Will AI replace Writers?
If you are concerned that Artificial Intelligence will replace you as a writer, hopefully, this will reassure you:
The rise of AI has sparked debate across numerous industries, and there is a great deal of concern about its usage and implications for the future of mankind. Anybody covering AI in its broadest terms is encouraged to at least acknowledge its benefits before going deeper into the dangers. To give just two examples, trials are currently underway for an artificial intelligence tool capable of predicting the side effects of patients treated for breast cancer, and AI is also proving to be a formidable ally in the battle against climate change.
Whilst the positives are compelling, it is indeed hard not to dwell on the apparent negatives. There appears to be an underlying worry that AI is a danger to mankind, though what this means isn’t always made exactly clear, and often gives the impression of ‘machines taking over’ in a sort of (yes, even this article is compelled to use the term once) ‘Skynet’ doomsday scenario. But the Terminator movies are, well… movies, and reality isn’t nearly as action-packed, but is infinitely more complex, hence the concept of AI as an existential threat being a lot more nuanced when explained sensibly.
There’s plenty of information regarding the pros and cons of AI available to read on the web, as well as how it actually works, and obviously, make sure that those sources of information are credible and reliable! But a prevailing anxiety that hovers over many is that of AI’s impact on creative industries such as art, photography and, of course… writing.
Again, this is a broad subject that covers several entities reliant on the imaginative brilliance of human writers, a particularly prolific example being the Writers Guild of America strike in 2023. One of the big reasons for the strike was writers wanting to ensure that generative AI would not replace them in the creation of scripts, but to only be used as a tool to aid in research and optimising ideas.
The generative AI in question was, as you probably guessed, chatbot systems such as ChatGPT, which launched in January of the same year. ChatGPT was designed for numerous purposes including keyword research, proofreading, and even writing and editing code. As one might imagine, it could also be used to generate scripts for videos, create copy for websites and undertake other such writing tasks which would allow users with broader ambitions in online business and the creator economy to save time and resources on the job of writing.
And yes of course, the writing industry itself, along with some of its most prolific publishers, have embraced AI in order to enhance how they do business. Big names such as Penguin Random House now use AI for marketing strategies, administration, customer service, audiobooks and much more. Quick as a flash, publishing and its relationship with Artificial Intelligence has become molecular and inseparable. The ethics of skilled professionals losing out to AI, and the regulations by which writers and publishers can guard their work from being infringed to train AI models (to name but two) continue to be important debates with a great deal of moving parts.
In light of all of this, I’d like to focus on the way outright misuse of this tool would come to pass almost immediately, and how the world of creative writing was no stranger to this misuse. I, for one, was soon to discover that a reputable publisher of short sci-fi stories which I had been following for some years had to close their doors to submissions due to the influx of AI-generated fiction. A story also popped up on the web concerning a writer whose identity had been counterfeited to sell AI-generated books under her name. I myself would also receive a book of short stories (we will get back to this) supposedly written by a human author. But as it turned out, not only was this person not real, but the entire publisher turned out to be a content mill of chatbot-written stories, with bogus social media profiles of non-existent writers, fake comments, AI-generated covers, and press images that didn’t make a lick of sense when you looked at them closely.
The binding factor in all of these occurrences is dishonesty, for if people willingly choose to read and enjoy AI-generated fiction, they should not be criticised for it. Likewise, the benefits of ChatGPT appear predominantly ethical just if used responsibly. However, the fact that it was used almost immediately on launch to lie, scam, and absolve oneself of putting in the proper work was predictable and tacky. For me at least it prompts the necessity to state that anything I write, this article included, is 100% by my hand, save a quick pass through Grammarly in order to spot the occasional misuse of a comma and make sure I didn’t miss any ‘the the’s or ‘and and’s when proofreading. At the risk of coming across as a luddite, I will never use ChatGPT or its peers for anything in my writing... Likewise, I will never criticise anybody who does, except of course if they are being disingenuous with the technology.
There is an extremely intelligent and credible article by The Harvard Gazette in which science author and cognitive psychologist, Steven Pinker is quizzed on whether or not chatbots will supplant humans as writers. It’s an enlightening and objective read that discusses the topic with more insight than I could, and a recommendation for anybody seeking a more academic response to the subject.
My own thoughts on this topic are as follows:
I mentioned receiving a compilation of short stories, and to my embarrassment, it wasn’t until I read the second story in this book that I realised that the premise of this being a flesh-and-blood human creation was in itself a work of fiction. Once I twigged, it was impossible not to see, with the events of the stories reading more like a list of bullet points, the narrative consisting of a generic beginning-middle-end structure, the characters having zero emotional clout and the closing paragraph rounding off the plot like it was the evaluation of a science experiment. ‘Following the incident, this happened to the character. The experience made them a better person, and they were very happy about it. The end.’, or clauses to that effect.
Though, through various descriptive methods, the stories were able to evoke certain images, sounds, and other sensations, and whilst it’s undoubtedly interesting what AI can build around a simple prompt, the two stories were completely devoid of any feeling or meaningful content. And after skimming through the book and realising that every story ended with the same deadpan roundoff of events, I had no reason to believe that any other short story in the book would be different.
The summary of this experience is that there was one good thing about observing AI-generated fiction. It really hits home how profoundly incredible literature is and always has been, literature written by humans that is. The connection between writer and reader is indefinable, an emotional network that a machine in this day and age can surely have no hope of understanding, for how do we begin to understand it ourselves? The way books express and communicate remains as important now in human society as it did centuries back, before cinema, recorded sound and digital media. With their ability to inform, entertain, humour, enlighten, enhance and in any other way irrevocably alter the feelings of the reader, books remain a vital part of humanity, and a cathartic experience for both the subject and the author who toiled in order to pen it.
AI chatbots continue to develop and will most likely evolve to perform new and impressive feats, But unless this tool built on the amalgamation of existing knowledge and nothing more can ever find its own voice and dare to bring something new into the world, it is unlikely that ChatGPT, Grok or others like it, will succeed in besting the intricate germination of a good idea and the emotional journey taken to fully express it. Whilst it can conflate vast composites of data with a speed and accuracy that supersedes the capability of our own grey matter, it cannot access and/or utilise the dreams, inspirations and mood of the individual, and can therefore never create the great literature of the mind which is surely the essence of credible creative writing.
Simply observe Masterclass.com’s 10 Rules for Writing a Good Novel. Only the first rule, ‘Read Voraciously’ may be applicable to the capability of a chatbot. Others such as ‘Write for Art’s Sake’ and ‘Rules are Meant to be Broken’ are ideals way beyond the thinking of a machine. Whether or not AI will one day become sentient enough to understand these rules and think outside the parameters of what is taught, is a concept the writer of this article has read too much science fiction to discount completely. But whilst no such animal exists in our present day, you as a writer can rest assured that you are irreplaceable, and the skills you learn to make good novels and the imagination, creativity, and characterisation that comes with it, is an asset to yourself and the audience experiencing it.
As the publishing industry continues to demonstrate a willingness to sew AI into the patchwork of its economy, anybody disheartened by that marriage should remember that they still have the freedom to control their work however they want, and therefore exclude any usage of Artificial Intelligence if that is where they feel most comfortable. Chances are, you write to make others feel. Consider that a gift rather than a tool, an organic process exclusive to yourself, and the thing that makes your storytelling talent unique. So the number one priority should be to make your own rules as you see fit. Whimsical? Maybe. Counter-productive? Most likely, according to some…
… but pointless and futile? Never in a thousand years. passion projects and labours of love are what makes us human after all.
