Building a machine’s “understanding”

This article focuses on the production of automated dialogue systems in the fields of marketing and customer relations, exploring the implementation of these systems via the case of two commercial chatbot projects. These systems, which are sometimes considered to be “artificial intelligence”, require the establishment of a background context, the construction of knowledge bases, and continuous supervision. While these devices promise mass interaction with a brand’s customers, the need for permanent supervision raises the question of the scale of the mass development of these devices, which are unable to operate 100% automatically.