The enduring schism between culture and technique: Historical (dis)continuities between the Daguerreotype and Generative AI

Abstract

This article contends that the promises and perils that people ascribe to so-called generative AI are in fact not so singular and that the debates around AI and the arts can be situated within a longer discursive lineage. We claim that antithetical responses to emerging new media historically tend to activate a perceived schism between culture and technique. As such, whenever a new technology appears to impact the artistic process, there is a tendency to present human culture as a constellation of specific friction-laden tasks that should either be protected from or supplanted by technology. To support this claim, we engage in a ‘topos study’ (Huhtamo, 1997) to compare current discourses on the effects of generative AI on the arts with debates about the relationship between media and art that took place nearly two hundred years ago: François Arago’s 1839 report on the daguerreotype and Charles Baudelaire’s 1859 denunciation of photography. We identify four recurring topological commonplaces – usefulness of a new medium for the arts, democratization through technological means, antagonism between industrialization and art, and the impoverishment of artistic imagination, and argue that, when combined, they contain discursive contradictions about friction and frictionlessness that historically define the meanings ascribed to technology and its impact on the realm of culture.

Introduction

Since the release of so-called ‘generative AI’ models such as ChatGPT, Dall-E, and Midjourney, public debate has been flooded with discourses about the benefits and perils of these technologies for the arts. Many CEOs, investors, and entrepreneurs sell and trade such technologies by claiming that generative AI ‘will change just about everything’ (Kaplan, 2024: 8), that it ‘will extend our creativity and knowledge, our collective imagination, and the ability to do anything’ (Murati and Stern, 2024), and that we are ‘heading towards the best world ever’ (Altman, 2023). These statements amplify existing ideas put forth by Silicon Valley that suggest artificial intelligence will become the most significant human invention – even ‘more important than fire or electricity’, as Google’s CEO Sundar Pichai put it in 2018 (Clifford, 2018). Within these hyperbolic narratives of technological revolution lie the evergreen desires for labour-reducing automation and the enhancement of human creativity and productivity, and the proselytizers of AI do not hesitate to make trails through this forest. At the same time, the rise of generative AI is creating concerns about the future of ‘white-collar work’ (Lowrey, 2023), existential anxieties regarding the definition of art and creativity as linked to our ‘human essence’ (Cave, 2023; Chiang, 2024; Kelly, 2019), and concrete struggles related to questions of authorship, creative labour, and copyright infringement (Chow and Celis Bueno, 2025; Mueller, 2023; Vincent, 2022). Even if some of these accounts may seem exaggerated, evangelists and sceptics alike assert that ‘AI’ is a singular technology that uniquely transforms the landscape that lies between technology and creativity.

In this article, we contend that the promises and perils that people ascribe to AI are in fact not so singular and that the debates around AI and the arts can be situated within a longer discursive lineage. Furthermore, we claim that behind this longer lineage lies a recurring tension between culture and technique that is mobilized every time a new technology appears to impact the artistic process. In this sense, there is a tendency to perceive culture and technique as separate domains, and to define human creative practices as a constellation of specific friction-laden tasks that should either be protected from or supplanted by the emerging media. To support this claim, we follow the work of media historians such as Marvin (1988), who explores the complex relation between the new and the old in media technologies, and Huhtamo (1997), who claims that media scholarship should grant as much relevance to the discourses surrounding technology as to the technologies themselves. More specifically, Huhtamo argues that the role of media historians is to identify particular topoi (also referred to as commonplaces) that recur in different historical contexts and in relation to different media.

The aim of this article is therefore to engage in a ‘topos study’ (Huhtamo, 1997) to compare current discourses on the effects of generative AI on the arts with debates about the relationship between media and art that took place nearly two hundred years ago. In particular, this article focuses on François Arago’s 1839 report to the French Chamber of Deputies and Charles Baudelaire’s denunciation of photography in his review of the 1859 Salon de Paris. ¹ We compare these antithetical responses to those made by contemporary technology corporations and artists. In doing so, we identify four topoi or commonplaces that recur: ‘usefulness of a new medium for the arts’, ‘democratization through technological means’ and their antithetical formulas of ‘antagonism between industrialization and art’, and the ‘impoverishment of artistic imagination’. Importantly, we argue that, when combined, these enduring commonplaces deserve attention specifically because they contain discursive contradictions about the schism between culture and technique that have defined the modern Western response to technology and its impact on the artistic process. This schism is neither resolvable within a single generation nor through a single medium, and therefore endures throughout the history of technology. Moreover, we argue that each new technological disruption in the arts reactivates these antithetical responses and thereby revives the schism between culture and technique. Hence, we believe that it is only through the identification of these historical commonplaces that we are capable of highlighting what is new in contemporary discourses about AI and art. As such, the contemplation of the old becomes the necessary complement to understanding how these discourses are also subject to historical change, which provides precise terms for identifying shifting understandings of human culture and the role of technology within it.

New media histories and enduring topoi

Studying the history of new media is often considered synonymous with studying the history of technological changes. Conceptually, this includes investigating media as they progress through biographical stages (Lesage and Simone, 2019; Natale and Balbi, 2014), evolutionary lifecycles (Scolari, 2023), or project-based transitions from utopian plans to ideological technologies (Flichy, 2007). Of course, new media have also been venerated for their capacity to fundamentally change the very basis of culture, a position that was famously championed by Marshall McLuhan. However, even though Carolyn Marvin was directly inspired by McLuhan’s provocative probes concerning electric light as a medium of mass communication, her book When Old Technologies Were New (1988) stands out in countering his suggestion of schismatic change. According to Marvin, the concept of ‘new media’ should be understood as a ‘historically relative term’ (1988: 3), where the experience of change is in fact ‘real’ and ‘unique’, but the structure of this experience is a recurring pattern for each new generation who encounters technological development. For Benjamin Peters, likewise, the notions of ‘new’ and ‘old’ must be enlisted in ways that ‘complement one another: the historian’s eye for contingent change can lead to a fuller understanding of the contemporary relevance of media; so too can new media scholars engage the present more forcefully with historiographical cautions in mind’ (2009: 15).

To accentuate this imbalance between the old and the new, Olivier Driessens (2022) developed a ‘taxonomy of continuity’ based on a thematic review of communication and media journals. In this study he found that scholars use the ideas of active continuation, stasis, incorporating change, history, and endurance as conceptual offsets to notions of the new (2022: 37). Of these types of continuity, the notion of endurance resonates most clearly with Marvin and Peters’ arguments; scholars who focus on this notion often attend to the ways that ‘[c]hange masks continuity’ by establishing ‘the deep historical roots of “recent” phenomena, or how “new” media practices are actually a re-articulation of much older practices’ (Driessens, 2022: 38).

This recent description of endurance finds good company with a particular genealogy of media theory. In 2017, Liam Cole Young distinguished three branching lineages stemming from communication studies, anthropology, and literary studies. He described this last branch as a way of thinking about media that employs the methods of close reading, privileges analytical depth over breadth, and treats discourse and texts ‘not as founts of human meaning but as indexes or traces of technical systems that structure ways of knowing’ (Young, 2017: 20). In other words, instead of being limited to the analysis of media content, it includes the analysis of underlying forms, patterns, and repetitions that shape long-lasting meanings. This literary inclination within media theory can also be explicitly read into the way media scholars have used literary concepts to study the histories of media through myth (Bory, 2020; Mosco, 2004; Natale and Ballatore, 2017) and metaphor (Wyatt, 2004; van den Boomen, 2014; Puschmann, 2014; Farkas and Maloney, 2024). In addition to these two concepts there is a third, although relatively overlooked, literary term that has been repurposed to study media histories. We argue that the concept of topoi – or commonplaces – provides a productive framework for answering questions about continuity that are different from those that emerge when studying media myths and metaphors. ²

Topoi have their origins in ‘the rhetorical traditions of classical antiquity’ that serve as ‘storehouses’ for ‘the composition of oration’ (Huhtamo, 1997: 222). This Aristotelian meaning of topoi continues today under argumentation theory where a topos is a commonly understood topic that is rhetorically used to make a connection between an argument and its logical conclusions (Žagar, 2010: 20). However, this rhetorical meaning was subsumed by literary studies in the twentieth century when E.R. Curtius and the New Rhetoric used the term to describe literature (Huhtamo, 1997: 222). What this twentieth-century re-articulation of topoi brought was a consequential shift in its meaning, one that identified a key characteristic of commonplaces that gives them a unique value when assessing cultural history. According to Igor Žagar, the most salient topoi are composed of ‘antithetical couple[s]’ and demonstrate a society’s ‘preference of certain values’ (2010: 22). For this reason, the topoi that are worth studying are those that contain and organize a society’s enduring contradictions. This identification of antithetical couplings as topoi enables an examination of the way ‘desires and fears from earlier historical moments reappear in unexpected ways to complicate linear media histories’ (Young, 2017: 26).

Accordingly, a ‘topos study’ of media seeks to abandon the temporal linearity that informs the idea of technical progress and, instead, to identify the antithetical and ‘cyclically recurring elements and motives underlying and guiding the development of media culture’ (Huhtamo 1997: 224). In light of this, Huhtamo elaborated on the literary use of individual topos that are assembled as topoi, ‘cultural patterns’ that become ‘a stereotypical formula evoked over and over again in different guises and for varying purposes’ (Huhtamo, 2011: 28). For example, the content of a topos may be inspired by ‘stereotypical metaphors’ that, through their recurrent use, represent ‘cultural patterns’ (Huhtamo, 2011: 38). As these cultural patterns intersect with one another, they join as long-lasting topoi: ‘commonplace elements or cultural motives, which have been encountered in earlier cultural processes’ (Huhtamo 1997: 224).

In this article we employ Huhtamo’s notion of topoi to trace the commonplace elements that structure the antithetical responses to technological change in the arts by comparing two particular moments: the development of the daguerreotype in the mid-nineteenth century and the current emergence of generative AI models. We argue that these two moments share similar antithetical responses that are structured by an enduring commonplace: the fundamental opposition between the domains of culture and technique. According to this commonplace, culture belongs to the realm of values and goals (or ends), while technique needs to be placed in the realm of means (Simondon, 2015: 18). This conflict between technique and culture is reactivated every time a new medium enters the domain of artistic production, informing both the rejection of the new medium in those discourses that want to preserve the purity of culture (as the domain of pure ends) and its integration in those discourses that define progress as the gradual expansion of technique into the domain of culture. More specifically, the topoi that activate this schism discourage an integrated perspective that views culture and technique as co-constitutive and instead stimulate positions in which one side prevails over the other (i.e., the artistic ends of the cultural domain are suggested either to dissolve entirely into or to be fundamentally incompatible with the new technological means on offer). As we will demonstrate, the topoi that propel this schismatic conflict tend to take shape around particular frictions – specific skill-demanding and time-consuming elements within the creative process – that are presented as thoroughly inimical to or entirely dissoluble by the emerging medium.

Arago’s report (1839)

In 1839, the revolutionary capacities of mechanical reproduction caught the imagination of François Arago, a nineteenth-century French politician and scientist (Tresch, 2012; Wood, 1997). During this year, the French Chamber of Deputies was discussing a bill proposal to grant a State pension to Louis Daguerre in return for the purchase of the ‘secret of his daguerreotype’ (Arago, 1980; Tresch, 2012; Wood, 1997). To communicate his own technological dreams, Arago drafted his Report which made the case to the Chamber for three fundamental values of Daguerre’s invention for French society: speed, fidelity, and simplicity.

As Arago begins his report, his first concern is to establish the originality of the daguerreotype – that is, its radical break with previous media. To do so, he compares it to previous attempts to produce photographic images, arguing that the daguerreotype is an ‘entirely new process […] capable of reproducing images sixty to eighty times more rapidly than the earlier process’ (Arago, 1980: 16). He then praises the ‘usefulness of this discovery’, and the ‘extraordinary advantages […] derived from so exact and rapid means of reproduction’ (17). To illustrate its usefulness, especially as a labour-saving device, Arago gives the example of the 1798 French expedition to Egypt: ‘To copy the millions of hieroglyphics which cover even the exterior of the great monuments […] would require decades of time and legions of draughtsmen. By daguerreotype one person would suffice to accomplish this immense work successfully’ (17). Towards the end of the report, Arago returns to this aspect and concludes that ‘the rapidity of the method has probably astonished the public more than anything else’ (19). Secondly, Arago’s report praises not only the daguerreotype’s speed but also its ‘exact means of reproduction’ (17). In this sense, the images produced by the daguerreotype are characterized by an ‘unimaginable precision’ and ‘accuracy of the lines’ (Delaroche in Arago, 1980: 18). He then establishes a comparison between photography and painting, arguing that the daguerreotype ‘excels the works of most accomplished painters, in fidelity of detail and true reproduction of the local atmosphere’ (17). Furthermore, he claims that since Daguerre’s invention ‘follows the laws of geometry, it will be possible to reestablish […] the exact size of the highest points of the most inaccessible structures’ (17). Beyond speed and precision, Arago’s report also draws attention to the simplicity of the process. According to Arago (19), the Daguerreotype calls for manipulations that ‘anyone can perform’. This means that it ‘presumes no knowledge of the art of drawing and demands no special dexterity’ (19). If ‘a few simple rules’ are followed step by step, ‘there is no one who cannot succeed’ (19). As we will see further below, the simplicity of the process is relevant not only because it reinforces the claims of speed and efficiency, but also because it represents a ‘democratization’ of the ability to produce images.

Based on these three characteristics of the daguerreotype – speed, fidelity, and simplicity – we can identify two topoi or commonplaces deployed by Arago to convince his audience of the value of Daguerre’s invention. The usefulness of the daguerreotype for the arts defines the first topos. Arago builds his case by referring to a report written by a ‘celebrated painter’, Paul Delaroche, aimed at answering the question of ‘whether art may expect further progress from the study of these images drawn by nature’s most subtle pencil, the light ray’ (Arago, 1980: 18). According to Delaroche, the painter finds in the daguerreotype ‘an easy way of making collections for after-study which otherwise are obtainable only at great expense of time and labour, and yet less perfect in quality, no matter how great his talents may be’ (Delaroche in Arago, 1980: 18). As we can see, Arago refers to the three aspects of speed, fidelity and simplicity as key characteristics that allow artists to increase their ability to collect images for their ‘after-study’. Painting is seen here as a laborious process that requires constant training and the close examination of its objects. Photography complements the painter’s laborious processes because it offers an efficient, fast, and precise way to produce images. For this reason, Arago claims that ‘after having opposed with excellent arguments the opinions of those who imagined that photography would be detrimental to our artists, [Delaroche concludes that] the remarkable invention of Daguerre is a great service rendered to the Arts’ (1980: 18).

Democratization through technological means comprises the second topos. Two aspects are key here. First, the fidelity of the photographic image is linked to its ‘objectivity’ and hence to a democratization of truth (Dinius, 2012; Thwaites, 2013). Second, and more relevant for our analysis, the simplicity of the daguerreotype’s operation is presented as a key argument for a democratization of artistic production. The fact that the successful operation of the daguerreotype ‘presumes no knowledge of the art of drawing and demands no special dexterity’ means that the new medium can be seen as a means that lowers the entry bar to both the production and ownership of images, a prospect that until then remained limited to those with particular training and/or resources. As we present in the following section, this democratization through technological means was a point of strong contention in the debates regarding the relationship between photography and the arts at the time (Doan, 2023).

We argue that these two topoi or commonplaces are brought together by an underlying assumption that the daguerreotype entails the technological mitigation of certain frictions within the artistic process. As such, the daguerreotype is understood as an expansion of technique into the realm of culture in which the frictions that are intrinsic to the latter (effort, skill, and resources as the result of a laborious process of ‘cultivation’ that requires time) can be mitigated by the new technological medium. Indeed, with each of his arguments – whether focused on speed, fidelity, or simplicity – Arago makes a case for how the daguerreotype exceeds certain human capacities, and how this yields a desirable minimization of the frictions that characterize culture. This friction can be temporal (e.g., the time that it takes to produce an image), or linked to particular skills and complex technical processes that one needs to master in order to produce appealing and/or faithful images (e.g., training and observation). In light of this, the daguerreotype is presented as a new medium that can reduce these different forms of friction by speeding up the production of images, bypassing the need to learn and train particular artistic skills, and replacing complex technical procedures, rendering artistic practices more accessible and democratic. In this regard, Arago’s account can be placed within a longer media-historical trend of conceiving of culture as a constellation of frictions that can and must be mitigated by technical means. As mentioned above, however, the topoi that have the strongest cultural significance are those that are composed by ‘antithetical couples’ (Žagar, 2010). This means that the idea of a technological reduction of friction in the arts acquires its full significance when examined not only through its proponents, but also through its detractors.

Baudelaire’s 1859 Salon

Unlike François Arago, Charles Baudelaire’s (1859) reception of photography was far from enthusiastic. Writing twenty years after Arago’s report, the French writer and poet described photography as ‘art’s mortal enemy’ and a crucial reason behind the ‘impoverishment’ and ‘corruption’ of both French artists and the French public alike (1956: 232). The context of Baudelaire’s denunciation of photography was his Salon de 1859, a review of the Parisian Salon published in four instalments in the Revue Française (Baudelaire, 1956; Blood, 1986). According to Susan Blood, the uniqueness of this version of the Parisian Salon was that it ‘saw the first exhibition of photographs, not yet included among, but adjoining the exhibition of paintings’ (1986: 820). In this context, Baudelaire defines photography ‘as a sterile technology with no future in the fine arts’ (Grøtta, 2013: 80). Informed by his judgement, Baudelaire ‘excludes photography from the aesthetic domain […] and consigns it to the domain of technology’ (Blood, 1986: 821). But his criticism is not purely contextual; rather, it can be read as revealing a more profound conceptualization of the relation between technology and the arts, that is, between technique and culture. To unpack this conceptualization, however, we first need to identify two notions that inform Baudelaire’s position: progress and imagination. When brought together, these notions allow us to identify two topoi or commonplaces at the heart of his text: the antagonism between industrialization and art and the impoverishment of artistic imagination.

In the 1859 review of the Salon de Paris, Baudelaire describes his present time as ‘plung[ing] into the track of progress’ (1956: 229). ‘By progress’, he adds, ‘I mean the progressive domination of matter’ (1956: 229). As Susan Blood (1986: 828) puts it, Baudelaire's conception of progress is the triumph of a ‘materialist viewpoint’ that he associates with conceptual confusion and moral decadence. This was a common theme in Baudelaire’s latter writings – as opposed to his earlier attitude in the 1840s (Blood, 1986). In 1855, for example, he referred to progress as ‘a grotesque idea’ which was flowering ‘upon the rotten soil of modern fatuity’ (1956: 198). His rejection of progress is in fact a rejection of the two underlying ‘materialist viewpoints’ of ‘naturalism’ and ‘industrialism’ (1956: 199). Both, he claims, prioritize ‘nature’ and ‘matter’, forgetting the essential difference that separates the physical from the moral world (1956: 199). The idea of progress is thus linked not only to technological development, but to a materialist viewpoint that believes purely in nature (Baudelaire, 1956: 230). Reproducing the schism between culture and technique mentioned above, Baudelaire claims that the superiority of the former – which encompasses not only the moral world, but also the domains of art, beauty, the imagination, and the intangible – was ‘plunging’ into a worldview dominated by technologically driven ideas of progress, nature, and industry. Photography, needless to say, was for Baudelaire an essential component of the latter.

The second concept to highlight in Baudelaire’s Salon de 1859 is that of ‘imagination’. He calls imagination ‘the Queen’ of the human faculties (1956: 233) and describes it as a ‘mysterious’ faculty, a combination of analytic skills, sensitivity, and the power to create the new (1956: 234–235). There is no human endeavour, he adds, that does not require the ‘laborious’ effort of this ‘cardinal faculty’, and no other human faculty can properly work without it (1956: 235). Furthermore, his account of imagination as a subtle but laborious process plays an essential role in his definition of art. According to Baudelaire, there are two types of artists,

those who call themselves ‘realists’ […]; and they say, ‘I want to represent things as they are, or rather as they would be, supposing that I did not exist’. In other words, the universe without man. The others however – the ‘imaginatives’ – say, I want to illuminate things with my mind, and to project their reflection upon other minds. (1956: 242)

Based on this distinction, Baudelaire argues that nature should only be seen as ‘a storehouse of images and signs’ to which the imagination of the painter should provide value, ‘a sort of pasture which the imagination must digest and transform’ (1956: 241). On the one hand, the radical forms of naturalism associated with industrialism and progress reduce the world to a materialist viewpoint; on the other, the ‘work of the imagination’, that laborious effort of ‘digesting and transforming’ matter, elevates nature to the domain of the intangible. Thus, the superiority of imaginative art proclaimed by Baudelaire is not to be read as a mere aesthetic problem but rather functions as a reminder of the superiority of culture over technique.

This antagonism between culture and technique informs two specific topoi or commonplaces present in Baudelaire’s Salon de 1859. First, we refer to Baudelaire’s claim according to which ‘photography is art’s mortal enemy’. This claim reflects a commonplace according to which industrialization and art are antithetical – and incompatible – forces. For him, photography needs to be restricted to the domain of ‘industry’ (1956: 230-31) and its operation should be understood as a mere ‘industrial process’ (1956: 230). Baudelaire then condemns the fascination of the French public with the daguerreotype as idolatry and fanatism, a result caused by the public’s inability to separate art from industry (1956: 232). Art belongs to the ‘domain of the impalpable and the imaginary’, whereas photography belongs to the physical world of nature (1956: 232). Photography replaces the laborious effort of the imagination that grounds the superiority of the artistic (and moral) domain with an automated process that reduces art to the mere imitation of nature. For this reason, Baudelaire concludes, industry and art ‘are like two ambitious men who hate one another with an instinctive hatred, and when they meet upon the same road, one of them has to give place’ (1956: 232). Whereas Arago saw no antagonism between these two domains, Baudelaire warns us against the ‘invasion of photography’ and the ‘deplorable results’ caused by the ‘industrial madness of our time’ (1956: 233).

This leads us to the idea that informs the second topos: the impoverishment of artistic imagination. According to Baudelaire, ‘photography, like all other purely material developments of progress, have contributed much to the impoverishment of the French artistic genius’ (1956: 231–232). In his well-known passage against photography (1956: 230), he writes,

In matters of painting and sculpture, the present-day Credo of the sophisticated, above all in France […], is this: ‘I believe in Nature, and I believe only in Nature […] I believe that Art is, and cannot be other than, the exact reproduction of Nature […] Thus an industry that could give us a result identical to Nature would be the absolute of art’. A revengeful God has given ear to the prayers of this multitude. Daguerre was his Messiah. And now the faithful says to himself: ‘Since photography gives us every guarantee of exactitude that we could desire (they really believe that, the mad fools!), then photography and Art are the same thing’.

Photography led to a ‘fanatism’ for ‘realist art’, glorifying mere ‘mimetic effects at the expense of true aesthetic [imaginative] beauty’ (Grøtta, 2013: 82). As Grøtta puts it, while Baudelaire ‘saw images as an inspiration for the imagination, the public was more fascinated with their verisimilitude, and the technique of photography appeared to cater to this particular preference’ (2013: 82). As mentioned above, Baudelaire saw photography as a mere copy of nature, a displacement of the laborious effort of the imagination, source, and essence of true artistic creation. Moreover, Baudelaire’s critique of this displacement can be read as antithetical to Arago’s topos of democratization through technological means. Whereas Arago celebrated the simplicity of the daguerreotype and its ‘democratizing’ potential, Baudelaire saw this simplicity ‘as the refuge of every would-be painter, every painter too ill-endowed or too lazy to complete his studies, this universal infatuation bore not only the mark of a blindness, an imbecility, but had also the air of a vengeance’ (1956: 231). In short, Baudelaire conceives artistic creation as a process that takes time and effort and condemns photography as the attempt to bypass this laborious effort by those who are too ‘lazy’ or ‘ill-endowed’.

Despite their different aesthetic, philosophical, and political perspectives, we claim that both Arago’s and Baudelaire’s responses should not be taken in isolation: both are structured by a schism between culture and technique. Furthermore, they both conceive of culture, and hence artistic creation, as a process rife with frictions. At the same time, they differ in their estimation of the aesthetic, philosophical, and political value of these frictions. While Arago’s report can be read as a celebration of the daguerreotype’s ability to reduce the frictions that underlie artistic production, Baudelaire saw these frictions as the source of the value of art (as the domain of the intangible). For the latter, artistic value comes not from the direct imitation of nature or from the acceleration of artistic production, but emerges as the result of the laborious process through which the imagination ‘digests and transforms’ (Baudelaire, 1956: 241) the elements provided by nature. As such, attempting to replace the time and effort necessary for artistic creation with the frictionless and immediate results allowed by the new medium can only result in the ‘corruption’ and ‘banalization’ of culture; a topic that is still being discussed today.

Generative AI and the arts: Between friction and frictionlessness

A recurring schism between culture and technique endures within the ongoing responses to generative AI. This schism has long shaped how people are encouraged to understand the impact of a new medium on the artistic process. By comparing current discourses on generative AI to those of Arago and Baudelaire, we examine how notions such as speed, simplicity, progress, imagination, and democratization continue to inform debates regarding the relation between culture and technique. At the same time, in line with the play of continuity and discontinuity that guides media history, there are also notable differences between the two media. In reading media history in such a way, we find continuity in how technologies are purported to smoothen, augment, or even obviate the frictions that characterize artistic and creative practices, but we also find discontinuity regarding the specific processes that this concerns and the ways in which the reduction of friction is connected to wider discussions about art and the human condition. In the remainder of this section, we highlight some consistencies and inconsistencies between Arago’s positive report on the daguerreotype and affirmative narratives about generative AI on the one hand, and Baudelaire’s critique of photography and pessimistic interpretations of generative AI on the other. In doing so, we aim to demonstrate how the concept of friction(lessness) (Kemper 2024; Kemper and Jankowski 2024) helps to illuminate historical patterns and to better position a critical stance on generative AI today.

What unites the various responses to the daguerreotype and generative AI is that both proponents and detractors frame the artistic process as an assemblage of high-friction tasks that should either be optimized by or safeguarded from the technology in question. Whereas the technology’s boosters suggest that adopting the new medium will mitigate certain frictions (often presented as ‘menial’ or ‘tedious’ tasks), detractors caution that these frictions form an indispensable part of creative practice and that eliminating them will dilute the value of the artistic process and product. The new medium, then, is understood less in terms of its discrete technical characteristics and more in terms of its connection – for better or for worse – to prevailing ideas about what the artistic process is and should be (Caramiaux et al., 2025).

As with the daguerreotype, affirmative narratives about generative AI tend to focus on the question of speed, with the underlying suggestion that the temporalities that hitherto constituted the creative process are frictions that should be overcome technologically. In both cases, the new medium is framed as a tool that will save labour-time and reduce temporal horizons. Canda (2024), for example, speaks of generative AI as an enhancement of human creativity that takes place at the level of ‘speed [...] and efficiency’. Similarly, Spair (2024) claims that AI tools ‘can help automate certain aspects of the creative process, freeing up artists’ time to focus on more complex and creative tasks’. This allows artists to explore more creative ideas ‘in less time’ (Opasinski, 2024) and ‘to work more quickly and make better use of their time’ (Marr, 2024). Finally, Chui et al. (2023: 18) argue that generative AI can ‘significantly reduce the time required for ideation and content drafting, saving valuable time and effort’. These claims resonate with Arago’s emphasis on the rapidity of the daguerreotype examined above. In both cases, technological innovation is seen to respond to and furnish a general acceleration of life that has come to define artistic and cultural production (Lijster, 2018). This acceleration belongs to a broader demand for speed characteristic of modernity – we could call this the ‘capitalist imperative’ that shapes a particular understanding of the relationship between technology and temporality (see e.g. Chui et al., 2023; Microsoft and LinkedIn, 2024) in which moments of creative protraction, deceleration, and delay are seen as frictions that are superfluous to or even inimical to the artistic process.

Strongly related to this, as with the daguerreotype, generative AI is promised to deliver new avenues for efficiency. By implication, celebratory narratives about AI seek to present established forms of creative practice as inefficient and plagued by friction. For example, both Spair (2024) and Marr (2024) suggest that generative AI can help artists to overcome ‘creative blocks’. Moreover, Marr adds that AI tools can make digital technologies ‘more user-friendly and intuitive, meaning artists and designers spend less time working out how to get the software to do what they want and more time creating’. In these examples, speed and efficiency are presented as the desired output of technological progress. In this sense, they reproduce enduring topoi according to which human progress is equivalent to the gradual expansion of technique into the domain of culture.

Because of the way the technology in question short-circuits frictional and time-consuming processes, the endorsements considered here tend to affirm the commonplace of democratization through technological means. By implying that conventional media of aesthetic expression have been impractical at best and exclusionary at worst, the new technology’s patrons claim that it will enable everyone to become an artist. Arago (1980: 19) suggested that the ‘simplicity’ of the technical process behind the daguerreotype would have a ‘democratizing’ effect in terms of the access to produce images. This echoes with current discourses that celebrate the ‘democratizing’ value of AI in general (e.g. Gates 2023), but also of generative AI in particular. As Spair (2024) puts it, generative AI has ‘democratized creativity by making it more accessible to a wider range of people. With generative AI, anyone with access to the technology can create their own unique pieces of art or design, regardless of their background or training’. Accordingly, proponents of these technologies establish a direct link between democratization and the promise of a frictionless creativity:

Humans have boundless creativity. However, the challenge of communicating their concepts in written or visual form restricts vast numbers of people from contributing new ideas. Generative AI can remove this obstacle. […] Generative AI’s greatest potential is not replacing humans; it is to assist humans in their individual and collective efforts to create hitherto unimaginable solutions. It can truly democratize innovation. (Eapen et al., 2023)

In sum, both with the daguerreotype and generative AI, technology is framed as a tool that mitigates frictions within creative practice, purportedly leading to a faster, more efficient, and more democratic artistic process. As such, both discourses reproduce a schism between culture (realm of ends) and technology (realm of means) and frame human progress as the gradual enhancement of the former by means of the latter.

This is not to say that these glorifications of frictionless artistic creation all refer to exactly the same phenomena. For instance, whereas Arago ties the daguerreotype’s achievements predominantly to the faithful representation of nature, generative AI is concerned with the more general process of converting ideas into material objects. Likewise, whereas Arago accentuates the daguerreotype’s capacity to produce photographic images, generative AI systems are now lauded as all-purpose systems that can lubricate any type of creative practice. What we see, then, is that a focus on removing frictions from the artistic process endures, but is observable within the specific tasks and practices where friction is located. Moreover, discursive expressions should always be read in relation to the dominant technological ideologies and imaginaries of the time. In this light, it is particularly salient that today’s paeans to generative AI speak to and further expand what has elsewhere been described as an ideology of frictionlessness (Kemper, 2024; Kemper and Jankowski, 2024). Today, Silicon Valley has inherited and repurposed the idea that technology should be designed to reduce frictions in everyday life. According to this vision, the wealth of creative, cultural, and social practices that hold the world together can be captured, replicated, and optimized by processes of ‘datafication’ (Mejias and Couldry, 2019). While it may be argued that the emergence of photography comprised an initial step in the now entrenched belief that technology can objectively mediate reality (Malevé and Sluis, 2023), the visions behind generative AI describe a much more far-reaching understanding of the world as potentially subsumable to processes of abstraction through datafication (Mejias and Couldry, 2019: 2). This vision legitimizes a cultural logic in which creative processes are made entirely subservient to the commodities they yield (Caramiaux et al., 2025: 6). As Goodlad and Stone (2024) put it, according to the ‘myth of frictionless knowing’ on which Silicon Valley operates, ‘writing’ and ‘art-making’ are ‘not process but product, not provenance but output’. Indeed, James Bridle (2023) observes a general belief in Silicon Valley that AI will inaugurate a frictionless form of artistry where we no longer ‘need to know how to draw, or how to write computer code’; instead, we can ‘simply whisper our desires to the machine and it would do the rest’.

Such beliefs are precisely what many critics of generative AI take issue with, and in this sense their words resonate with Baudelaire’s argument that photography produced an impoverishment of the artistic process. The most striking topological endurance here is that both Baudelaire and various critics of AI caution that the frictions the medium in question eliminates are indispensable and affectively necessary elements of the creative process (and of culture more broadly). Philosopher Sean Kelly (2019) argues that creativity is ‘one of the defining features of human beings’. Hence, if we ‘treat machine “creativity” as a substitute for our own, […] we will have lost track of the fundamental role that creativity plays in being human’ (Kelly, 2019). Kelly’s sentiments are echoed by several notable artists: Brian Eno (2024), for example, states that ‘the joy of art isn’t only the pleasure of an end result but also the experience of going through the process of having made it’ and that the ease and black-boxed nature of generative AI undermines this experience; Nick Cave (2023) dismisses songs generated by ChatGPT as a ‘mockery of what it means to be human’ by claiming that songwriting arises from the ‘suffering’ that is ‘predicated upon the complex, internal human struggle of creation’; and Ted Chiang (2024) explains that AI services designed to generate ‘images with little effort’ erase the subjective and situated frictions of decision-making that give art its meaning. These topological formulas resonate with the work of numerous scholars: Aden Evens (2024: 92), for instance, argues that the digital’s discretizing logic ‘removes contingency’ from creative processes, ‘leaving no play of potential, but only the rigid rule of necessary determination’; and Michele Elam (2023: 281–283) describes ‘fiction as friction’, asserting that the realm of literature cannot be reconciled with the tech industry’s aims of efficiency and optimization.

These accounts suggest that the artistic process is sustained rather than hampered by the protracted, uncertain, and contingent grappling with mediation. The tech sector’s attempt to short-circuit these struggles is seen to denigrate rather than enhance cultural production; like in Baudelaire, the frictions that technology promises to erase are considered to be precisely what makes us human and what imbues culture with its value. It is, moreover, not just the case that the new technology is framed to diminish the vibrant role of human interiority in the creative process, but also that this supposed elimination of friction is seen to lead to unsatisfactory artistic products. Just as Baudelaire feared the emergence of a more lifeless art as the fancies of the imagination were curtailed by photography’s possibilities of direct and immediate representation, critics of generative AI caution that its data-driven promises of frictionless artistry cause an aesthetic impoverishment. One dimension of the topos of the antagonism between industrialization and art concerns the topological idea of ‘authenticity’. Cave (2023), for example, dismisses AI-written songs as ‘inauthentic’ and ‘replication as travesty’. Another reason behind this impoverishment would be the substitution of human creativity for probabilistic calculations, leading towards standardization. In this sense, several scholars, and artists (Bridle, 2023; Steyerl, 2023) describe how generative AI encourages artistic homogeneity, leading to more uniform, predictable, and mechanistic cultural expressions.

Again, there are also discontinuities in how these topoi are concretely imagined. If, for Baudelaire, the harm of the daguerreotype consists in how it privileges the rote representation of nature over the fanciful flights of the imagination, the situation with generative AI is more complex. While Baudelaire was driven by the belief that ‘true art’ should not be reduced to the mere imitation of nature, the contemporary critics of AI are less concerned with artistic questions of natural fidelity and more with the effects of automation on aesthetic forms and genres in general. Even if defamatory accounts also focus on how the technology bypasses productive frictions in the imagination, it is not so much the case that generative AI replaces imaginative modes of art with art that exhibits a stringent realism. Moreover, Baudelaire was concerned with how the daguerreotype’s ability to directly and faithfully represent nature would discourage people from using their imagination, but for artists like Nick Cave and Brian Eno the problems with AI are more far-reaching, as the imagination itself is now becoming subject to algorithmic automation. A related difference is that, where for Baudelaire the daguerreotype reduces representation to literality and faithfulness (which in turn dulls the creative frictions of the imagination), generative AI reduces everything to calculation. Aden Evens, for instance, argues that AI-driven technologies construct a ‘digital world’ that ‘restricts meaning to the predeterminations of the rules that govern its operations’ (2024: 206).

Conclusion

To understand media history through the schism between culture and technique is to identify processes of repetition and change in how the stakes of new media are articulated. To conceive of technology in this way is to avoid essentializing accounts that understand emerging media as uniquely beneficial or detrimental and instead to situate responses within a historical topology of antagonisms. It is also to assert that the discourses that emerge around new media enduringly encourage an image of the artistic process as an assemblage of high-friction processes that should either be optimized through or inoculated against technological developments.

We developed this argument by showing how the claims pertaining to generative AI are continuous with the discourses that were used to advocate for the daguerreotype in the mid-1800s. To this point, our ‘topos study’ has highlighted how the topoi of the usefulness of a new medium for the arts, democratisation through technological means, and their antithetical couplet of antagonism between industrialization and art, and the impoverishment of artistic imagination endured in discourses pertaining to both the daguerreotype and generative AI. Some salient and recurring questions that structure these topological commonplaces are: which aspects of human culture can be satisfactorily outsourced to or augmented by technical operations and which aspects are worth preserving in the face of technological innovations? Is the point of media to generate more ‘authentic’ forms of cultural expression, or to make it easier to make more art? Across the history of technology, these concerns have animated technological boosters and critics alike. However, we argued that it is only after establishing these enduring topoi that it is possible to assess how discourses about AI substantially differ from previous narratives.

As we pointed out, the different discourses examined in this paper exhibit a persistent schism between the domain of ends (culture) and the domain of means (technique). Both Arago and Baudelaire agreed on what photography as a means was capable of: the mechanical imitation of nature. They differed, however, on the relation between means and ends. For Arago, the new medium represented an aid to and a democratization of the realm of culture, whereas for Baudelaire it meant a form of replacement that led to the impoverishment of culture. Similarly, critics and supporters of generative AI agree on the fact that the new medium accelerates and simplifies artistic processes, but disagree on the attached value that this acceleration and simplification may entail. Beyond these continuities, we have suggested that the pivotal difference between discourses surrounding both media is that with generative AI the focus shifts from the ability to mechanically imitate nature towards the promises of an unmediated – and therefore frictionless – automation of the imagination itself. Regardless of this sharp difference, we see that discourses and debates about new media persistently boil down to an existential provocation: within the vast edifice of human expressivity, which frictions can be productively erased by technology and which frictions are indispensable to a fulfilling creative life? It is our hope that illuminating the schism on which this provocation often hinges and the various topoi that have historically converged around it will provide more context to contemporary debates about generative AI and will open up pathways toward more integrated perspectives that underscore the constitutive entanglement of culture and technique.