Your browser does not meet the minimum requirements to view this website. The browsers listed below are compatible. If you do not have any of these browsers, click on the icon to download the desired browser.
The history of human-computer interaction is a history of an ongoing attempt to bridge the "inter," the gap, between human action and machine response.
In 1982, the computer scientist Ben Shneiderman describes how the direct relationship between a user's action and the displayed effect on the computer screen in graphical user interfaces (GUI), transform an initial "hostility" towards the computer into a "glowing enthusiasm" and a "thrilling sense of power."1 In this new paradigm of "direct manipulation," users can directly interact with on-screen objects through instantaneous, incremental, and reversible actions, replacing complex command language syntax. For example, in display editors, full-page visibility, labelled buttons, and a blinking cursor provide clear, dynamic indicators of where the user's actions will take effect.
Media theorist Wendy Hui Kyong Chun describes this directness as a "compensatory gesture" that gives users a pleasurable sense of mastery while concealing the machine’s complexity. She argues, that this experience of control is a "carefully crafted simulation, deliberately programmed into the interface, rather than genuine mastery."2
See the Pen The Magic Orb 1 by CharlotteLen (@CharlotteLen) on CodePen.
Today, our interaction with so-called artificial intelligence (AI) brings us full circle back to the written word. In leading generative AI tools and Large Language Models (LLM) like Midjourney and ChatGPT, the text field becomes the primary interface between human action and machine response. This new paradigm challenges users to express their intentions through description rather than direct manipulation. Instead of guiding the computer through incremental steps to gradually arrive at a desired result, users now prompt an outcome without engaging in or getting insights into the process of its creation. With AI, the user skips to the end.
At the same time, this bigger leap from input to output confronts us anew with the fundamental challenge of bridging the gap between human and machine. The relationship between prompt and output is far from straightforward. The chat-like interface creates an illusion of dialogue, while users engage in writing gymnastics of perpetually reimagining, adjusting and refining their prompts to yield a machine response that matches their expectations.
AI interfaces respond to this widening gap with new compensatory gestures. The “inter”, in interaction, becomes the space of the spinning orbs, magic wands, and purple sparkles.3 The opaque realm between prompt and result is obscured and mystified behind a veil of magical iconography; a welcome alternative when a reference to human intelligence seems too dull. It is little surprising that the term AI is often evoked like a spell, suggesting technological magic beyond human understanding and intelligence. The magic wand becomes a way to suspend our disbelief; to transform an initial “hostility” into a “glowing enthusiasm,” to say it in Shneiderman’s words.4
AI therefore, brings the compensatory gesture of human computer interfaces to another level. Every click on the send/upload button in AI interfaces is a leap of faith into a magical unknown. Only the blinking cursor remains as a vestige of human agency, of our ability to anticipate, to navigate, to control.
See the Pen The Magic Orb 2 by CharlotteLen (@CharlotteLen) on CodePen.
While today’s user interfaces seek to minimise, compensate for and even mystify the gap between human input and machine output, the absurdly complex machines of both William Heath Robinson and Rube Goldberg deliberately stretched and emphasised the ‘inter’ in elaborate and often times comical ways. In the early 20th century, their illustrations of intricate contraptions were humorous parodies of the efficiency and mechanisation mania of the time, which promised to simplify everyday life—an obsession that has reached new heights in today’s era of machine intelligence.
While Robinson and Goldberg’s devices were designed to perform simple tasks such as eating peas or operating a napkin, they did so through ridiculously convoluted networks of winches and pulleys, and in Goldberg’s drawings sometimes even incorporated unpredictable elements such as animal behaviour.5 The adjective, “Heath Robinson” even made it into the Cambridge dictionary, ‘to describe a machine that is very cleverly made and is complicated in a silly or humorous way, but has no practical use’. Similarly, the term “Rube Goldberg” has become synonymous in American English with performing simple actions in an unnecessarily complex or laborious way. Robinson and Goldberg's machines make the process not only elaborate but also visible and transparent. They involve the viewer in the process, inviting them to follow the sequence of steps with their eyes, to consider their plausibility and to savour the humour in each ridiculously elaborate solution to a simple problem.
See the Pen The Magic Orb 3 by CharlotteLen (@CharlotteLen) on CodePen.
As the waiter uses a spoon to place the cooked peas into the heated tank at one end of the table, they roll down the chute and into the funnel, which tips each pea onto the conveyor belt. By turning the crank with his right hand, the restaurant guest, sitting on the other side of the table, turns the drum, which activates the conveyor belt that takes each pea up one at a time and drops them onto a spoon mounted on a hinge. By pulling on the chain with his left hand, he tilts the spoon forward, releasing the peas directly into his mouth.
Robinson's illustration of the pea-eating machine as described above, does not simplify or replace the act of eating, it merely changes it. The machine exhibits a paradoxical relationship to automation: while ostensibly automating a basic task, it complicates the act and requires more human involvement, not less. Eating peas becomes an action that requires both hands: One to pull down a spoon mechanism, the other to turn a handle. And yet the input and output remain the same: a spoonful of cooked peas in. A spoonful of cooked peas out.
The process is not a means to an end—it does not change or add to the product. The "inter" here is not primarily about efficiency but speaks to a desire to do something simple the hard way, to do things differently, to challenge repetitive ways of doing things, to take detours. In Robinson and Goldberg's illustrations, the space between human and machine is not a gap to be compensated for, but a space to be explored, celebrated, and even deliberately expanded. Their contraptions offer a different view of human-machine interaction, one that embraces complexity and process, and engages users in the choreographic potential of the “inter”.
See the Pen The Magic Orb 4 by CharlotteLen (@CharlotteLen) on CodePen.
The practice-based element of this paper is an interactive online programme that deliberately extends and explores the “inter” in human-computer interaction through a series of experimental text interfaces. Visually mimicking a chat window, the interface transforms the simple act of typing into an elaborate choreography.
The programme consists of seven individual interactive interfaces written in HTML and JavaScript, each providing a different context that prolongs the act of typing into an extended sequence of interactions. These district interfaces are interconnected, with each send/upload button leading not to the transmission of the message but to a new iteration of the action, keeping the user suspended in the “inter”. Each interface challenges the habitual act of typing through simple interactive elements—buttons, checkboxes, mouse drags and clicks.
When a user presses a key, they have to navigate through deliberately complex interactions. In one programme, pressing a key randomly generates a series of additional characters that must be deleted before proceeding. In another, the user encounters a series of pop-up windows—literal interruptions into the space between—each requiring specific manipulation before a single character is displayed on the screen. In yet another, the letters appear in random positions and must be rearranged by dragging and dropping them with the mouse. Through these often absurd steps, the user becomes immersed in what would otherwise be a simple keystroke, attending to the usually bridged gap between human input and machine output.
This deliberate complication subverts the compensatory gestures of today's interactive interfaces. Rather than obscuring complexity behind a magic orb, the programme celebrates the gap between human and machine as a space for exploration.
See the Pen The Magic Orb 5 by CharlotteLen (@CharlotteLen) on CodePen.
See the Pen The Magic Orb 6 by CharlotteLen (@CharlotteLen) on CodePen.
See the Pen The Magic Orb 7 by CharlotteLen (@CharlotteLen) on CodePen.
Your browser does not meet the minimum requirements to view this website. The browsers listed below are compatible. If you do not have any of these browsers, click on the icon to download the desired browser.
