by E.M. Blake

(This is part three of a three-part series. Read part one here and part two here.)

“If I was a professional writer rather than professional thinker, I’d feel slightly threatened by ChatGPT; not because it would replace me, but because of what it would do to intellectual property and the ability to make a living from writing,” says Dr Blay Whitby.

Dr Whitby’s 1984 paper on AI explored some of the immediate dangers of AI falling into the hands of the elite. His first book in 1988 considered the social implications of AI. I read Dr Whitby’s Artificial Intelligence: A Beginners Guide (2008), while researching AI. His book and our conversation highlights the impact the writer’s imagination has on technology. Dr Whitby can’t think of a prominent AI researcher who hasn’t read an awful lot of science fiction and, one of the jokes is that “AI undergraduates think the AI degree is about making real computers like the ones in the movies, and there’s some truth in that.”

In my fourth interview for this three-part feature on ChatGPT, Dr Whitby tells me he is a member of the Advisory Board of the All Party Parliamentary Group on AI, an ethics expert for the European Commission and a member of the Strategic Ethics Committee of BCS, The Chartered Institute of IT.

He says governments and technical advisors, “Feel copyright is out of date and they would like to move away from it.” I discovered the Society of Author’s website is a valuable source of information on the subject.

A Google search informs me that John Grisham, Jodi Picoult and George R.R. Martin are among 17 authors suing OpenAI. Meanwhile, an open letter signed by authors, including Margaret Atwood and Philip Pullman, called for compensation from AI companies for using their work.

Dr Whitby confirms my suspicions that, “The writing profession will have to adapt. They probably need to give more weight and prestige to the raw creative process than the craft of putting words together in a spectacular way.”

Watch Omniscient here © (Instagram: @_kays.journal)

He adds, “The craft part (of writing) will undoubtedly get better. I can tell when a student essay has been written by ChatGPT because it has impeccable spelling and grammar in British English, which is something students rarely achieve. That sort of everyday skill, or that everyday bit of the craft, will be done by AI and it will enable them to work faster.”

Dr Whitby asserts, “I’m not against ChatGPT or even reluctant. I just never seem to have a use for it.” He was on the ethicist panel for local small and medium enterprises and was the only one in a room of at least 50 that didn’t use it.

Aren’t You Glad Electric Guitars Were Invented?

We discuss music too. He says, “When electric guitars were first produced, people tried to play them like an acoustic guitar. It was quite some years before a new generation of guitarists who had learned on the electric guitar, like Hendrix and Clapton, did things with the electric guitar that you couldn’t do with an acoustic guitar. They learned to play it and I feel a lot of this technology is like that.”

I contemplate the riff in Sweet Child O’ Mine and who can forget Marty McFly in Back To The Future?

But back to talking about our future, Dr Whitby says, “The new generation will be able to do things with ChatGPT because they’re not treating it as the older technology. They can do new things with it.”

How can we adapt and make the most out of ChatGPT? Dominic Jarkey, a senior R Programmer and data scientist, provides a fascinating perspective in my fifth interview. He finds ChatGPT primarily beneficial for, “Coding tasks I do infrequently, the things I don’t have a complete mastery of.”

ChatGPT Makes Research Easier

I love time-saving efficiencies and share Dominic’s enthusiasm for AI. He highlights how, “ChatGPT really shines at summarising existing academic literature and debates.” In, for example, political science research, follow-up questions might include: “What did this particular author argue? How did they argue it? What study did they use?” It saves time, it provides information about the top ten scientific papers and you don’t have to read 50 papers.

He adds, “I’d fact-check anything I use directly that I’m relying upon to make a decision.” He also says, “Facts become easier to verify after you know what they’re supposed to be. You ask ChatGPT about ten birds that live in a habitat. You don’t know the name to verify until you’re given it.”

Are We Interacting With ChatGPT In The Wrong Way?

Dominic says ChatGPT, “Is a statistical model, literally predicting the next word in a sentence. The sentence is being spoken by an all-knowing assistant which is very good at guessing. It’s not, in itself, the assistant. It’s a model guessing what the assistant would say. If you see the assistant as a being in a computer, you will interact with it in the wrong way.”

What does he mean by statistical model? “It’s a probabilistic guessing game. If I say ‘the cat sat on the,’ then asked you what the next word is, you might guess mat, floor or tree. You’d be more likely to guess floor or mat than tree, because you have a notion of what sentences are like. A statistical model is a way of apportioning different probabilities to an outcome. ChatGPT and these large language models treat language like a pattern. Statistical models are no different to what we use to predict the weather. It’s not right all the time, but it has a decent chance of guessing that,  if clouds are coming, there will be cloudy days ahead.”

Dominic foresees that, “ChatGPT might be held up as the flagship bringer of the movement. AI will become specialised.”

I reflect on my other interviewees’ comments.

Dr Kholeif says, “To dream of a different possible future, you can’t be antagonistic about it from the outset. One needs to be open to the possibilities of change. Change is a gradual thing. It emerges with patterns of behaviour and acceptance, where we need more of a mindset shift through literacy.”

Imo contextualises ChatGPT by glancing back at our history. “Were people suspicious of a spell check or computers in the beginning? Were they thinking, it’s not a typewriter, it’s cheating? In ten years, will we be more obviously relying on it, and at that point, will be it be more accepted?”

Catherine infers, “AI can be helpful but after a certain point people will ignore it or be annoyed by it.” She uses the example of art from an actual artist, where she’d examine it more critically, “Rather than ‘oh that’s neat’ and scroll past.”

In 1984, Dr Blay wrote: Humans are an amazingly adaptable species. Attenborough lists us as one of five opportunist species. We will adapt to this technology, just like we’ve adapted to every other technology.

Now, 40 years later, what if ChatGPT empowers us to become kick-ass writers, editors, artists, thinkers, parents, children, friends and humans? One of my favourite sayings is Walt Disney’s: “It’s kind of fun to do the impossible.”

Are we like butterflies flapping our wings, buffeted by wind, or caterpillars about to undergo an extraordinary metamorphosis from a chrysalis?

Rie Kudan admitted ChatGPT helped write five per cent of a novel that won her Japan’s most prestigious literary award. It unlocked more of her potential as a writer by influencing her creative process, including sometimes inspiring dialogue in ‘Tokyo-to Dojo-to’ (‘Sympathy Tower Tokyo’).[1]

AI could have churned out this series of features but, instead, I invested three months nurturing my pieces, using ChatGPT for minor editing at a sentence level. Maybe it’s nostalgia, a desire to create the best experience for readers and that, perhaps, the journey is more important than the destination.

(c) Ellie Blake, 2024

*****

Ellie M. Blake is a regular contributor to Write On! and the author of the children’s picture books Dilly Dally Sally and This Silly Bench.

Website: mycreativeeveryday.weebly.com Connect: Instagram, TikTok, YouTube & X

The Interviewees

The Interviewees

Catherine Sheffield and Imo Scrimger are completing their second year of an MFA at City University.

Dr Omar Kholeif’s latest book, Internet_Art: From the Birth of the Web to the Rise of NFTs is published by Phaidon. Website: www.omarkholeif.com Connect: Instagram & LinkedIn

Dr Blay Whitby is a professor and technology ethicist. Website: blaywhitby.org

Dominic Jarkey is a Senior R Programmer and Data Scientist with a Masters of Political Science at Washington University. Connect: Github

[1] Averre, David and AFP, “Literary prize winner says ChatGPT wrote some of her sci-fi novel which judges described as ‘almost flawless,”  Daily Mail Online, Updated: 16:07, 18 January 2024 (29 January 2023), dailymail.co.uk/news/article-12979117/ChatGPT-wrote-prizewinning-novel-flawless.html

*****

You can read Write On! issue 20 online here and find it in libraries and other outlets. Previous editions of our magazines can be found here.

You can hear great new ideas, creative work and writing tips on Write On! Audio. Find us on all major podcast platforms, including Apple and Google Podcasts and Spotify. Type Pen to Print into your browser and look for our logo, or find us on Podcasters.Spotify.com.

*****

If you or someone you know has been affected by issues covered in our pages, please see the relevant link below for ​information, advice and support​: https://pentoprint.org/about/advice-support/