A car screeches through busy downtown streets. Controlling a character who sits in the passenger seat, I look behind me and see a host of pursuing vehicles, their occupants climbing onto bonnets to leap in my direction with hostile intent. I take aim with a pistol, and a circular targeting pip flags up exactly when I should pump the trigger on my PS5 remote to blow out tyres and shoot the besuited enemies. The roof of the car I’m in is torn off during this combat, and now I must also contend with a helicopter launching a hail of rockets toward me. Upgrading to a larger weapon, I indiscriminately spray bullets everywhere in sight. The result is an immense explosion which seems to take out all adversaries and draw a line under this action sequence and section of gameplay.

At this point an onscreen logo – Unreal Engine 5 – comes and goes, and after this I float above the same city in which combat just took place (seemingly now clear of the explosions and rubble I so recently generated). Prompts invite me to ‘Toggle Mass AI Visualization’, ‘Toggle Day/Night’, or ‘Toggle Nanite Visualization’, while accompanying text states seemingly related concurrent information about ‘automatic open world streaming’, ‘dynamic global illumination and reflections’, ‘rule based object placement’, ‘particle system’, ‘virtualized geometry’, ‘procedural audio generation’, and more. Pressing x on the controller indeed toggles these values, making traffic appear and disappear, altering the time of the world from midday brightness to night-time halogen glow, or even rendering the entire, extremely verisimilitudinous world into a fractal, primary-coloured cavalcade of tes-sellated triangles (Figures C1a–b). After this aerial sweep, I am plunged back to street-level, where I control a character who can walk, run, and drive through this Western metropolis. No longer pursued by agents, possible activities to undertake are suggested by map markers (if I reach them I will gain access to more toggling options), but instrumental task completion does not seem to be the primary purpose of this world. Rather, these streets seem to be about wandering and languid aesthetic appreciation – I am invited to be awed at the size of this map, the realism of the physics, the scale of the AI, the quality of the audio, and so on.

‘Oh’ appears on screen eight times in eight different fonts, each one superimposed over the other, at the same time as the word is sung over and over, in a light and fruity vocal, by singer-songwriter Charli XCX. A tone in the background, like in inverted cymbal crash, grows and takes us into the start of the song proper, as Charli sings about the thrill of getting calls from her crush. Like a karaoke video, the lyrics continue to appear on screen at the same time as they are sung. But unlike karaoke, they manifest in a stagger-ing array of fonts, and they pulse and stretch to correspond with the aural qualities and inflections of their vocal articulation. At times they become completely illegible, a mass of neon globules and baffling superimposition. After the first chorus – in which Charli describes how her boy's unique ring-tone is the only one she knows – pop trio Kero Kero Bonito arrive into the song and ruminate on how a ringtone ‘could be anything’, even a ‘barking dog’ or a ‘revved engine’. The key, as they sing, is that the noise must capture the user's attention. Each line of this verse is plastered across the screen: the word ‘ringtone’ dominates in a chunky 1970s font, purple outlined with orange, beating like a heart with the rhythm of the song; subsequent lines appear over this, with ‘dog’ and ‘engine’ accompanied by low-resolution images of these things.

This is the lyric video for the song ‘ringtone (remix)’ by the experimental hyperpop duo 100 gecs (2020a). Available on YouTube, this extremely lo-fi video relies on a tight intermingling of image and sound – and little else. As a lyric video rather than a more traditional or official music video, it prioritises the words of the song, and seems to have very little production value (a single talented individual might have put it together in an afternoon using After Effects). But this does not mean it is sparse, simplistic, or that the relationship of the visuals to the music is purely that of semantic illustration. Rather, the bubble-gum baroque designs on display articulate the aural qualities of the voice, key, rhythm, timbre, and pitch of each particular moment of the song in peculiar but tightly bonded ways.

Deep inside the Chicago headquarters of technology company Kinetic Solutions Incorporated (KSI) is a top-secret military wing. During a tour of this facility, the company's supercilious CEO Joshua Joyce (Stanley Tucci) exhibits the ‘quantum leap’ that he believes will give KSI the edge on all global competitors. ‘It's the Holy Grail: Transformium’, an associate ( James Bachman) calls it, which Joyce jumps in to explain is a catchy, focus-grouped and trademarked moniker. What is so special about this substance? It is ‘programmable matter’, and now that KSI have ‘mapped’ its ‘genome’ they can control it as they see fit. To demonstrate this, Joyce lifts a small white ball from the table in front of him; the ball begins to float above his hand, before suddenly disassembling into a seemingly weightless, endlessly squirming cloud of small white cubes and intersecting geometric lines. ‘Sensual, almost’, Joyce comments as this cloud moves with sinuous elegance in the space between his hands. ‘Do you like music?’ he then asks, and the undulation of cubes suddenly coheres in his outstretched hand as a Beats Pill, a brand of portable Bluetooth speaker. ‘Perhaps something a little more violent?’ he prompts, and the substance morphs into a handgun.

This sequence takes place halfway through Transformers: Age of Extinction, a 2014 blockbuster. The subject of Kevin B. Lee's Premake video essay discussed in Chapter 1, the film is either a barrage of nonsensical and offensive spectacle or a trace of the ‘reinvention of cinema in the age of polymorphous and ubiquitous screen culture’, depending on who you ask (Koepnick 2018: 19). Indeed, it may well be both. This sequence, and the sub-plot it kick-starts involving an army of KSI-produced robot soldiers hijacked by the consciousness of franchise master-villain Megatron, indicates the film's own preoccupations with polymorphism. The enormous robot aliens of the series are heavy and machinic: they wreak palpable physical destruction and emit mechanical clunks and whirrs. But they are also light and digital, products of VFX work which, for all the perceptual realism in play, can seem disconnected from all physical surroundings. They shapeshift fluidly, impossibly, hypnotically, and fling themselves through space in inconceivable ways.

On an airfield in France in 1918, a group of desperate soldiers run through the open doors of a large aircraft hanger. As they emerge, the hanger begins to explode behind them, blasts of flame and debris flaring up in the doorway. But this exciting scene from a blockbuster Hollywood film is rudely interrupted – the image slows and then freezes, before detaching from the edges of our viewing screen. The now-mobile film frame swings to the side, so we view it as a parallelogram within an otherwise blank void. Multiple other parallelograms quickly manifest in a line in front of the original, each one containing different pieces of that first image. Before we have a chance to digest this carved-up series, the numerous slices re-converge into a single plane, and this plane swings back to once more occupy the entire frame of the video we are watching. The soldiers resume their running, albeit with an even greater fireball behind them, a fireball which we now understand – albeit only quite dimly and generally – to be at least partly the product of digital visual effects work.

Generated using programs like Autodesk 3ds Max, Maya, Adobe After Effects, and NUKE, the digital effects of contemporary cinema and television are most often outsourced to effects companies rather than made by the production personnel of a given film or show. Larger media productions will contract a range of companies to complete the many hundreds of visual effects shots, insertions, and clean-ups required by today's digital media ecology of spectacle. Described above is a brief moment from a short video by the effects and post-production studio Universal Production Partners (UPP), showcasing their work on the major blockbuster film Wonder Woman (2017). It is available on the company's website, where it is listed as a ‘Showreel’ (UPP n.d.), and also on YouTube, where it was re-posted in 2018 by the FilmIsNow Movie Bloopers and Extras account under the title ‘Wonder Woman – VFX Breakdown by UPP’. In the former context, it advertises UPP's technical capacities to prospective clients; in the latter, it is likely to be encountered by a more general audience (including fans of the film and those broadly interested in visual effects spectacle).

We get stuck to our screens for many reasons. Watching television, taking photos, creating albums, filling baskets, messaging friends, playing games, writing documents, mapping journeys: all can all be undertaken using the same screen in the space of a few minutes, or even simultaneously. This functional convergence is extremely convenient. The screen-based desktop interface becomes a one-stop shop, our go-to destination for a range of work and leisure tasks. It also makes this interface a crucial element of contemporary visual and aural culture, the centre of a gelatinous web bonding together various media.

This book is about the aesthetic cross-pollination that occurs as a result of this integrated screen culture. The graphic user interface, or GUI, is installed as standard on the vast majority of the world's desktop computers and smartphones. Whether using Windows, macOS, Android, or some other brand of technology, it is overwhelmingly a GUI of some sort which provides access to the applications which furnish us with our media environment. Through a GUI we can open a Chrome internet browser to watch Netflix; through a GUI we can access the Android app store to download Google Earth; through a GUI we can load up Steam to play a videogame; through a GUI we can use Maya to design buildings and even entire cities. The many tabs of an internet browser and the multiple programs scattered upon a typical desktop all testify to the multiplicity inherent in today's interfaces, the convergence they enact by bringing a range of media and functions together in a single 40-, 14-, or 4-inch screen. As a result, the GUI becomes gooey. The GUI sticks a multitude of media together, such that these media not only blend and interact with one another, but also increasingly adopt a fluid mass of aesthetic attributes that they share with the GUI itself.

For the last few decades, media scholars have been attentive to the ways in which consumer agency, corporate ownership, and industrial strategy have shifted thanks to digital production and distribution. Aesthetics have also been explored, but in less holistic ways. Several commentators have noted how, for instance, videogames import aesthetic qualities from cinema, and sometimes even seek critical validation through this process of appropriation (with ‘cinematic’ being synonymous in some videogaming cultures and scholarship with ‘good’).

We float above a sprawling yet strangely cramped metropolis. A dense collection of city blocks, factories, and leisure facilities stop suddenly at the edges of endless green fields, while a beautiful blue ocean runs along one side. Each of the buildings of this city has a specific, delineated function, which we can reveal by clicking on them. The residential skyscrapers – named things like ‘Wilkins Plaza’ and ‘Gomez Lofts’ – come with a caption and an emoji indicating how happy (or not) their occupants are. The population of ‘Pitts Place’ are displeased with their proximity to a factory (‘A factory near my home! Why!?’ they exclaim), and this is pulling down the overall satisfaction rating of the city. As unelected but all-powerful mayor, we resolve to do something about this, selecting the offending factory and dragging it several blocks away from the Pitts. Dropped into this new spot, it immediately resumes production of ‘seeds’ and ‘chemicals’, both of which will be ready in a couple of minutes. Pitts Place seems mollified, but to make our general populace even chirpier, we open up a menu of possible parks that can be built, dragging and dropping what is listed as a ‘Row of Trees’ into the gap vacated by the factory. Doing so temporarily turns the city into a mass of 3D vertical bars, each one visually telling us the level of satisfaction felt by every city block. Once dropped into position, the trees do their job, boosting all the adjacent happiness bars towards the sky, and adding to the overall contentment that the fictional populace feel about our mayoralty.

Just another day in Flexiville, the city I have created using the mobile videogame SimCity: BuildIt (2014). This game is just one of a host of cities that can be accessed using a GUI, but the malleability of its represented urban space is in many ways indicative of the kinds of newly flexible urban environments that can be imagined in the age of the desktop interface. Today's cities overlap with screens in many ways. Cities are sites of screen production, dissemination, reception and infrastructure. Images of their skylines and neighbourhoods circulate widely on screens in the form of films, television programs, videogames, apps, websites, and more.

I write this chapter using a MacBook Air. To accomplish this task, arranged on my computer desktop are the following: a Microsoft Word file containing a draft of the chapter; another Word file of notes; an internet browser with multiple tabs; a file management box, with lists of folders in which I can find further chapters, notes, and pdfs; and Spotify, a music player. I have listed these programs in their order of recession – that is, the Word file of the chapter is at the perceptual forefront (apart from the ever-present arrow of the cursor which I move around the desktop) and all the other programs are partly concealed, slid beneath both the chapter file and one another, such that sections of program windows peek out in an overlapping constellation on my 13.3-inch screen. I tell myself that this arrangement is convenient for switching between all these sources of information as and when the need arises: I can easily manipulate all the open programs to change their order and placement, bringing the file management window to the fore and moving the browser to the back, for example. The interleaved approach also allows me to keep a sneaky eye on my emails while engrossed in other tasks.

But this arrangement is by no means fixed. When the need arises, I can reposition all these windows both in space and in perceptual depth – for instance, I can bury the internet browser beneath everything else so that, while it is still present and open, I won't actually see if any new emails appear within its window. During such reshuffling, program windows and the onscreen cursor move fluidly, sliding around the screen with flowing clarity and laser-like precision. This rearranging can also reveal what lies behind all these open windows – namely, an image of Lone Pine Peak, a rock formation in California, captured in the early morning light. Bundled with the OS, this natural scene suggests even more depth – it is a distant landscape photo, after all – but the image itself signifies the very back and limit of the screen's navigable, overlapping space.

I am both at my desk and in a Jordanian refugee camp. An apparatus embraces the upper part of my head, slightly weighing down on my skull. I am able to visually explore the mediated space in which I find myself: a small, sparse, windowless room. The walls here have the sheen of light plastic material; slim mattresses are arranged on the floor; a CRT television is propped in a corner beneath a single shelf; a clothes hamper and a couple of mirrors are the only other furniture. I hear the voice of a young girl and look around to find the source – she sits on the floor, talking to me in a language I don't understand. Her voice is tinny and echoes, a marker of inexpensive audio recording equipment and the poor acoustics of this prefabricated space. But an English translation of her words begins speaking above this in the mix, and this other voice is much closer and richer, reverberating around my skull thanks to the in-ear headphones that accompany my PS4 virtual reality rig.

Through this voice-over, the girl introduces herself as Sidra. She tells me that she is from Syria, but has lived here in the Zaatari camp in Jordan for the last year and a half. I am at her eye-level, seemingly sitting across from her, sharing space on the Persian rug. As I settle in to hear more, my vision darkens – a fade through black takes me to the same room, but later, and I presume that the other children of various ages suddenly in this space are the family Sidra is currently describing. Unlike Sidra in those earlier moments, her siblings gather themselves to leave without acknowledging my presence at all, a snub which seems aligned to my new position, which is higher than them, looking down (probably the height of an average standing adult). Over the following seven minutes I am granted a tour of the refugee camp, oriented by the translated voice-over. Routines of school, cooking, sports, videogaming, fitness, and eating are shown in such a way that I am placed firmly at their centre, albeit for the most part as an unacknowledged observer. At the end of this tour, Sidra comments that she feels she has been here long enough, and would like to leave.

Accepting a prompt to restore information on the device I’m holding, I swipe right, accessing the phone's home screen. The intelligent assistant suggests that I look through the available programs to identify the location of the phone's owner, Sara. I scroll through the text messages, read the emails, and scour the photos that belong to this stranger. In doing so, I encounter dissatisfied mothers, quirky friends, and dumped boyfriends. This all seems fairly innocuous, apart from some casual mentions of witchcraft, midnight meetings, and deadly viral media (and then there's the Siri-like assistant, which seems oddly pushy …). Suddenly, my investigations are interrupted as the phone rings and, without thinking, I answer. A muffled, desperate voice asks if I’m Sara. Before I can reply, the call cuts off. A few minutes later, after more investigative work and coaching from the phone itself, I find myself having to choose which of two strangers is killed. The predictive text I’m using to communicate my decision only allows me to state one of their names – there is no option to protest.

This is Sara is Missing (2016), an immersive sim for Mac, Android, and Windows. If played on a smartphone, the game takes over the entirety of the user's screen, and mimics many of the operations and expectations of this kind of interface. Answering a call in-game requires the exact same swipe on the same screen that the player would undertake to answer a call coming in from their actual life (which could well also occur as they play). The result is a videogame that finds deeper leverage in the player's reality than might be expected given its more outlandish elements, leverage achieved precisely because the game's diegesis overlaps with, or rather consists openly and entirely of a specific interface, the one of the device on which it is being played. That the game can mimic a smartphone screen so successfully, and yet still function as a videogame, hints at the extent to which computing interfaces overlap with other media.

As discussed in the introduction, the screens we use to watch media, do our jobs, and plan and enact our social lives are in many cases not discrete, single-use devices. Whether a phone, laptop, tablet, or even a television, screens are multipurpose, whatever their size.

What is the relationship between digital media and the material world, and how might our use of the former change our understanding of and place within the latter? A central concern of digital media theory from the late- 1990s onwards, the ethical aspects of this technological and (potentially) ontological change in relation to film is summed up by Markos Hadjioannou in From Light to Byte: Towards an Ethics of Digital Cinema (2012). As he describes, analogue cinema's indexicality ‘enables a realization on the part of the spectator of her or his existential position within the world and so qualifies an ethical implication in the image as the potential for responding to, and acting in, the world’ (2012: 177). Digital media, by contrast, offer not indexical trace but interactive, present-tense simulation, and so, for Hadjioannou, ‘cannot conjure up an image of the world as an existential guarantee’ (2012: 177). While assertions around the radical differences between analogue and digital capture have been challenged (see, for instance, Cubitt 2011), it is certainly the case that digital images and the networks that produce them offer a new kind of visual modality. In a digital age, vision becomes a method of filtering, rather than encountering, and the image becomes interactive and navigable (Manovich 2001; Verhoeff 2012). Images are no longer only observed, they are swiped, clicked, tagged, filtered, and shared. If they provide evidentiary proof, it is less as an authentication of a past moment than a trace of a flexible present, a conditional marker of temporary conditions.

All of this alters our ethical relationship to images. The new digital visuality is not principally voyeuristic or scopophilic but rather cartographic, defined by an informatic mode of automated recording and registering that seeks legibility and apparent usefulness. This is not to say that this is entirely novel: the digital's ‘cartographic gaze’ of remote presence certainly inherits impulses to catalogue and quantify familiar from colonial visual modalities – image-making and image-circulation remain methods for distant control and ideological persuasion (Specht and Feigenbaum 2019). Through interactive digital images, the perception of the user is extended beyond their temporalspatial confines. This is achieved through the digital generation of an algorithmic world, one which is less concerned with aesthetic appreciation than it is with bodily immersion and virtual presence.