01/11/2023

I suspect I’m not the only contributor to this edition of The Ship who didn’t give a sly thought as to whether they could get an AI chatbot to write their article for them this year. This would be a handy resource in several ways. The fee for this annual writing job is payable in the form of an occasional lunch with the editor (not that this isn’t a charming and enjoyable prospect) and the copy deadline always seems to fall just as the weather’s getting nice and one’s thoughts are turning to the summer holidays. Chatbots don’t eat lunch or go on holiday. Obviously we Ship contributors are dedicated professionals who would never resort to such a scurrilous subterfuge, but I was just wondering: would I have got away with it if I had? As a professional writer I’m hoping not, for obvious reasons.

The development of AI has been as significant as the invention of the atom bomb. Eighty years ago mankind (I can use this antiquated and sexist term, as I believe it was a bunch of men who thought up the idea) worked out how to destroy our physical world. Now we have figured out how to destroy our intellectual world too.

But this was not how it was meant to be. Back when I was a kid in the 1970s we were told that technology would be our friend. Stiff-suited Raymond Baxter on Tomorrow’s World on BBC1 on a Thursday evening would inform us that in the future machines would be the servants of humanity, ushering in a new golden age of leisure. Robots would do our bidding, performing all the boring manual tasks and freeing us up to spend our days sitting around composing symphonies and writing poetry. Unfortunately it’s turned out the other way round. Chatbots can now write poetry and symphonies. And we humans are left with the washing up.

The classic hit-list of human jobs that are vulnerable to AI includes all those nice, sensible middle-class professions that our mums dreamed of us ending up doing when they dropped us off at the Gatehouse all those years ago: accountancy, the law, medicine and journalism. As a journalist and writer I am one of the people whose livelihood is clearly at risk. This is doubly unfair because, as far as we creative individuals are concerned, AI is the second phase of a double whammy perpetrated on our profession by new technology. Ever since the invention of printing, if people wanted to read stuff they expected to have to pay for it, in the form of a book, a newspaper, a magazine, a theatre programme etc: all these things cost money. But then at the end of the Twentieth Century along came the internet and took away a lot of the paid gigs from writers. Suddenly there was all this free stuff online. What was the point in paying for writing? It’s now got to the stage that no person aged under forty would consider buying a physical newspaper. Why would they want to pay money to read yesterday’s news, when they can get today’s for free online?

All that was bad enough, but now AI has crashed the party too and driven the second, fatal nail into the writer’s coffin. Many of us hacks have already been reduced to writing at a knock-down rate, or for free, but now there’s no need for us to even do that any more: because bots will also do it for free and instantaneously to boot.

But before I consign my own profession to the scrapheap, there is a small crumb of hope, which is expressed in something else which has been worrying me over recent years. As I’ve got older it’s started to bother me that I can’t teach what I do for a living. It’s not so much me feeling that I have a mission to pass on the skills I possess to the next generation, but more about me being able to earn a bit of money when I’m too old to do anything else. Various friends of mine are drifting into semi-retirement (no one seems to actually retire completely these days). In the main these are the sensible people who followed their mum’s advice all those years ago and got proper, quantifiable skills as bankers, accountants or lawyers. They now spend an increasing proportion of their time sitting as non-executive directors on company boards or working as business consultants or life coaches. They have a transferrable expertise that is in demand and can be cashed in.

But I write humour for a living. You can’t teach that. Okay, I know that creative writing courses exist, but I think their main utility is giving wannabe writers a deadline to produce stuff by and encouraging them to show their work to others. I can’t teach what I do, because I don’t even know how I do it myself. What happens is that I sit down at a desk each morning in a boring room with a cup of coffee and a blank sheet of paper (or sometimes a blank Word document) in front of me. Then I start writing stuff. I write for a while until I get something that’s good (usually I have to replenish the coffee cup quite a few times in the process). How can I tell when I’ve got something that’s good? I don’t know.. I just know. It feels somehow connected. There’s a tension there. It’s a gut feeling that’s impossible to describe. Of course other people may not agree that what I’ve written is good, but that’s a matter of opinion (unless that person is my editor, in which case it’s a problem).

But now in the face of the rise of AI, I can see the advantage of the fact that what I do for a living is unquantifiable. It means that it can’t be reduced to an algorithm. AI can’t write jokes, because although it can spew out prose much quicker than any human writer, it has no judgement. It can’t tell what’s interesting and original from what’s boring and obvious. Or maybe different things are interesting and boring to machines than to humans. So, on balance, perhaps my job is safe, at least until next year’s edition of The Ship. Or till the machines decide to stop bothering to do what we humans want them to do and start doing what THEY want instead…

In the interests of balance and avoiding a human-centric bias to creep in I have decided to let ChatGPT have the last word and to write the final paragraph of my piece, explaining why AI is not a threat to humanity. Here’s what it had to say:

“AI is not a threat to humanity, primarily due to the crucial role played by humans in its development and application. As a tool created by humans, AI systems rely entirely on the data and instructions they are provided. They lack consciousness, intentionality, and emotions, which are fundamental aspects of human decision-making. AI is designed to assist and augment human capabilities rather than replace or overpower them. Additionally, ethical frameworks and regulations are being developed to ensure responsible AI development, deployment, and use. With proper oversight, AI can contribute significantly to areas such as healthcare, education, and scientific research, enhancing human lives and fostering societal progress.”

Ah yes, but it would say that, wouldn’t it?

Russell Taylor MBE (St Anne’s 1979)