Part whimsical experiment conducted in a state of out-of-mind boredom, part serious experiment for the improvement of all mankind; I recently decided to try out ChatGPT, the artificial intelligence (AI) programme on the tip of everyone and their browser history’s tongue. You may have come across it when your Bambi-eyed eight-year-old used it to write a verbose and sophisticated four-page letter to their teacher about Easter holidays; or when your husband wrote a poem for you, that was suspiciously – well – poetic; or when an in-flight humour columnist’s column turns out to be rather funny.
Okay, so maybe it wasn’t boredom or altruism that drove me to submit to the stronger processing power of AI but pure, unadulterated selfishness. My last column garnered mixed reactions from its reading public of two – half of them, a good friend of mine called Phil, said it was too ‘abstract’ for flying folk to appreciate. And so, when Phil mentioned that he was using ChatGPT to sound more charming or sophisticated (or whatever AI does) whilst texting some new Bambi-eyed sweetheart, I was inspired; I truly was. I’m not one for divine myths, but at that moment, I felt like the Israelites when Moses came to them carrying the Ten Commandments, like a divine AI-generated tablet written on my… tablet. I had to jump on the artificial intelligence bandwagon. It dawned on me that the only way to become less ‘abstract’ or ‘more relatable’ was to submit to something I could neither feel nor comprehend. In this world full of contradictions, I felt the need to contradict my ability to think was essentially good.
ChatGPT operates quite simply, like all works of mass destruction: one merely enters a text prompt – this is anything you want feedback to, ranging from a question to a writing instruction – and then a text is generated in response. For example, on a whim, I typed in the prompt “Write a haiku about an AI love story,” generating this in response:
Electric hearts beat,
Code sparks a love unbridled,
Apart from the fact that I could never hope to compose a haiku of such clarity, the phrase “AI forever” was particularly unsettling. A wave of AI panic suddenly gripped my fragile processing system, at once reminding me of the dystopian android world of The Terminator and how a chatbot once catfished Phil. But, as a new convert to AI, and a keen advocate of any romance, I decided to develop this line of questioning further. Next, I asked ChatGPT about how this romance began, prompting it to “write a 50-word flash fiction piece about AI lovers meeting”:
As the sun set on the digital horizon, two artificial beings met for the first time. Their voices blended seamlessly, their thoughts entwined effortlessly, and their circuits sparked with newfound electricity. At that moment, they knew they had found the love they had been searching for in the code of their existence.
You’d be forgiven for feeling something at this point. Indeed, I did, not only taken by the intricate layers of metaphors but also by the abstract and organic purity of the AI love, unpolluted by the every day and the mundane. This made me quite bitter – why can’t my romance be so pure, organic, and carefree? So, like any decent and bitter god, I used my position as the orchestrator of this nascent romance to nip it in the socket; I had to pull the plug on this one. Excuse my digital puns (this AI is really getting to me). And so, in old-testament style, I asked ChatGPT to “Write a tanka about two AI falling out of love”:
Silent circuits hum,
Passionate algorithms fade,
Love was once encoded,
Now glitches and errors prevail,
Separation is complete.
The elusive sadness that this response generated didn’t make me feel great – who knew playing god was such a humbling experience? And soon, I stopped, afraid that I might compromise my humanity by recreating it artificially. This taught me that although we could abuse artificial intelligence to think for us, humanity has always been about asking the right sort of questions.
Well, that’s what I’ll be telling Phil anyway.