אור ('or) Translate investigates the creative possibilities of noise and its ability to interfere with the learning process of machines for collecting and analyzing data. The work invites reflection on the relationship of users with Google and allows the breakdown of the usual use of this type of platform: information search. אור ('or) Translate presents a noisy comedy that aims to push the limits of acceptance Google's use terms and privacy policy; for this, it encourages the user to play with the translator and try to confuse it.

The Google Cloud Platform API is deformed to translate text in 50 different languages, modifying its programming so that it generates sequences of automatic translations and modifies the languages of the text to be translated and of translation in each execution, randomly selecting another language. The design of the installation degrades the semantic content of the information, brings the communication code closer to the limit of non-functionality and causes the breakdown of the usual feedback between user and machine. An interactive process is proposed that aims to generate interferences, interruptions and confusion in the Google platform by pathologically feeding its database, offering it noisy information related to a single profile created for the sound installation.