From human to binary and back: how can we translate the digital machine?

Résumé En

From human to binary and back: how can we translate the digital machine? Everything that happens in the digital environment is about writing and translation: it can be from one human language to another but it is even more from computer languages to human ones and back, in a translation loop that goes both directions. Hence, everything digital is a form of mediation through different languages. These processes of translation between code and human language are done automatically, but this is true for more and more translations between human languages, as machine translation services are growing and perfecting. Since most of the times we do not know how these systems work, we need to develop tools to translate the machine. How can we use creativity to translate it for a better human understanding and use? How can we improve the relationship between us and the computing machines? To help tackle the emerging tensions in these relationships, especially with so-called "AI", we should first find a better way to define these technologies: from the "stochastic parrots" to new proposals to replace the deceiving terminology of "AI", critical efforts by scholars in this sense have been manifold. We should then look inside these machines, inside their code: if we know how they work we, as researchers and/or translators, will be able to intervene on a deeper level to coordinate and correct machine output. And that is pivotal if we are to embed these technologies in our everyday life, as it is already happening. We will then go through different examples of how we can open "AI" black boxes and engage with them in a creative and critical form.