"Machine", "Tool": Confusion in the very words employed

We no longer know what words to use. The sudden appearance of AI in the daily lives of organisations has created a problem of definition. But agreeing on a reality is the first step towards overcoming it.

By Kevin Erkeletyan

“The emergence of these tools”, “the help that the machine gives you”… Even with the experts we interviewed, the words “tool” and “machine” seemed to come up interchangeably to describe Artificial Intelligence (AI). It’s as if things between us and AI are moving a little too fast. What do you call this relationship? What does AI mean for us? What is our relationship with it, and its relationship with us? It erupted so suddenly into our daily lives that we had no time to think about it or reflect on it. AI makes us confused. It almost makes us talk gibberish.

Are these fuzzy terms also due to the fact that the language of Molière is not the language of AI? The leaders in the technology are Google, Microsoft and OpenAI; they are based in English-speaking countries and disseminate their innovations and work in English. The main academic publications, international conferences and educational resources are also mainly in English. English expressions like “machine learning”, “deep learning” and “neural networks” are thus becoming part of French vocabulary, sometimes for convenience, and often without an appropriate literal translation. French struggles to provide precise equivalents for English technical terms. “Machine learning”, for example, is often translated as “apprentissage automatique” (“automatic learning”): an expression that does not capture the depth of the original concept. So, in the interests of precision, French people stick to the English.

BORN BEFORE (THE) TERM

But can we visualise what is said or written in a foreign language as well as we can in our own? The technical terms associated with AI are often opaque and complex. Technological jargon, coupled with the increasing use of English, is widening the gap between the initiated and novices. This inability to find precise terms in French can lead to a loss of meaning and further confusion. English contributes to creating a barrier for those with a poor command of it.

As we discussed AI with the experts we interviewed, however, one point seemed to emerge: the vague impression that the word “machine” is used when we talk about AI in general as a meta-human abstraction, and that the term “tool” comes up when we talk about its applications. Turning this feeling into a certainty will be one of the challenges facing AI education and digital literacy, which should be widely shared. Otherwise, AI will have the effect of a machine on us.

A BRIEF ON-BOARD GLOSSARY OF GENAI

Generative AI :

A type of artificial intelligence capable of creating original content, such as images, text or music, in response to requests.

Large Language Model, LLM :

Advanced algorithms capable of understanding and generating text in natural language, such as ChatGPT.

Deep Learning :

A branch of machine learning using artificial neural networks to model high-level abstractions in data.A branch of machine learning using artificial neural networks to model high-level abstractions in data.

GPT (Generative Pretrained Transformer) :

A language model developed by OpenAI, capable of generating text in a consistent and contextual manner.