“We propose the term ‘quasi-text’ for word sequences produced by ChatGPT”

Lhe automated generation of language elements is a major topic in the field of artificial intelligence (AI), and the latest version of ChatGPT constitutes a significant step forward in this field.

Developed by OpenAI companythis model of deep learning generates sequences of words that look like human-produced text, by predicting the next word from a given sequence.

As we move towards a future where our writing practice is aided by algorithms, it is essential to consider the potential consequences of this type of production.

Competences and communicative intention

The rapid increase in the supply of assisted writing can indeed lead to a feeling of disorientation and uncertainty as readers try to adapt. Certainly, AI bots will have a significant impact on many industries, but they may also lead to valuation and decision-making difficulties for some players due to the sudden change in the writing supply curve. . To navigate this new information landscape, these sectors will need to reorient themselves and adapt to the changes.

And also in our file: Article reserved for our subscribers Artificial intelligence: the many faces of deep learning

To understand the impact of this form of AI on writing practices, the art of writing can be broken down into two aspects: writing skills and communicative intent.

Writing skills thus refer to the technical aspects of writing, such as syntax, punctuation, structure, and fluency.

The communicative intention refers to the reason or the message of the text. In other words, it is above all about the reason why a person will try to communicate with another.

See as well : Artificial Intelligence: Does ChatGPT open a new era?
Related Articles:  Kuwaiti newspaper newspaper | Cemetery in the Milky Way

It’s entirely possible that ChatGPT could one day surpass humans in generative skills, given that its AI models use hundreds of millions of texts to emulate styles and produce texts similar to the corpus on which they have been trained.

These models might also be able to mix different writing styles until they create their own style. However, a limit must be drawn: they undeniably lack the “communicative intention” which underlies, in humans, the writing and circulation of texts.

Neither awareness nor ability to interact

Individuals seem to experience a certain enjoyment in the use of language robots (their use becoming that of tools or toys) by offering command prompts in the form of questions, precisely because they inject a minimal communicative intention into the system. and see the rest come true.

You have 55.18% of this article left to read. The following is for subscribers only.

Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.