When the appearance of synthetic images generated through multimodal models or popular text2image models is even considered, it is in terms of style and or kitsch, or as a provocation towards photographic indexicality. Parallel to critiques of artificial imagination I want to place a critique of the appearance of the artificial. Consequently, in this talk I pay attention to the phenomenal appearance of AI generated digital image, to speak to the way in which the experience of their appearance can constitute the appearance. This entails a media phenomenology for aesthetics. The many considerations in media philosophy of the digital and discrete as closed off from human experience, or of digital media as occurring on a pre-phenomenal level, makes it only more pressing to look for the presence of the computational and/or digital in the appearance. Taking everyday and ubiquitous AI generated images as examples, I trouble the question of the aisthetic appearance of digital image synthesis - as determination or indetermination, as formatting or mediation.

Layers of abstraction and discretization, rigidly determined by rules and measure, are the conditions of possibility for the appearance of synthetic images. In what sense is the hard determination and application of rules visible in the appearance? This is an objective process of mediation and determination then, but perception and experience is never objective, and the experience of an image appearance is never fully determined. This is a process and tension of determination and indeterminacy, that covers both the generation and the phenomenal appearance of the image output - the analog indeterminacy of the image vs the discrete determination of the generative process that produces the image. The adjustable rules for different text2image models are determinations of measurements and constraints, they are formatting rules.

As synthesis in multimodal models is both discrete (computational) and unifying (generating a statistical average), an aisthetic experience and appearance of the synthesis of generated images would be both analog (unified) and digital (discrete). If determination formats the technical media output, indeterminacy formats the experience of the appearance. How can this be seen in the image?

This essay analyses the interplay of indeterminacy and determinacy in the experience of images generated through text-to-image (T2I) models. Through an interdisciplinary approach, it uncovers three layers of indeterminacy: the computational indeterminacy inherent in text-to-image model processes, the indeterminacy of imagination in Husserl’s concept of protean phantasy, and finally the visual indeterminacy that figures in meaning making in all images. Generated images pass through these stages of indeterminacy, transforming indeterminate phantasy into determined visual objects, resulting in a conflict of consciousness between potential and actual. A distinction emerges between artificial phantasy, characterized by quasi-experience, and artificial imagination, grounded in images both as training data and perceptual image objects. As mediators between indeterminacy and determination, T2I images appear as technical media that mediate multiple forms of indeterminacy, showing the circulation between phantasy and imagination, between continuous and discrete. The generated image marks the limit of the unlimited indeterminate imagination.

The blockbusters of the digital present are iOS and Windows 11, TikTok, Instagram, and Fortnite. They constitute an aisthesis of habit, the sensible articulation of medial appearance. This paper is concerned with the appearance of computation in everyday visual media. Digital visual media are computational media, as generated and supported through processes of discretisation and determination that are often seen as nonsensuous and closed off from sensuous experience. In light of a wide set of media theories that posit the precognitive, the extra-phenomenal, the discrete, and the invisual, humans are nevertheless still apprehending media consciously – we are still seeing and experiencing visual media. Emmanuel Alloa’s medial phenomenology and Mark Hansen’s media entangled phenomenology are explored here as approaches towards the differential mediation of computationally grounded visual media – the appearance of the discrete. This entails a focus on the appearance of the middle, appearance as the locus of both aesthetics and aisthesis, and the possibility of a media aesthetic approach to the shaky distinctions between digital and analogue, discrete and continuous. Art is often seen as offering a critical distance to the order of life and society – but the images encountered by the average screen media user today offer no critical distance, rather they are a generic presence of the real and a flattening of the critical and unique aspects of art. What these images or visual media are is not necessarily poor images, rather mundane mediations. Everyday mediation is here exemplified through analyses of the smartphone homescreen, loading screens in apps and websites, as well as the everyday use of text-2-image generation models. These sites of mediation share formal expressions such as the grid, rounded edges and a flatness of resemblance and reproductions that approaches abstraction. Through these examples and the concept of the generic, this paper explores the ways in which a technical medium of display makes its own appearance visible, as a generic aisthesis. The computational process of discretization structures the continuous appearance of the mediation, a medial expression of nonsensuous computation and algorithmic governance.

Satellite images concretise several interwoven levels of ensembles and milieux, the purely technical of its capture, the networked technical of its dissemination and storage, Google Maps images are documentations of a specific place, but also documents the relations between certain technologies, mediums and formats, as well as specific ways of seeing and experiencing maps, landscapes, cityscapes, and images. The invisual technical processes in the background of these images create an artificial presence of the city in our image consciousness. (Wiesing 2011) The individuation of these images as objects relies on a process of both difference and unity (Hui 2016), as the medium and format of these images oscillates continuously between representation and function, aesthetics and operation. Google Maps is given to the viewer as an image of coherent unity, and present themselves as authoritative documentation of the landscape. But the images are composed of a differential variety of processes, files, and objects.  A relational ecological materialisation takes place which gives a new perspective on documents of the physical city, as both depictions and versions of the city in their own right. While documenting the city, satellite images intervenes in the materiality of the actual city as a media ecology.