A guest post by Adrian Raffill from Caversham Writers
Whenever we read a news headline there’s a part of our mind that simply asks ‘Yes, but how does this affect me?’
Recently a group of Artificial Intelligence experts expressed their concern about the risk of AI by issuing the following statement:
“Mitigating the risk of extinction from AI should be a global priority alongside other societal-scale risks such as pandemics and nuclear war.”
Yes, but how does this affect me?
It might be a while before our very existence is threatened by AI and better minds than mine are clearly working on the problem, or at the very least, thinking about it. But in the meantime, how does AI affect us now? And by us, I mean us as writers.
I wear two writing hats. The first one is a workman’s cap I don in my day job as a freelance copywriter, writing marketing material for technology companies. If I’m replaced in that capacity it won't be by AI - it’ll be by a person using AI. It’s in my own best interests, therefore, to ensure I become a person that knows how to use AI.
My other hat is worn in the pursuit of creative writing (and I like to think of it as a fedora I wear at a rakish and jaunty angle.) The path to publication is littered with the bodies of writers who ignored advances in technology, their handwritten manuscripts piled onto the scrapheap, soon to be joined by the typewriter and the floppy disk. Once a thing has been invented, it cannot be uninvented. It’s here to stay, at least until the next Latest Thing is invented to replace it, so ignoring it means at best we have no say in how it’ll be used. At worst we become redundant.
But does AI fall into this category? This is a technology that changes not so much the way we record or deliver our work, but instead forces us to look at where the work comes from. It forces us to look at the very concept of creativity.
What is creativity in writing anyway? In his book Creativity: A Short and Cheerful Guide, John Cleese describes the processes at work when we create. He discusses the cognitive functions of the left and right hemispheres of the brain and the roles they play in creativity. There is the childlike playfulness and curiosity of the right brain leading to idea generation and the analytical engine of the left brain that shapes and moulds those ideas. When engaged in creative activities, we switch constantly between these two modes of thinking.
In my view, it is the ability to integrate these two ways of thinking that still gives us the edge over anything that current iterations of AI technology are capable of. It comes down to the type of reasoning that machine learning algorithms use - and also what type of reasoning they don’t use.
Early AI systems were based on a method of reasoning called deductive inference. This is a method of drawing a definitive conclusion from known facts:
All pigs have four legs.
Wilbur is a pig.
Therefore Wilbur has four legs.
ChatGTP is the current poster boy for AI. Like most of its counterparts it has moved on from simply using deductive inference and now also uses a form of reasoning called inductive inference, which in essence means knowledge through experience or observation. It’s a way of drawing general conclusions from incomplete data:
Every pig I’ve ever seen has four legs.
Look! Here’s another pig and it has four legs too.
Therefore all pigs have four legs.
Both deductive and inductive inference are examples of left brain logical thinking.
What makes human thinking special, though, is a different type of reasoning called abductive inference. This is the ability to “think outside the box”, to hypothesise and theorise. We guess, we make stuff up based not on what we already know, but what we intuit. We daydream, we wonder, we imagine, and we go beyond the scope of the information at hand:
A pig wanders out into a farmyard.
George Orwell: 1, 2, 3, 4… yep, that’s a pig, alright. They’re quite intelligent actually. In fact, if the animals ever rose up and took over the farm, I bet it’d be the pigs that led the revolution. Ooh, wait! That might work as an allegory for the Russian Revolution. I wonder if I could work that up as a critique of totalitarianism and abuse of power…?
I may have embellished that last example slightly. But there are two things to note about abductive inference. First, it is an example of right brain imaginative thinking.
Second, there is currently no good working theory that explains how the human mind does this. And if we don’t know how we do it ourselves, we can’t teach an AI to do it either.
Intelligence requires creativity. Creativity requires both left and right brain thinking; it needs abductive reasoning too.
We don’t yet have true Artificial Intelligence. We have Artificial Inference, and only a limited form of that.
One final thought. The creative human mind gives us an ability to communicate ideas through storytelling that sets us apart as a species. But for all that, none of us can really know what it is to be inside the mind of another human being, to see the world as they see it. Our attempts at conveying that, at communicating what it is to be ‘me’, is a pretty good working definition of art. One day the boffins might work out abductive inference and we may yet see the world the way an AI sees it, but until then I’ll want to know what it’s like to be in your head. So tell me a story.