The proliferation of automated visual systems has been transforming not only how we produce images but especially how we perceive the act of seeing. These processes do not function merely as tools but displace the subject’s position in relation to the image. We no longer look from a sovereign place: the gaze becomes mediated by invisible architectures that interfere, suggest, and sometimes decide what is visible.
The image of a plush orangutan, seated as if observing, becomes the center of a sequence digitally expanded. The figures that emerge—partial bodies, suspended gestures—never existed. They were inferred by a visual generation system, without any textual prompt. And yet, they seem to belong in the scene. The software does not capture: it projects. It assumes. And with that, a sense of strangeness emerges—not from error, but from the suggestion of a logic that operates outside human intention.
Throughout this process, the essay proposes an aesthetic reflection: what happens to the image when the human no longer holds exclusive control? As Mitchell (2005) observes, images are not passive; they act, affect, and react. When formulated through computational means, this power does not vanish—it is simply redistributed.
Benjamin (2008) anticipated that technical reproduction would dissipate the aura of the work. In the case of these figures, there is no aura to lose. The image is already born as a copy, but with enough density to be perceived as real. The visual pact is not broken—only displaced.
In some of the compositions, shadows and gestures emerge whose arrangement provokes discomfort beyond the visual—almost symbolic. The hands resemble insinuations, the contours hover between the banal and the involuntary. None of this was asked for, yet all of it was generated. The machine invents, but it is not neutral. What appears reveals not just statistical algorithms, but echoes of cultural patterns. As if, in inferring the world, the visual generator also exposed what we prefer not to see—the imagetic residues that already inhabit our collective imagination.
Simondon (2005) proposes that the individual is not substance, but process. The image is also that: a field of machinic and cultural individuation. Authorship dissolves, giving way to algorithmic curatorship—a gesture of selection more than of creation.
Sontag (2003) reminds us that images both move and desensitize. These automated figures oscillate between affect and numbness. But it is not the algorithm that deceives—it is our reading that falters.
This essay does not aim to resolve the tense field these images inhabit but to sustain the gaze before what manifests beyond our intention. It is not about a new aesthetic in formal terms, but a shift in how we begin to recognize—or distrust—the image. Computational generation does not inaugurate a distinct visual regime; it merely intensifies the noise that already existed in the production of the visible. What disturbs us is not the machine’s presence, but its reflection upon us. What we see, even without human authorship, reveals less about the technology and more about the patterns that preceded and continue to shape it. It is not the image that lies—it is we who have forgotten how to read it.
This visual essay stems from an artistic practice focused on images automatically generated from an original composition featuring a plush orangutan. Using visual expansion processes—without textual commands or explicit direction—unexpected figures emerge: fragmented bodies, shadows, and gestures that suggest human presence, yet do not originate from any photographed reality. This intervention does not aim to demonstrate a technique but to create a critical space for reading what is generated outside the author’s intention.
More than a digital manipulation exercise, it is about making visible what escapes: automatisms, glitches, and symbolic residues that reveal another layer of visuality. Rather than correcting the image, this essay chooses error as material. The artist does not determine the content but selects and reacts to deviation—constructing meaning from what resists predictability. Here, the image does not represent: it challenges.
The reflection shifts from technical control to the act of seeing as an ethical and aesthetic experience. If the image no longer depends on the human gesture to be formed, what role is left for the gaze? Far from offering a conclusion, the essay leaves open the suspicion that perhaps it is not the image itself that unsettles us most, but the way it keeps appearing—even after we’ve stopped seeking it.
