The algorithmic interpretation of everyday images in this project becomes the basis for creating atypical, lively typographic forms that arise not through strict construction, but through the process of distortion, malfunction and subsequent visual assembly. Each glyph is the result of the interaction of photography, neural network, and 3D processing - not a literal reproduction, but an abstract reconstruction. It becomes a sign in which the visual memory of the original image is preserved in a recycled, multilayered form.
The project operates on the boundary between control and rejection: the author sets parameters, selects images, edits the result, but the final form turns out to be unpredictable. The neural network here is not a tool, but a co–author, whose "thinking" forms a new morphology of the sign.
The created glyphs are not static letters, but dynamic structures capable of interpretation, transformation and updating. They form the alphabet, a process in which typography is perceived as a visual system living at the intersection of technology, mistakes, and intentions.