This AI-generated video explores the language's performativity notion, and the loss (or addition) of information between transmitter and receiver, what Shannon & Weaver call noise. This destructive noise is reminiscent of the very workings of image-generating AIs - which add and remove noise to create - in which the Generator and Discriminator are 2 entities dialoguing with each other to arrive at the final result, one being the transmitter and the other the receiver. The same goes for the human-artist (transmitter) dialoguing with the AI (receiver) through the prompt, for the AI (transmitter) dialoguing with the human-spectator (receiver) through the final result, just as the human-artist (transmitter) is dialoguing with the human-spectator (receiver) through this symbolic handshake, whose textual hallucinations - haunted by corporate watermarks - are merely the result of the noise introduced, literally and figuratively.

And everything I just said here is exactly like this paratextual hallucination : lost in translation.