Over the last few weeks the French publishing world has been in turmoil. Accusations of AI-use have been levelled at M.Orélien, the author of C’était ça ou mourir, which has won the FNAC literary prize and was shortlisted for more.
The whiff of prejudice
First, a whiff of prejudice surrounds the singling out of this book by Samuel Fitoussi, an anti-woke crusader. M.Orélien is Black, and a refugee from Haiti. His book describes the wrenching journey of its protagonist from Haiti to Montreal.
Are all books by middle-class white authors, those dealing with issues that do not question the way that we (i.e. middle-class Global Northerners that M.Fitoussi is part of) treat destitute minorities, subject to such scrutiny? Why was this one chosen for scrutiny?
We impute too much capacity to AI (which AI gurus want us to!)
Second, and more fundamentally, what, exactly, is the scandal about? It defies credibility that the book could have been written by simply feeding a few prompts into an AI machine.
So, even if we suspect AI has been used, the only credible way it could have intervened is to flesh out the writing, maybe re-work a rough manuscript into something more readable. If AI has been used, it may even have been as a language corrector, a grammar and spell-check. In any case, it is difficult to see how this is reprehensible.
Many authors obtain writing assistance : is AI assistance any different?
Third, let’s assume that AI is more literary than I think, and can sustain a train of reasoning, a storyline, and coherence over a few hundred pages. It will then have acted like a ghost writer or a good editor.
Do we condemn the hundreds of prominent ‘writers’ who use ghost writers, research assistants, or competent editors to help them express themselves? At what point does all this help amount to authorship not resting with the author? Whatever the reply, we have managed to survive the last few centuries without this being particularly problematic.
Should we trust AI to act as (possible) author, judge and jury?
Finally, the accusation was made on the basis of an AI model (Panagram) claiming that the book is written by AI.
Prior to this accusation, readers were moved by the book, which reflects some of the authors’ own experiences and feelings as he left Haiti, and realised the toughness and brutality of the world he travelled through, and the sheer luck that governed his survival (and many others’ deaths).
Should we leave readers’ opinions (and those of literary prize committees) at the door, and allow AI to tell us what a good book is?
I think not.
