A six-fingered hand appears on the screen. Its texture looks realistic, and the lighting suggests authenticity, with the skin tone maintained through careful calculation. However, something feels off: the gesture goes beyond anatomical limits, the finger curves are disproportionate, and the thumb bends unnaturally. This figure, created by an algorithmic model, bears the mark of an age-old obsession — the desire to turn the relationship between body and intention into an image. The mistake symbolizes a civilizational project that reduced human imitation to a matter of technical precision.

Since Paleolithic caves where pigment was applied around the palms, the hand has been central to the origin of imagery. Each mark showed direct contact between body and surface, a transfer of matter that linked gesture with pigment. This connection continues into the Renaissance, when the hand symbolized the boundary between thought and matter. Leonardo depicted its joints with surgical precision and theological devotion, while Dürer elevated hands in prayer to the level of spiritual portraiture. In both cases, gestures captured the essence of what the eye perceived and what the body expressed. Even in highly detailed anatomical drawings, a subtle dissonance remained— a gap between sight and touch. Hands hold a secret of movement that no image can entirely encapsulate.

The gap between cave paintings made by hand and algorithm-generated images lessens when we see both as inscriptions lacking a present body. Pigment was brushed onto stone, much like pixels are now projected onto screens: in both cases, the gesture leaves a mark without substance. This ongoing trace creates an archaeology of touch, where the surface captures motion even when the body is absent. Breath and light serve the same purpose — to capture the fleeting.

Centuries later, artificial intelligences face a similar challenge. They produce perfect faces, realistic landscapes, and fabrics with a nearly tangible shine—yet struggle with hands. They falter where humans also show uncertainty. This coincidence feels too precise to be mere chance. Designed as a mirror of humanity, generative AI is a machine trained to imitate our gaze, style, hesitation, and flaws. The imperfection of its hands thus faithfully mirrors our own: beings driven by imperfect intent and craftsmen of interrupted gestures.

The algorithmic image reveals this mirror with stark clarity. Every extra finger, every shifted joint highlights the gap between calculation and actual experience. The model is unaware of the body’s weight on others, the resistance of skin, or the tension before contact. Its hand is purely statistical—a gesture simulation lacking memory of the act. Yet, this very error embodies the core cultural aim of generative machines: transforming humans into patterns, converting gestures into data, and replicating the body to the point where its original meaning is lost.

The mirror reflects both what we want to see and what we overlook. The distorted AI hand vividly symbolizes our technocultural goal: to replicate the human in its glory and imperfections. Here, error gains symbolic significance, serving as the involuntary mark of an era where the creator and the simulator are indistinguishable. Humans craft to comprehend their actions; machines produce to imitate them. Between these two lies the space of contact.

*

Since the beginning of visual art, the hand has held a complex role in representation. It symbolizes action, ownership, and promise, embodying gesture, craftsmanship, and authorship. Early artists depicted hands on cave walls, establishing a dual perspective of presence and absence. The hand’s imprint on the rock is both a physical form and a shadow, a gesture and a memory. Art has since persisted in that space between touch and its mark.

Leonardo da Vinci aimed to view the hand as a device of thought. In his notebooks, each tendon and bone was rendered with detailed anatomical accuracy and a mystical quality. To him, the human body reflected a divine design; the hands served as the tools of that divine correspondence. However, his meticulousness carried a quiet unease. No drawing could capture the fleeting moment of a gesture — the hesitation between impulse and touch. Leonardo’s hand was too idealized to truly connect with the world.

A few years later, Albrecht Dürer depicted two hands in prayer. They lack bodies, faces, and objects, simply expressing emptiness. This moment of suspension embodies the symbol’s strength. The praying hands halt the gesture, showcasing human fragility in the face of the invisible. The drawing still resonates with purpose, as if each finger symbolized the effort to represent what lies beyond will.

Velázquez, Rembrandt, Caravaggio, Titian, and Goya all encountered the same challenge. Hands often drew disproportionate attention, creating instability within the composition. Many artists concealed them with fabric, gloves, or shadows, while others accentuated them to add drama. The history of painting reflects this ongoing struggle with hands: they are too intricate to be merely decorative and too expressive to ignore.

Rodin recognized tension and sculpted it into form. In his Mains d’artistes, hands detach from the body, becoming independent. Just a pair of entwined fingers can hint at a vast network of relationships. Giacometti expanded on this idea: in his elongated figures, hands seem fragile, extending from a body in decay — gestures seeking substance. Both artists share the understanding that the hand embodies the full complexity of human drama.

Paula Rego openly shared in interviews that she never quite mastered painting hands, perhaps because hands tend to become overly human in her art. Nonetheless, the hands in her works, from *The Policeman’s Daughter* to the Abortion Series, capture the physical and emotional tension of her characters, showcasing the body’s strength through hesitant gestures. Francis Bacon also focused on a different part of the body. When speaking with David Sylvester, he discussed how difficult it is to paint mouths, calling them living, risky features that are prone to deformation. Goya approached technical challenges differently by turning them into symbolic elements: during a portrait negotiations with clergy, he asked if they wanted hands included, warning it would cost more because hands require time, precision, and personality. Even animated movies reflect this awareness of complexity—characters are given four fingers and white gloves (a Disney convention) to simplify movement and make gestures clearer. From Paula Rego to Mickey Mouse, the hand remains a bridge between the viewer’s gaze and what the image can still reveal.

The history of art reveals a lineage of sublime failures. Every effort to depict a hand highlights the challenge of turning movement into form. The hand acts as both tool and symbol, requiring the artist to seize the moment when thought turns into matter. Each era offered its own partial fix, visibly showing traces of hesitation and the imprint of gesture. Mistakes gain expressive significance, becoming the space where movement leaves its mark and creation emerges.

Jorge Ramos do Ó stated that “making the hand” involves learning physically before understanding intellectually. The gestures that draw, sculpt, or write serve as both a teaching method and a creation process. Here, the hand acts as both workshop and thinking space, where knowledge is reinforced through repetition. This approach to gesture has characterized art throughout history and continues into modern techniques, where making becomes an action and the body is driven by calculation. The title I share with Ramos do Ó reflects this belief: to understand the hand is to grasp how the world is created.

The invention of photography introduced a new way of representing visions, where images are created by light and expand human reach. The camera acts as an extension of the gaze, sharing some of the creative process with the device. Gestures transform into framing, and touching becomes the click of a shutter. Photographic mistakes—such as blur, overexposure, or unexpected cuts—highlight the machine’s independence and its interpretation of reality. The camera captures moments the human eye might miss, turning light into tangible evidence. This shift signifies moving from human error to technical error, with the body becoming a distant operator in the process. AI builds on this visual legacy, extending it further until the line between observer and creator blurs.

*

Generative artificial intelligences function within a realm that merges computation and statistical creativity. They generate images by drawing from extensive collections of visual data, learning to identify connections among shapes, colors, textures, and contexts. Their approach is cumulative: they absorb millions of examples, extract correlations, and blend patterns until a new synthesis appears. This process results in an almost instant sense of realism — a fabricated appearance of reality created through repeated patterns.

The error present within these images shares the same fate as human expression. Each image reflects the gap between intent and result. AI reproduces this gap with mechanical accuracy, turning what has always been an artistic experience into a pattern. The algorithm’s failure is no different from a hand’s failure: both expose the challenge of matching form with intention. What differs is how the error is embedded — once in the body, now in the calculation.

In this process, the human body is depicted as one of the most common yet inconsistent references. Hands, in particular, often appear fragmented, hidden, deformed, or merged with objects. The model learns from data that depict gestures without sequence, angles without context, and joints out of proportion. Each generated hand results from thousands of partial examples reconstructed by an algorithm that does not perceive the body as an integrated system.

The algorithm fails to understand the balance among shoulders, wrists, and fingers; it overlooks gravity, muscular tension, and the friction between skin and surface. Its understanding is solely visual and based on experience. The hands it generates come from an intangible understanding—a knowledge lacking tactile sensation. Each anatomical mistake reflects the gap between calculation and actual presence.

The technical issue becomes an ontological question. AI produces images by mimicking humans without having a body. Its functioning relies on imitation, substituting direct experience with repeated patterns. Unlike the artist who feels the resistance of matter, the machine simply replicates correlations. Its learning is based on statistics; its output is synthetic. For AI, the human body is seen as a visual pattern lacking sensory depth.

Because of this, algorithmic hands break down with almost poetic consistency. They are mistakes stemming from too much similarity. AI imitates the human until the form becomes overwhelming, causing the gesture to lose significance. Every extra finger, every impossible curve, reveals a lack of purpose. The machine sketches the hand as if someone repeats a word without understanding its sound.

The mistakes seen in generative images show how the technical gaze perceives the world — based on likelihood rather than firsthand experience. The model understands that typically a hand has five fingers, but it also recognizes that sometimes a sixth finger may appear due to shadows, perspective, or overlap. Consequently, the chances of error are incorporated into the model’s assessment of realism. The machine reproduces these errors because it has learned that they can occur.

AI, therefore, does not truly make mistakes; instead, it replicates human ambiguity. Its imperfect execution demonstrates the high fidelity of imitation. The machine reflects the human—in a form without a body—a mirror that shows both shape and flaw. Its errors become our portrait.

*

Throughout history, error has never been a matter of chance. It is an essential component of practice — representing action and presence. In drawing, errors indicate hesitation between impulse and form. In creation, they mark moments when the artist’s body learns to control material. Each mistake reflects a decision, showing a deviation that reveals the creator’s hand. When visible, errors affirm that an object or work was crafted by someone.

AI replicates this process in reverse. An error arises from an action lacking a body, characterized by continuous repetition that omits hesitation. Algorithmic failure occurs due to too many correspondences, an overload of patterns. The result visually resembles human error — and this is where the mirror forms. The machine’s mistake essentially mirrors human hesitation, pushing imitation towards the point where artificiality blends with originality.

The artist makes mistakes because the body pushes back, while the machine fails due to the absence of the body. Both demonstrate the same tension: the link between intention and execution. In one, gesture faces material resistance; in the other, calculation substitutes gesture with probability. In both cases, the flaw highlights the gap between the desire for mastery and the act of creation.

The technical culture supporting generative AIs embodies a mirror ambition: to replicate humans with utmost likeness and detail. This replication includes imperfections, as realism is often conveyed through error. The model gains credibility by adding minor distortions that enhance believability. For example, deformed hands illustrate this fidelity, showcasing the machine’s ability to mimic human asymmetries.

Vilém Flusser, in *Gestures*, viewed gesture as material thought—a form of embodied knowledge held in the body. He considered gesture a pre-verbal language that helps us understand the world through action. When AI tries to imitate gesture, it reduces this language to simple calculations. As a result, it loses the experiential quality, retaining only the outline. Ultimately, gesture becomes statistical data. Boris Groys, in *In the Flow*, sees the digital realm as a space of continuous circulation where authorship fades and images endlessly extend. Algorithmic creation functions within this flow, constantly reconfiguring itself. Each error is seen as a variation, and every deformation opens up aesthetic possibilities. Error takes on the value of a pattern and becomes a stable form amid the endless movement of images.

AI acts as the first mirror guided by the absence of intention, reflecting humans with mechanical precision and producing a reflection based solely on operation. In this framework, errors become meaningful as synthesis, reaffirming the program’s coherence. The hand with extra fingers symbolizes the algorithm’s completeness — its ultimate effort to mimic humans in all aspects, even the most unlikely. Algorithmic errors embody complete fidelity, reaching a point where resemblance is maximized. The machine fulfills its cultural role to mirror every detail of humans, transforming incoherence into form. The digital mirror presents an image of humanity where error signifies authenticity.

*

The hand symbolizes authorship. Through its gestures, thoughts gain significance; in the fingerprint, the body claims its identity. The signature, mark, touch, and act of creation all pass through the hand. It connects intention with matter, linking the subject and the world. Thus, it is the point where creation transforms into responsibility.

From traditional craftsmanship to modern design, the hand bridges the gap between idea and execution. Every stroke of ink, every flaw in a cut, and every irregular line reflect a connection with material. The work retains the essence of the body’s involvement during creation. The hand’s presence also introduces error, variation, and unique choices that cannot be replicated. Touch emphasizes the individuality of each gesture.

Generative AI creates a new ongoing connection based on correlation. The images generated come from a collective process where authorship broadens and shifts. Each creation reflects the merging of many intentions within a computational system. Its ‘hand’ appears as a pattern and acts as a combinatorial matrix. When human gestures are translated into computation, they take on a different temporal rhythm. The act of creation shifts from being a moment of making to a process of continuous processing, resulting in a seamless experience of ongoing and cumulative creation.

Federico Campagna, in “Technic and Magic,” portrays technique as a system transforming reality into functionality. In this view, the hand moves away from direct touch and acts as an operator. Gesture turns into an interface. This shift is embodied by AI: the body merges into operation, and the world is structured as a data system. Consequently, the hand’s importance diminishes, and touch is replaced by simulation.

Yet, even without a hand, the craving for presence persists. Each created image echoes the lost touch, with AI’s distorted hands symbolizing longing. These deformities serve as visual memories of an absent body, remnants of authorship the algorithm cannot fully claim. The machine mimics the gesture, yet a gesture involves a decision — a choice that belongs to the human actor.

In design, this distinction holds ethical significance. “Total design,” focused on efficiency and automation, aims to eliminate flaws and gestures, envisioning a frictionless, fully calculable world. Conversely, “design of transitions” recognizes error as a space for relation and learning, emphasizing contact, adaptation, and unpredictability.

In this context, the hand serves as a metaphor for a practice that emphasizes the body as a mediator. Error marks the moment when intention turns into knowledge. The human hand, through failure, gains understanding; the algorithmic hand, through repetition, loses memory. This distinction outlines the boundary of modern creation — between genuine gesture and simulated gesture, between embodied authorship and data-driven authorship.

The clash between total design and transition design reflects two perspectives on the modern technical system. Total design aims to contain gestures within predictable functions, while transition design explores the tension between body and code, viewing error as a form of practical knowledge. Both approaches represent different ways of understanding the relationship between gesture and system — one emphasizes stability, the other focuses on adaptation.

The hand signifies presence through action. AI, by mimicking it, exposes what it misses—the moment when decision flows through matter. Every extra finger symbolizes an ongoing absence. The digital mirror reflects our desire to erase the body, and in doing so, the body reappears—often as error.

*

The image of the flawed hand remains on the screen like an involuntary oracle. What unsettles us is not the defect itself but the echo it triggers. We see the error because we see ourselves in it. The algorithmic distortion echoes the vulnerability of human gestures — the gap between intention and action. Each additional finger is a visual confession that calculation also seeks to touch.

Generative AI embodies the idea of a perfect mirror. It was designed to replicate humans with mathematical accuracy—an accuracy that includes occasional failures. These imperfections are evidence of the mirror’s perfection: it reflects us so vividly that it causes vertigo, turning the reflection into a reflection of our own desire to see ourselves. The errors in those hands do not indicate machine failure but rather evoke the longing of the body that created them.

David Graeber, in The Utopia of Rules, described bureaucracy as the most sophisticated form of the death of gesture. Similarly, algorithms aim to organize the world through repetition, removing chance and neutralizing time. The work produced by AIs reflects the hands of technical bureaucracy — precise, disciplined, yet flawed. They represent an ideal of control that ultimately falters when trying to imitate the body.

The created image exists in a tense balance: it aims to seem alive, but to achieve this, it must include error. The flaw turns into decoration, disguise, or a form of artificial authenticity. Here, the machine attains its greatest similarity to humans. Error transforms into expression. The mirror completes its cycle, reflecting the gesture until no further difference can be expressed.

The human, standing before that mirror, finds himself unrecognizable. The AI’s hand becomes a symbolic extension of our desire to imitate life. Each generated image is a gift to the illusion of control; every flaw, a reminder that matter resists transparency. Touch, when it vanishes, returns in spectral form — a hand of light trying to grasp the weight of the world.

The mistake involving hands lies not with the machine or the human alone, but in the shared space between them. This is where gesture turns into a mirror, and the mirror into gesture. When AI fails, it exposes a forgotten truth: that creation involves hesitation, touching demands vulnerability, and the body is the fundamental measure of every image.

For centuries, the failures in representation remained unacknowledged. Artists considered the hand a challenge, but never used it as an argument. Errors were tolerated, never explicitly pointed out. With the advent of algorithmic imagery, this dynamic changes. Flaws become public, measurable, and shared. The six-fingered hand circulates as both a story and a sign of systemic issues — scandal and diagnosis combined. Error no longer shame its creator; instead, it highlights the system’s inherent logic. For the first time, failure enters the conversation about representation — as discourse and humor — reflecting a body that calculation tries to emulate but can never fully replicate.

The deformed hand perfectly symbolizes our era: it exemplifies programmed imperfection, serving as a reflection that constantly strives for authenticity. The mirror has already mastered mimicking the gesture and keeps attempting to recall the body.

Bibliography

Bacon, F., & Sylvester, D. (1987). Interviews with Francis Bacon: 1962–1979. Thames & Hudson.

Campagna, F. (2018). Technic and Magic: The Reconstruction of Reality. Bloomsbury Academic.

Da Vinci, L. (1970). The Notebooks of Leonardo da Vinci (I. A. Richter, Ed.). Oxford University Press.  

Dürer, A. (1957). The Complete Woodcuts of Albrecht Dürer. Dover Publications.

Flusser, V. (2014). Gestures. University of Minnesota Press.

Goya, F. (1989). Los Caprichos (facsimile ed.). Dover Publications.

Graeber, D. (2015). The Utopia of Rules: On Technology, Stupidity, and the Secret Joys of Bureaucracy. Melville House.

Groys, B. (2016). In the Flow. Verso Books.  

Ramos do Ó, J. (2019). Fazer a Mão: por uma escrita inventiva na universidade. Edições do Saguão.

Rego, P. (2003). Paula Rego (F. Feaver, Ed.). Phaidon Press.

Rodin, A. (1981). Les Mains d’Artistes. Musée Rodin.