A binary grid conceals two distinct forms in a single image

2x2 grid, AI infers topics, 16:9 Generate [CONCEPT_NEAR] and [CONCEPT_FAR] as a single hybrid-frequency image. Rules: - do not compose them side by side, blended, or as a collage - infer FORM_DNA for each independently: silhouette, mass distribution, rhythm - render [CONCEPT_NEAR] entirely in high-frequency information: fine line, small marks,   edge detail, texture - render [CONCEPT_FAR] entirely in low-frequency information: broad mass, coarse value   structure, soft tonal blocks - the two must occupy the same pixels; neither may be confined to a region of the frame - use CROSSOVER_DISTANCE to set the viewing range where one subject dies and the other is born - strictly limited palette, no mid-tones that would let both read at once - keep it clean, calibrated, and unsettling - the viewer at arm's length should see one subject, at ten paces the other, with a moment   of nothing in between - no hard-coded iconography, captions, or arrows unless supplied