white and blue butterflies illustration
Artificial Intelligence, Blog

The Artificial Intelligence Revolution—How Do You Feel About It?

I’ve been debating for a long time whether or not to take the plunge and start my PhD. I’ve sent for details several times. I bought new stationery last year, used it all on other projects… and now I’ve bought more stationery for this new academic year. Will 2023 be the year I begin my studies? Possibly. I have found a research subject close to my heart, so here goes. How do you feel about the Artificial Intelligence Revolution?

The Artificial Intelligence Revolution

We all use AI to some extent, if it’s only Google or Siri. Almost everybody has an opinion on the AI Revolution. Some say they will never use it. Others say it marks the end of writing, writers, and the writer’s life I’ve enjoyed for so long. Still more have yet to make up their minds. The big question is, if pressing a button will regurgitate enough text to support a computer-generated cover art and a 99 pence price tag, will human authors and the books they produce vanish beneath what Joanna Penn calls “a tsunami of crap”?

close up view of an old typewriter
Photo by Suzy Hazelwood on Pexels.com

That’s what I’ll be researching when I start my PhD. Throughout human history there have been sudden, enormous changes in lifestyle. Wars, social and geographical mobility (think the adoption of the horse, bike, train, cars, or planes), revolutions industrial and agricultural…the list goes on. Humans have always adapted, and survived. AI will change writing, for sure, but keen readers will never stop reading. Everybody loves a well-crafted story. You don’t have to read more than a few pages to discover whether a book is worth reading on your personal scale of enjoyment. Who or what wrote a brilliant book is immaterial—unless you are a human writer trying to earn a living. That’s when things get tricky.

I’ve never used AI to write a book or article and I can’t imagine doing so, but I have employed it to help with the aspects of writing craft that I find tricky or time-consuming. Distilling a two-page synopsis into a snappy tag line of only a few words is something that keeps me awake at night. The media expect press releases to take a certain form and contain particular information. AI will produce that kind of technical, rather than imaginative, writing quickly and easily.

In these cases, I craft a prompt that will result in ChatGPT creating half a dozen examples of tag lines, or whatever text I need. Some will be rubbish, some will be quite good, but none (so far) have ever been perfect. They’ve all lacked something. For want of a better word, you could call heart. I choose the best line, then tweak it until it’s a perfect fit. As the famous writing instruction says, it’s far easier to edit something than to break a blank page. AI does that. It can blaze a trail. The quality of what you eventually produce from that starting point is up to you.

AI Assisted v. AI Generated

When human input exceeds the amount of machine-created work, Amazon and other players in the market call it AI-assisted. Broadly speaking, that’s a widely acceptable use of AI although (in my opinion anyway) you should probably ‘fess up. In contrast, if a computer was asked to produce a 50,000 contemporary romance in the style of Barbara Cartland with the intention of marketing it as such, that would be AI Generated. The actual percentage of human input needed before “generated” can fairly be categorised as only “assisted” is a point that will be debated by everyone, and will keep lawyers in business for years.

When creating ad copy distracts writers from real creative writing, that’s where AI can help. Feed the final draft of your two-page synopsis into ChatGpt, and ask it to create half a dozen tag-lines. It will respond in seconds. If you want a blurb for your book, AI can do that too. But of course, it’s not all good news.

Garbage In, Garbage Out

The most advanced AI system in the world can’t read your mind (yet). That means the resulting output can only reflect the quality of the prompts you gave the system. The Large Learning Models used to train AI systems are exactly what they sound like–vast quantities of data used to “teach” the AI system to recognise patterns in computer code, words, or symbols.

AI in the form writers use it at the moment can’t reason or think. Therefore, it can’t improve on your input, only suggest combinations of words and phrases most likely to agree with the prompts you gave it or the questions you asked. That proves the truth of an old computing motto: Garbage in, Garbage out.

Satisfaction (not entirely) Guaranteed?

There’s also the problem of possible plagiarism. Nobody should ever claim work as their own unless it has been throughly checked for plagiarism. Anything created by AI should be meticulously fact-checked, too. There have been cases where students have submitted beautiful assignments, complete with full footnotes, references, and citations. When checked, Artificial Intelligence has been rather too helpful. As well as doing a wonderful job of creating feasible original text, it has gone on to back up its arguments with made-up quotes attributed to fictional authorities and citations from non-existent textbooks.

If you want to read a real book by a real author (me!) while there is still time, my escapist romance Royal Passion is released on 19th October. To order your own copy from your favourite ebook supplier, click here.

You can find out more about me here, and see some of my books here.

4 thoughts on “The Artificial Intelligence Revolution—How Do You Feel About It?”

  1. It’s a fascinating and terrifying subject, Christina, and a brilliant one for a PhD. I wish you every luck with that!

  2. I think your approach is very sensible. When new things like AI come along it’s smarter to say ” How can I use this?” rather than “Oh no! this is terrible!” AI, like trains and smart phones, is here to stay. We should master it, not run from it. 😊

    1. Wise words, Sheila. Trains, cars and planes were once seen as terrifying technologies. Like fire they can all be as destructive as they are useful, but used carefully all are invaluable. Let’s hope the same thinking can be applied to AI.

What do you think?

This site uses Akismet to reduce spam. Learn how your comment data is processed.