounters; Kaufman, 2013; McCrae, 2007; McCrae & Greenberg, 2014; Silvia et al., 2015), personal attitudes toward AI (which we hypothesized could predict…”

“, those that were false in cases wherein the labels indicated that the artworks were created by a human), although we can nevertheless minimize these concerns (at least to some extent) by considering the past literature showing participants’ inability to discriminate between AI- and human-created paintings (Chamberlai…”

“Critically, however, other studies have yielded results that are at odds with those from the aforementioned studies, finding little-to-no differences in evaluations of human- and AI-created artworks (Hong & Curran, 2019; Israfilzade, 2020; Xu et al., 2020).”

“…nevertheless emerged across peoples’ aesthetic judgements of such artworks (e.g., blue is the most preferred color, whereas yellow is most disliked [Bornstein, 1975; Komar & Melamid, 1999; Palmer et al., 2013]; and representational art is enjoyed over abstract art [Heinrichs & Cupchik, 1985;…”

“In addition, openness has been shown to influence judgements of art depending on painting type, such that people who are more open to experience tend to have an increased appreciation for abstract art compared to less-open participants (Feist & Brady, 2004).”

“In other words, people consider non-sensory aspects of art, like context and background, in their judgements and evaluations of art (Blank et al., 1984; Chatterjee & Vartanian, 2016; Winner, 1982).”

“…considers creativity to be not an absolute construct, but a consequence of changing interactions between subjective criteria and socio-cultural factors like technology (i.e., “in the eye of the beholder”; Amabile, 1982; Cseh & Jeffries, 2019; Hennessey et al., 2011; Kaufman & Sternberg, 2010).”

“It can encapsulate and communicate our emotions, be used to express our individualistic and communal experiences, and serve as a social commentary on these experiences, all of which are elements that are commonly thought to be human-specific (Chatterjee, 2014).”

“…across peoples’ aesthetic judgements of such artworks (e.g., blue is the most preferred color, whereas yellow is most disliked [Bornstein, 1975; Komar & Melamid, 1999; Palmer et al., 2013]; and representational art is enjoyed over abstract art [Heinrichs & Cupchik, 1985; Kettlewell et al.,…”

“Other personality traits explored in aesthetics and creativity research include empathy and embodied cognitions (Freedberg & Gallese, 2007; Rusu, 2017), and beliefs of creative mindsets (Hass et al., 2016; Karwowski, 2014).”

“, those that were false in cases wherein the labels indicated that the artworks were created by a human), although we can nevertheless minimize these concerns (at least to some extent) by considering the past literature showing participants’ inability to discriminate between AI- and human-created paintings (Chamberlai…”

“Art created by AI has been met with both success and controversy (see Chatterjee, 2022); and, very recently, these creative products have permeated into mainstream culture like museums and artistic industries (Diaz, 2022; Kaleagasi, 2017; Kinsella, 2018; Zulić, 2019).”

“Similar findings emerged in a more-recent study in which participants were shown, and asked to provide judgements of, two abstract paintings (Chiarella et al., 2022).”

“…whether artificial programs are capable of producing “creative” products (in line with the computer science principle known as the Lovelace Objection; Natale & Henrickson, 2022), it is nonetheless critical to understand how audiences perceive such art given that contemporary psychological…”

“This is supported, in part, by the growing discussion of art not only as a bottom-up appraisal of only visual features that induces liking or judgements of beauty, but a device for communication and socio-epistemic value (see Sherman & Morrissey, 2017).”

“Such work (e.g., Margulis et al., 2019; McAuley et al., 2021) finds interesting cross-cultural differences, implicating the role of media in our consumption of art and underlying associations (see Wu et al., 2020 for a cross-cultural study that finds higher appreciation for AI products of art…”

“In fact, several recent studies have reported similar anti-AI findings in music (Shank et al., 2022), creative writing (Raj et al., 2023), dance (Darda & Cross, 2023), poetry (Köbis & Mossink, 2020), and non-art texts (Darda et al., 2023).”

“…(e.g., blue is the most preferred color, whereas yellow is most disliked [Bornstein, 1975; Komar & Melamid, 1999; Palmer et al., 2013]; and representational art is enjoyed over abstract art [Heinrichs & Cupchik, 1985; Kettlewell et al., 1990; Knapp & Wulff, 1963; Mastandrea et al., 2011, 2021]).”

“Art created by AI has been met with both success and controversy (see Chatterjee, 2022); and, very recently, these creative products have permeated into mainstream culture like museums and artistic industries (Diaz, 2022; Kaleagasi, 2017; Kinsella, 2018; Zulić, 2019).”

“In other words, people consider non-sensory aspects of art, like context and background, in their judgements and evaluations of art (Blank et al., 1984; Chatterjee & Vartanian, 2016; Winner, 1982).”

“Other emerging models have since provided more nuance to Berlyne’s traditional behaviorist approach (see Berlyne, 1975; Silvia, 2005), including more cognitive principles, and an increased understanding of the processes people utilize to engage with art.”

“…(e.g., blue is the most preferred color, whereas yellow is most disliked [Bornstein, 1975; Komar & Melamid, 1999; Palmer et al., 2013]; and representational art is enjoyed over abstract art [Heinrichs & Cupchik, 1985; Kettlewell et al., 1990; Knapp & Wulff, 1963; Mastandrea et al., 2011, 2021]).”

“…(e.g., blue is the most preferred color, whereas yellow is most disliked [Bornstein, 1975; Komar & Melamid, 1999; Palmer et al., 2013]; and representational art is enjoyed over abstract art [Heinrichs & Cupchik, 1985; Kettlewell et al., 1990; Knapp & Wulff, 1963; Mastandrea et al., 2011, 2021]).”

“Such work (e.g., Margulis et al., 2019; McAuley et al., 2021) finds interesting cross-cultural differences, implicating the role of media in our consumption of art and underlying associations (see Wu et al., 2020 for a cross-cultural study that finds higher appreciation for AI products of art…”

“, 2023), dance (Darda & Cross, 2023), poetry (Köbis & Mossink, 2020), and non-art texts (Darda et al., 2023).”

“…(e.g., blue is the most preferred color, whereas yellow is most disliked [Bornstein, 1975; Komar & Melamid, 1999; Palmer et al., 2013]; and representational art is enjoyed over abstract art [Heinrichs & Cupchik, 1985; Kettlewell et al., 1990; Knapp & Wulff, 1963; Mastandrea et al., 2011, 2021]).”

“In fact, several recent studies have reported similar anti-AI findings in music (Shank et al., 2022), creative writing (Raj et al., 2023), dance (Darda & Cross, 2023), poetry (Köbis & Mossink, 2020), and non-art texts (Darda et al., 2023).”

“Critically, however, other studies have yielded results that are at odds with those from the aforementioned studies, finding little-to-no differences in evaluations of human- and AI-created artworks (Hong & Curran, 2019; Israfilzade, 2020; Xu et al., 2020).”

“In fact, several recent studies have reported similar anti-AI findings in music (Shank et al., 2022), creative writing (Raj et al., 2023), dance (Darda & Cross, 2023), poetry (Köbis & Mossink, 2020), and non-art texts (Darda et al., 2023).”

“…(e.g., blue is the most preferred color, whereas yellow is most disliked [Bornstein, 1975; Komar & Melamid, 1999; Palmer et al., 2013]; and representational art is enjoyed over abstract art [Heinrichs & Cupchik, 1985; Kettlewell et al., 1990; Knapp & Wulff, 1963; Mastandrea et al., 2011, 2021]).”

“Other personality traits explored in aesthetics and creativity research include empathy and embodied cognitions (Freedberg & Gallese, 2007; Rusu, 2017), and beliefs of creative mindsets (Hass et al., 2016; Karwowski, 2014).”

“Research on AI-art perceptions has primarily used abstract paintings as stimuli, which could limit our understanding of the true role of painting type in aesthetic judgements (Chiarella et al., 2022; Israfilzade, 2020).”

“However, for experts to have a harsher anti-AI bias than non-experts also goes against claims that “novices are… more non-aesthetically biased in their aesthetic judgments compared with experts,” which suggests that experts are more resistant to label effects (Lundy & Smith, 2017, p. 140).”

“Specifically, Mockros (1993) reported that general aesthetic judgements were rated higher by novice professional adults than novice undergraduates.”

“Various studies have examined the differences in perception and valuation between human-created and AI-generated artwork (Bellaiche et al., 2023; Chiarella et al., 2022).”

“…by AI than when told the poem is written by a human poet (two-sided Welch’s t (4571.552) = –17.398, p 0.0001, p Bonf 0.0001, Mean difference = –0.814, Cohen’s d = -0.508, 99.5% CI –0.945 to –0.683), confirming earlier findings that participants are biased against AI authorship 2,7,15 .”

“While most studies indicate a bias against GenAI-generated art (Bellaiche et al. 2023; Magni et al. 2024; Millet et al. 2023), Hong and Cur-ran (2019) observed that participants’ evaluations of art were generally unaffected by whether they knew the work was GenAI-generated.”

“Also, in studies conducted on AI disclosures and evaluations of various types of content, most results showed a negative effect (Bellaiche et al. 2023; Jakesch et al. 2019; Lim and Schmälzle 2024; Shank et al. 2023; Toff and Simon 2023; Wang and Huang 2024).”

“Bellaiche et al. (2023) examined whether participants' judgments might be explained by their ability to empathize with other agents, including AI.”

“More in general, it underscores the importance of considering individual differences in future research examining anti-AI bias regarding music and arts (Bellaiche et al., 2023), such as trust in technology or personal experience with AI as well as personality traits or cognitive fl exibility.”

“Critics argue that while AI can emulate the technical aspects of art, it lacks the subjective experience and emotional depth that characterize human-driven creativity (Bellaiche et al., 2023).”

“Research on AI aversion in Social AI like Chat-GPT has found some evidence of a similar resistance (Böhm et al. 2023; Lim and Schmälzle 2024), for example, in creative domains (Bauer et al. 2024; Bellaiche et al. 2023; Shank et al. 2023).”

“Information about human authorship in creative domains such as art, literature, and music has been shown to enhance perceived authenticity and quality [13, 55].”

“These results suggest that stereotypes about technological savviness may cause perceptual harms, in the form of AI suspicion, against groups seen as likely to uptake these new technologies, including groups like White men who usually stand to benefit from social biases.”

“A longitudinal approach allows it to make the distinction between AI generated and human generated art based on deep historical insights [32].”

“This aligns with research [6] discussing biases when viewers are aware that a piece of art is AI-generated, impacting their interpretation and emotional connection.”

“When comparing AI artwork with human artwork, even with labels swapped (i.e., a human artwork labelled as AI-generated), raters tend to disparage the AI art more often than not [3].”

“For example, Bellaiche et al. (2023) highlighted that disclosure significantly affects individuals’ evaluations, with the “AI” label frequently associated with negative perceptions.”

“Additionally, to identify prominent reasons for negative comments, topic modeling was adopted to analyze the 42199 comments for AIGP and 39763 comments for human paintings, respectively.”

“However, people have a more favorable judgment of AI-generated art when a human is engaged in the making process (Bellaiche et al., 2023).”

“…the experience of AI-attributed artworks, with those artworks being appreciated less than their human-attributed counterparts (Agudo et al., 2022; Bellaiche et al., 2023; Chamberlain et al., 2018; Gangadharbatla, 2022; Hong and Curran, 2019; Millet et al., 2023; Ragot et al., 2020; Wu et al.,…”

“Bellaiche et al. (2023), for instance, found that people preferred works of art they thought were produced by a human over those produced by an AI, even though they were the same works , seemingly because they thought the human work more morally authentic – that is, produced by an entity capable of…”

“The aesthetic acceptance of AI art by humans is an area worth exploring (Bellaiche et al., 2023).”

“Similarly, when AI-generated paintings were labeled as AI-generated, they received lower ratings of Liking, Beauty, Profundity, and Worth than when the same paintings were labeled as human-created (Bellaiche et al., 2023).”

“The empirical evidence suggests that capturing AI attitudes that are deemed “general” has predictive value across a variety of contexts (e.g., Alruwaili et al., 2024; Bellaiche et al., 2023).”

“In the broader AI-art context, Horton et al. 81 and Bellaiche et al. 82 show that people often display a systematic bias in favour of human authorship: human-created artworks are preferred over AI-created ones, and the presence of AI output can even lead observers to upgrade their evaluations of human creativity.”

“Furthermore, the evidence on the potential negative bias people may have against AI-generated artwork compared to human-created art (Bellaiche et al., 2023) adds another level of complexity to this debate.”

“Following Bellaiche (2023), the Aesthetic Judgment Scale was used to evaluate participants’ aesthetic judgments across four dimensions: liking, beauty, profundity and worth.”

“Moreover, in the specific art domain, a study demonstrated that people prefer labeled human-created art over AI-created art, particularly when it is perceived as conveying deeper meanings (Bellaiche et al. 2023).”

“In recent developments, core marketing activities such as segment targeting, content customization, and customer interaction have been redefined by AI’s capacity to process complex data sets and anticipate user behavior patterns [2][3].”

“While AI can imitate esthetic qualities, it will struggle to replicate the deeper emotional and narrative elements provided by human creativity, which consumers value highly in art products (Bellaiche et al., 2023).”

“This may explain why AI-generated art is often evaluated less favorably than human works (Bellaiche et al., 2023; Shank et al., 2023), even when nearly indistinguishable from them (Chamberlain et al., 2018; Hong et al., 2022).”

“In traditional art markets, value has been attributed to its technical execution, origi - nality, and the identity of its creator, with the physical act of creation providing observable evidence of skill and labor (Bellaiche et al., 2023 ; Hernando & Campo, 2017 ).”