My 5 Metre Writing Certificate
Professional validation is worthy of both celebration and experimentation.
Earlier this week I attended one of Hampshire Writers’ Society’s monthly meetings. Writing must be one of the loneliest occupations and the Society is one of four local writing groups I participate in to help deter my anchorite tendencies. One part of the meeting is devoted to announcing the winners of that month’s flash fiction competition, set and adjudicated by the meeting’s invited guest author. This month’s author was Joanna Barnard, who offered some refreshingly transparent insights into the hardships of traditional publishing.
I’d entered the competition a few weeks earlier, forgotten even the title I’d used, and so politely clapped the announced winners and runners-up, until I realised I was clapping for myself.
I drove home like a first year pupil running up to a parent clutching a five metre swimming badge and certificate. It's my first ever formal writing accolade and sometimes you need a psychological boost to get you through the week. This was certainly a welcome fillip.
But, me being me, I thought I’d use what I’d written to experiment more with ChatGPT - y’know, that super-intelligent ‘bot thing which will soon take over everyone’s job and then the world. By next year. Unlike the printing press and typewriters and word-processors; trains and planes and automobiles; radio and TV and home computers; record players and cassette tapes and video recorders; the web and smartphones and social media; and… well, anything that causes wholesale disruption to the established (and profitable) order of things.
Here’s my original flash-fiction text which was highly commended:
Entwined in Time
“Ava, please look at me.”
“I’m reading.”
“I don’t care.”
I snap my book shut and look up. “What do you want?”
No matter what everybody said, it wasn’t like looking in a mirror. A mirror should only reflect yourself. Your rights are its identical lefts, and your ups shouldn’t be its downs. But looking at my twin sister was different. It’s more of a translation, from my flaws into her perfection – if I allow it. As our gazes connect, those three wayward eyebrow hairs, bristling with my frown, vanish into her flawless complexion.
“Yesterday – in the garden – did you feel anything different?” she asks.
“Different to what?” I reply – adding, “or when?”, cautious at what she might say; how wide a division she might force me to make. Arguing with my twin often mirrors arguing with myself. If I’m forced into making a digression, I’ll pressure her to do the same. Amplifying instead the differences in our faces; cementing our individuality into separate spaces. Even when it’s about us, we should remain ourselves. It’s the only way to remain sane.
“It was as if… as if we were treading on ourselves,” she says.
A curious way to put it, but I had felt it. Like walking on a frozen pond still capable of a perfect reflection. “Yes, but only when I looked at you,” I say, giving away too much.
My eyes stare back at me. Into another mirror – like our name.
We both say, “Ava, don’t look at me like that.” But her words come first, her voice differing not in pitch or loudness, but in time. She’s divided us anew with a sharp slice of our temporal knife. She’s pulled herself into our future, as I fall again into our past.
But then she replies, “No, Ava. Not this time. This time it’s you, in the wrong present.”
I then emptied the above “inventive… mind-bending” text into the maw of a giant word-sucking device and asked it to traverse its language-based landscape of weighted and layered connections using some sophisticated statistical equations1.
You know that’s what it does, right? It has absolutely no clue about reality or emotions or feelings, human or otherwise. But this won’t stop people ascribing such characteristics to ‘their’ ‘bots and turning them into sentient creatures, or even other people. Hypocritical confession: I already call my robot hoover ‘Poomba’ and scold it when it misses a dust bunny.
This was the bot’s summary analysis when prompted with the ~310 word fictional piece I’d written above:
“That's a very intriguing piece of flash fiction. I like how you use the mirror metaphor to explore the relationship between the twin sisters and the concept of time.”
I have to admit I was impressed, despite my 25+ year pedigree as a jaded ‘seen it all before’ technologist. The bot had correctly spat back out the subtle theme and metaphor underpinning my story. It had done so in mere seconds, and likely better than I could, even if given more time. After all, I only wrote the piece, not deconstructed or judged it. This is why such ‘bots are fast becoming such a useful tool for both work and play.
Then I committed writing heresy and asked told it to “please extend the story”.
This is what it came up with:
I blink, confused by her words. How can she be in the right present, when we are both here, in the same room, at the same moment? What does she mean by wrong present? Is there another present that I'm supposed to be in? A present where we are not twins, where we are not mirrors, where we are not Avas?
She smiles, a sad and knowing smile that I don't recognize on her face. She reaches out and touches my hand. I feel a jolt of electricity, a shock of recognition. She is me, and I am her. We are one and the same. But how? And why?
She speaks again, her voice soft and gentle. "Ava, listen to me. We are not normal twins. We are not even human. We are experiments, created by a mad scientist who wanted to play with time. He made us from his own DNA, but he altered us in different ways. He gave me the ability to see the future, and he gave you the ability to see the past. He wanted to see what would happen if he put us together, if he made us share a present that was neither ours."
I shake my head, refusing to believe her. This is insane, this is impossible, this is a nightmare. She is lying, she is crazy, she is trying to trick me. But then I look into her eyes again, and I see something that makes me doubt my own reality. I see myself, but not as I am now. I see myself as I was yesterday, in the garden, feeling something different. And I see myself as I will be tomorrow, in a lab, feeling something worse.
She squeezes my hand and says, "Ava, please trust me. We have to escape from here. We have to find a way to break free from his control. He is coming for us soon, and he will do terrible things to us. He will try to separate us, to erase us, to destroy us. He will try to make us forget who we are."
I feel a surge of fear, of anger, of defiance. I don't want to lose her, I don't want to lose myself, I don't want to lose our present. I nod at her and say, "Okay. Let's go."
We stand up and run towards the door. We don't know where we are going, or how we will get there. We only know that we have to stick together, that we have to find our own present, that we have to be Avas.
Again, impressive and took only moments to generate. Would change some bits, like the newly created ‘mad scientist’ and ‘lab’ backstory and any American spelling. Also tell it to tone down the anger emotional element, and perhaps request an alternative last paragraph ending. I’d of course also undertake my usual editing steps with my preferred piece of ubiquitous writing software.
In employment, when technologies appeared on the corporate horizon (or, more likely, I journeyed towards it in search of them), I’ve been paid to analyse and suggest how best to adopt them. So it would be foolish not to do the same for myself, even if unpaid.
Here’s what I plan to do:
Refuse to be cowed. Resist being subsumed.
Operate a passive policy of if you can’t beat ‘em, join ‘em.
Learn how to command, wield and compete.
Ensure such tools are more of an anchorite chamber pot emptier than a critique partner. It’s people who want to give and receive love and embrace happiness; who suffer conflict and endure sacrifice; who can inflict pain and trauma. Not ‘bots. So far.
That’s all for this week, folks!
Please take care and be ‘bot aware.
JR
The Large Language Models (LLMs) used by modern chatbots are typically a combination of natural language processing, deep learning, and generative models. Some of the techniques they use to understand textual inputs, generate outputs, and evaluate the quality of outputs include:
Natural language processing (NLP): A branch of artificial intelligence that deals with analysing and processing human language. NLP can parse text, extract keywords, identify topics, sentiments, emotions, and intents. It can also generate natural and fluent text to match a particular style and genre.
Deep learning: A subset of machine learning that uses artificial neural networks to learn from large amounts of data. Deep learning is used to train language models on various types of content such as novels, poems, essays, songs, images, and more. It is also used to fine-tune models on specific domains or requested tasks.
Generative models: A type of deep learning models that can produce new data based on existing data. I use generative models to create original and diverse content that is relevant and coherent with your input. I also use generative models to modify or improve your existing content based on your feedback.