Inside Dementia

The entire filming site is transplanted into the gallery: a FHD video loops on a 13-minute cycle, rewinding memory each time. Between two opposing 43-inch vertical displays sit the wheelchair, care chair, bed, camera, and lighting exactly as they were used on location.

I adored my grandmother, eating mandarins while nestled beneath the kotatsu — a low, quilt-covered table with a built-in heater that keeps one’s legs warm — as we watched Shōten (a long-running Japanese variety show beloved by older viewers) together—an exchange now dismissed by statistical models as an outlier. When my grandmother became bedridden, a kiss on her cheek drew a smile, yet the doctor said it was “only a reflex.”

AI maps events within a high-dimensional feature space, assigning them probabilities and distances, discarding anything beyond a threshold as “noise.” Extending that gaze, our own interpretations can flip just as easily. The recording equipment in the installation inverts seeing and being seen, pulling you inside the frame. In the instant you decide what image to accept as truth, whose voice is silently cut away?