Rethinking the role of writing in the age of AI

Novels written by Artificial Intelligence are beginning to fill shelves in bookstores, even if they are digital shelves. Universities that have focused on writing skills in various specialisations, including Sciences, Humanities and Liberal Arts, are having to face tough questions about their courses. What is the point of teaching writing, the argument goes, if software can write within seconds what a human would take hours or days to write?
The questions are not far-fetched.
Data across higher education have shown increasing frustration on the part of faculty members with the way in which learners are using AI technologies to write reports, essays and even full-fledged research papers.
Band-aid, short-term solutions to this issue have been to bring back in-class writing, oral assessment, and, overall, to try to focus on process rather than final product.
That may be easier said than done when dealing with large classes where individual attention becomes challenging.
On a more positive note, it may be that this is a passing fad. The fancy with such easy writing is already seeing some kind of pushback. When it comes to creative writing, the attitude is even more pronounced.
A research study on attitudes to AI-written stories found that readers responded unfavourably once they were told of the authorship of the fiction.
Research conducted by the University of Michigan found that 'New research finds that people evaluate creative writing less favourably when they learn it was generated in whole or part by artificial intelligence.'
There is another reason why such writing will feel soulless and dry, even when it is not creative writing.
A study conducted by Sultan Qaboos University found that learners needed to voice their opinions about texts in a language that felt immediate and reflected their own feelings. This was particularly true when writing opinion pieces such as literary essays.
As the authors suggest, “In the analysis of literature, interpretation and evaluation make writing more sophisticated than summarising and explaining, and this is achieved through mastery of individual voice.'
To what extent AI could generate such individual evaluation is not clear yet. Nor is it very clear that we want to read the interpretations of a software over those of individual readers to understand, reflect, and critically think about a character, a plot, or even a theme.
Throughout history, writing has been about more than putting words on paper. It has been about the process of thinking, of processing ideas, rejecting them, and contemplating them.
The sentence was only the final product of those thoughts, not the beginning.
In many ways, AI just skips over that struggle. There is no tussle over what we really want to say, nor is there any conflict for the reader about what the intention of the author really is: where are they coming from? What ideas are actually being questioned? How does a novel, or even an essay on a novel, point to new ways of seeing the world?
These are questions that come and are conveyed through a range of language, at times even creating new words for emerging ideas. Large Language Models (LLMs) use already existing language. That does not mean that anything new is being created; it is mainly recycling existing ideas. Surely, we want to read more than that, especially in the creative arts.
Rejecting AI may not be very practical anymore, but focusing on the process, on finding new ways of expressing old and new ideas, will make it more possible for a writer to find pride in their work, to claim ownership of the practice of critical, out-of-the-box thinking.