By Sue Moberly
– A postscript to the earlier blogpost The Robot Will Write Your Blogpost Now
An online article on the BBC news website (Jill Martin Wrenn, business reporter) asked ‘Are romance authors at risk from book-writing chatbots?’
It opened by saying that, although sales of romantic fiction are continuing to boom, the genre is ‘often accused of being formulaic,’ which may suggest that a robot could easily take over.
The article includes an interview with Julia Quinn, the author of the bestselling Bridgerton series of novels. With a reputed twenty million books in print in the US alone, Ms Quinn is reputed to be one of the world’s bestselling authors. The TV adaptation of Bridgerton is one of the most-watched shows on Netflix.
Ms Quinn, and other authors interviewed, suggest that these futuristic tools will keep human authors on their toes. Certain sectors of the fiction market are allegedly already getting away without crediting (or even using) a human author.
As I indicated in my earlier post, the ‘problem’ is the release, last autumn, of ChatGPT – the advanced language processing technology developed by OpenAI.
When trained, ChatGPT can produce intricate writing that is difficult to distinguish from that written by a human. It has made headlines partly because it can be used by students to author their essays. The Captcha clause asks you to declare ‘I am not a robot’ by ticking the box next to that statement. However, if robots are so damn clever, they are surely clever enough to put a tick in a box. When I was doing my MA in Creative Writing a few years ago, and even now, the academic software (‘Turnitin’) for essays, would catch you out even if you tried to quote sentences from your own previously written work. This new software could potentially help in achieving passing grades, however it is still a long way from being able give students the desirable instant high grades. For more on this, check out this recent BBC news article here.
Open AI launched its latest version of ChatGPT back in March this year and Google, unsurprisingly, released its rival system called Bard around the same time.
In an interview on BBC’s PM news programme recently, Simon Armitage, the current Poet Laureate, was dismissive of Bard’s early attempts at writing poetry, which must be reassuring for human poets. However, in an article for The Washington Post, Seth Perlow pointed out that Alan Turing originally started his test of ‘sentience’ for robots with a literary request: “Please write me a sonnet on the subject of the Forth Bridge”, which he expected any artificial intelligence in his lifetime to fail. When Perlow sent the same request to ChatGPT, in less than a minute the programme created a full rhyming Shakespearean sonnet which began “Upon the Firth of Forth, a bridge doth stand”.
If you find all this very scary, you are in good company. On 1st May Dr Geoffrey Hinton, ‘The Godfather of AI’ resigned from Google, after working there for a decade on AI. He said he was concerned about the dangerous speed at which generative AI is progressing. Many are calling for a pause in AI Research, but many others accept that the ‘Genii is now out of the bottle’ and that there is no going back.
In an interview, available on YouTube, the late world-renowned Physicist Stephen Hawking said something to the effect that once AI could replicate itself, it was ‘game over’ for humanity. In a more recent interview with Piers Morgan, also on YouTube*, the world-renowned Biologist, Richard Dawkins, conceded that Hawkins could be right, but no one could say that for sure. The way Dawkins framed it was that any future AI would have been ‘conceived’ by a human mind, even if it were ‘born’ in a different form, which he did not seem to think was the disaster envisaged by Stephen Hawkins. As an 82-year-old Atheist, I suspect that Dawkins cares far less about this than he might have in his younger days. Although I am only just over ten years younger, I care deeply about what happens. Maybe it’s because I have poetry in my soul. Or maybe it’s because I find I cannot keep up with the new developments in editing software and feel that they are muscling in on my autonomy as a writer. There is now potentially so much ‘interference’ between the editing software and the final article that it often seems much easier to write with a pen and paper. This might make it difficult to publish anything but is that any worse than having one’s expressed thoughts at best ‘corrected’ or at worst ‘altered’?
* This discussion into AI begins at 25:38.
ABOUT THE AUTHOR
In another life, Susan Moberly worked as an Advertising Copywriter. She trained for the job at the ‘Oxbridge’ of Advertising as the Watford Copywriting course was referred to all those years ago. After the one-year course, she did a short internship at Ogilvy, Benson and Mather in London and ended up as a junior copywriter headed up by Salman Rushdie (yes, him). She then spent ten years or so at various London Advertising agencies writing radio, television, and press advertisements. Susan gave up work when she had her son because she suffered from severe post-natal depression due to a very traumatic birth. Life became a struggle thereafter. When her son got married and was expecting his first child, she started an MA in Creative Writing at Surrey, finishing it when her second grandchild was due to be born. She now has three lovely grandchildren. They, along with their parents, are her motivation for everything. Susan has always written poetry and kept a journal, and still lives in hope of being published one day. If not, she would like to win the lottery to move nearer to London again.