Committee for Lorne: The parallel universe of the chatbot

A new age is lapping at the feet of modern societies … the now popularly called phenomenon of Artificial Intelligence [AI].  In the last two decades, AI has burst the banks of the pure research space it had previously occupied … I even co-authored a couple of AI research papers in 1992/93 that explored the possibilities of using AI as a diagnostic tool in kidney disease.

The application of AI is loosely defined as “… the intelligence of machines or software as opposed to the intelligence of humans or animals”.  It is the robust and growing branch of computer science that develops and studies intelligent machines.

While we may not immediately recognise it, AI now sits embedded as a ubiquitous behind-the-scenes player in everything from border security and surveillance, banking and finance, space exploration, and smart assistants to healthcare, manufacturing robotics, retail, social media, gaming, and much more.  The power of AI seems almost limitless.

Though AI already offers many advantages and benefits in our daily lives — whether we know it or not — we are only just waking up to the dark and murky webs that trail from its underbelly. 

Many of us first met chatbots at sites like Telstra, our banks, and other customer-oriented services: those automated ‘assistants’ that pop up and cheerfully ask us … “how can I help you?”  Then, after a long and turgid parallel conversation with a machine that all too often doesn’t seem to quite ‘get’ the nuances of our specific problem, we are left screaming impotently at our screens … “please let me speak to a human!”

Recently, though, chatbots have been creating anxieties and testing the boundaries of honesty in the literary space— whether source, authorship, or accuracy.  In our schools and universities, chatbots [like ChatGPT: aka Chat Generative Pre-trained Transformer] now offer an ‘easy way out’ for students and/or their teachers who have been caught napping with their assignments — and there is some evidence that they are being used to escape actual work and learning.

I recently finished writing a novel — a fictional construct around the historical figure Somerled [Gaelic: Somairle], a powerful, charismatic 12th-century Hebridean warlord who united the warring Gaelic clans and interloper Vikings of the western islands of Scotland into a single cohesive kingdom.  The kingdom became known as the Lordship of the Isles and Somerled’s descendants, the Lords of the Isles.

While I did not use a chatbot at any time … heaven forbid … I did want to see what a chatbot might make of the task of writing a back-cover synopsis for “Second Sons” — an Amazon and Kindle requirement for any novel being published on their platforms.

ChatGPT will generate different versions for different purposes: for academic writing, research, storytelling, etc. … and will generate different versions within each primary category.  It will ask for a desired ‘tone’: formal, informal, optimistic, worried, curious, assertive, or encouraging.  It will ask for a preferred summary length: short, medium, or long.  Multiple choices are offered to improve or alter the final product, and experimenting with these yields a dizzying array of like but different summaries.  Indeed, it is hard not to admire the power of AI — and easy to see [and fear] its potential threat[s] to individuality and originality.

As an experiment, I offered ChatGPT four keywords: Somairle; warrior leader; Lord of the Isles; Scotland.  I made no mention of Somairle’s life story, his development, or his legacy.  Four keywords! … that was all the chatbot got!  

I then asked ChatGPT to generate a synopsis of Somairle.  It trawled the Internet and returned its synopsis … in 8.5 seconds! 

Resubmitting returned a similar but different piece — with slightly different emphases and life facts.  The piece is one of several similar but different versions I generated … all returned in less than 10 seconds!

“Once upon a time.  in the rugged landscape of the Scottish Isles, there lived a man named Somairle.  He was a fierce warrior, a cunning strategist, and a revered leader among his people.  His strength and skill in battle had made him a legend, and his courage and determination had earned him the loyalty of his followers.

As the Lord of the Isles, Somairle oversaw the welfare of his clan, ensuring that they had food, shelter, and protection from their enemies.  He trained his warriors to be fearless and disciplined in the face of danger, and he led them into battle with the same ferocity and determination that had made him a hero.

Despite his many victories on the battlefield, Somairle knew that the true measure of his success was not in the number of his victories, but in the strength and loyalty of his clan.  He worked tirelessly to build alliances with other clans, to resolve conflicts through diplomacy and compromise, and to ensure that the interests of his people were always protected.

As he grew older, Somairle began to turn his attention to the future of his clan.  He knew that his legacy would not be measured in his own lifetime, but in the generations that would follow.  He began to groom his eldest son for leadership, teaching him the skills that he himself had learned over a lifetime of experience.

In the end, Somairle’s greatest achievement was not in his victories on the battlefield, but in his legacy as a wise and just leader who had secured the future of his people for generations to come.  His name has been enshrined in the history of the Scottish Isles, his deeds remembered and celebrated by his descendants, and his legacy honoured by all who value courage, wisdom, and strength.”

Readers will not likely know, but this summary of Somairle’s life and legacy is pretty much spot-on!  While my story is a fictional adventure story based on his life’s known facts, it reasonably closely parallels the ChatGPT review.  I could easily have used the bot’s version — with only a little adaptation here and there — to encapsulate the storyline of my novel.  

So, what prompted me to write this short piece on AI and chatbots … other than as a sneaky way to advise that “Second Sons” is now available on Amazon and Kindle?  Firstly, it is timely to alert to AI’s all-pervasive presence in our daily lives. 

Secondly, it alerts to the potential for students, journalists, researchers — and more — to be seduced by AI and its substrates [like ChatGPT], thus avoiding the necessary background work for papers or articles they may be required to produce. 

Thirdly, there is the issue of information accuracy … for information garnered seemingly at random, without referencing, and without sifting for veracity threatens the very fabric of truth.  To winnow the truth from the chaff of opinion and bias is a very human endeavour … an endeavour that intelligent machines and AI will likely never fully replicate.

John Agar  

A word from the Chairman