Do AI chatbots dream of electric awards?

“Don’t go all Skynet on this,” my boss said.

I lit an imaginary cigarette and stared into the dystopian rain; maybe this was me being programmed. Maybe this was, ironically, a prompt that exposed the pre-existence of the concept all along. You prompt. You push the button. The copy monkey delivers.

We’d already been doing it. Like a ChatGPT – but slowly.

Really slowly. And with bad breath. And skin complaints. And hangovers. God… a hangover. Did ChatGPT get them? Could we prompt it to write some copy as though hungover? Or is the future tee-total? No…unrealistic. Curse this debate.

My handlers cared not for the liminal explorations twixt the honest endeavours of the human scribes versus their seemingly inexorable usurpation by the AI army. All I knew was that I was no longer a mere copywriter (25 years, time served, as I reminded anyone who’d listen). No, we’d gone beyond knocking out a couple of paras for a pamphlet or a bit of social post chat.

I was now Rick Deckard, tasked with flushing out the replicants. I mean, perhaps this screed was also AI fodder?

“Chat GPT: write an ironic parody article about what ChatGPT means for copywriters, in the style of Ridley Scott’s Blade Runner.”

No, I knew that couldn’t be true; ChatGPT would have made it funnier for a start.

“But it’s just a tool,” my fellow writers cried, “it can’t channel emotions, references and facts in a well-crafted way to persuade folks to find out more about toothpaste.” Except, truth is, it can. Easily.

Thus exposing us charlatans for what we were: hacks, recycling the same old syntax, emphases, and filler terms – just changing the subject matter. I’ll confess, it’s easy to do once you have the hang of it. Just as with any job, there’s a groove, and once you find it – and you’re confident enough – the stuff just pours out of you some days.

Except, motivations matter.

ChatGPT has none, other than executing a task in response to a prompt. It knows not the pleasure of a turn of phrase. The absurd thrill of using a semi-colon correctly; something we should all revel in I believe. The geeky joy of writing to a word count or line count (and not making it look like you’re trying). The feeling of ‘a job well done’ to earn money, talk about it down the pub, pay a mortgage or buy a needy child some shoes… hang on, I’m meant to be Deckard, not Dickens… scratch that last one. But you get the idea. Does any of that matter though? Do you need such human baggage?

Well, not if you’re a soulless bean counter I suppose. It’s all about the profit margin. And if you can up that by ditching the meat puppets… hey, job done. Plus, who’s shedding a tear for the demise of advertising copywriters? Nurses, paramedics, teachers, firefighters… we can’t do without them. But people who knock out a clever headline about buying a bed or some word salad about fish fingers, few would mourn them surely? Suddenly, a knock at the door – or rather a thudding sound in my noggin. “You need to wrap this up, Montgomery, pronto… you’re losing them with your self-indulgent musings.”

The trail was going cold.

I lit another imaginary cigarette. And the dystopian rain beat down harder. Perhaps I could reignite this hunt for truth with some quasi-philosophical bullshit. It was all I had. ‘If a tree falls in the forest and no one hears it, does it still make a sound?’. ‘What is the sound of one hand clapping?’ God, I bet ChatGPT wouldn’t stoop to such lows to try and pad out an article. It was free from human foibles. Unless such weakness was programmed in right? “Chat GPT, write me some copy like a desperately insecure copywriter, who feels past his sell-by date.” The ultimate insult. The final blow.

Did any of this matter?

Did it matter that the emergence of this ‘threat’ pulled back the curtain, and exposed the copy wizards? Did it matter if it ruined the so-called magic? Was there any magic to ruin in the first place? Did it matter that one day, ChatGPT would or could win a D&AD? Without even trying. Well, maybe not. But would it take any pleasure in doing so? No. It wouldn’t. And that, was the turtle test. That was what would expose it in the end.

Oh yes, it could do as writers do – or as humans do. But it could never understand what it meant to actually ‘be human’.

Time to die? For us writers… not quite yet.