On Saturday So can you tell... did I write this column?

SOMETIMES newspaper columnists – and just about everyone else who writes for a living – will sit down at a blank screen and wonder how to fill the white space in front of them. Most of the time that screen doesn’t stay blank for long, but occasionally there’s a problem getting started.

Which is happening to me right now, because my charming 14-month-old grandson Teddy is wanting to play and, as I don’t see him often, I’m keen to indulge the little lad.

So I turn to my robot friends for help. Like millions of other people I have been experimenting with the new Artificial Intelligence platforms and have been impressed – and scared – by what they can do.

So I type the following instruction: “Write an amusing short story about a grandfather going to see his baby grandson.”

It took four seconds and this is what the robots replied: “In a quaint town, a spry grandfather named Herbert, known for juggling cats and playing the tuba while riding a unicycle, prepared to visit his baby grandson, Timmy. Excited, he embarked on his journey aboard his vintage penny-farthing bicycle, adorned with bells and ribbons.

“As Herbert wobbled through the bustling city, his appearance drew attention. A group of laughing teenagers inspired him to give a show. He juggled apples and played a borrowed tuba while cycling, mesmerising them. Realising he was late, Herbert hurriedly continued to Timmy’s house.

“Upon arrival, Herbert, out of breath, met his grandson. Enchanted by Timmy’s wide-eyed wonder, he juggled apples and balanced an egg on his nose. As Timmy giggled, Herbert knew he had found his greatest fan. Thus began the legendary tales of Grandpa Herbert and baby Timmy, delighting all for years to come.”

It’s not going to win a Nobel Prize, but it only took four seconds.

If I am being honest, I did start by seeing if I could cheat altogether... “Write an amusing newspaper column in the style of Martin Hesp about the experience of being a grandfather.”

This caused the robots to struggle a bit. It took a full two minutes to pen a 900-word column – and I rest easy (for the moment) in the knowledge it wasn’t good enough. For a start, it wasn’t in “the style of Martin Hesp”, being extremely glib and filled with cliche. Which, of course, some readers may think is exactly my style!

Here’s the first paragraph... “They say time flies when you’re having fun, and as I find myself in the midst of this maelstrom called ‘grandfatherhood’, I can certainly attest to that. One moment, I was swapping diapers and lullabies with my own progeny, and the next, here I am – a certified grandfather with two delightful little terrors, I mean grandchildren, who’ve invaded my once quiet and predictable existence.”

Not bad, but a bit formulaic. And it soon gets into what could be described as its ‘Disney-fied’ stride...

“There was that time when my granddaughter looked up at me with those big, innocent eyes and asked, ‘Grandpa, are your teeth real?’ I hesitated for a moment, caught off-guard by the question, before replying, ‘Well, darling, some of them are.’ She considered this for a moment before grinning and saying, ‘That’s okay, Grandpa. I still love you, even if you’re a little bit broken.’

“Ah, the sweet, sweet honesty of children. I must admit that I have now found myself examining my reflection in the mirror with a more critical eye, wondering if perhaps it’s time to invest in some Polygrip or, heaven forbid, a new set of dentures.”

It goes on, tugging at plenty of stereotypical heartstrings. Too obvious and, as I say, formulaic.

And of course it is. Because the Artificial Intelligence algorithms at work in systems like ChatGPT can only sift through the billions of words which have been written before and attempt to make assumptions on the rapid lessons it learns.

As it fine-tunes itself, though – or rather, as the human user learns to fine-tune the instructions he or she gives to the machine – so the results will improve. And improve beyond measure.

At which point – as I wrote a few weeks ago when focusing on the subject of AI – the editor of this and every other newspaper will say: “Why should we pay Hesp when we can get these columns for nothing?”

I would like to think we’re quite a way from that moment yet, because I truly believe that so far the robotgenerated stuff lacks a real and genuine sense of humanity. And I think this would be noticeable, and therefore irritating, to readers. Indeed, I believe anyone with a brain would find it highly conspicuous and so rapidly turn the page.

But for how long?

This week I heard some learned professors talking about the AI revolution on BBC Radio and there was a consensus that as many as 300 million jobs could be lost in Europe alone over the next three years.

All those paid jobs taken over by unpaid robots! That is scary. What are all those people going to do? How are they going to find the means to live?

Maybe I’m too old to worry about my own future, but what about those little grandchildren of mine? Maybe I ought to ask ChatGPT how their story is going to turn out, once the robots have taken over.