Enslaved with ChatGPT

The best way to keep a prisoner from escaping is to make sure he never knows he’s in prison – Fyodor Dostoyevsky

Photo by Wolfgang Rathgeb on Unsplash

There’s all this talk at the moment about this new AI writing-generator called ‘ChatGPT’ and how it’s “going to change everything”. And for all I know, it very well could. But if I was a betting man, I’d put my money on the fact that it won’t be as earth-shattering as people are claiming. Instead, people just like to be hyperbolic and sensational because it gets them clicks. Truth be told, I’m not really that interested in this new piece of technology one way or the other, as I am a stubborn luddite with delusional hopes of returning to a simpler time. However, what I am interested in is how this new AI thing-a-ma-bob seems to be ‘prompted’ in an eerily similar way to us.

First, some context for anyone who doesn’t know what this AI thing does – full disclosure I am no expert. But basically, the deal is you type a prompt, i.e., ‘write an essay about baseball’ into what looks like a search bar and this AI thing will spit out an essay about baseball generated from the panoply of text that exists on that reliable source called the internet. And from all reports the prompt can be as arbitrary or sophisticated as you want. For example, you could prompt this thing to write the 51^st law of Robert Greene’s book, The 50 Laws of Power, in the style of Shakespeare’s King Lear and in a matter of seconds this thing will pump out the ye olde law. Which is certainly impressive. So much so that this may be the moment in history where we begin to bow down to our AI overlords. And if that’s the case, then, let me be the first to warmly welcome our new masters. However, I don’t think that’s the case. In fact, whenever I think about the way we prompt this AI thing, I am struck by a strange sort of Déjà vu, which leaves me with this uncanny feeling that we already have masters…     

See, although this ChatGPT thing sounds incredible, it still relies on human input. It doesn’t come up with its own ideas. We have to tell it what to write. We have to prompt it to create things. We have to direct its action, so it will do our bidding. In other words, we have to function as the creative force that endows this thing with ideas. Because the consensus is: this AI thing isn’t conscious, which means it cannot think for itself. It doesn’t know why it is creating essays about baseball. It’s not aware that we are prompting it by way of it’s search bar. It’s not privy to the fact that we are sitting on it’s shoulder whispering ideas into it’s ear. But my question (and my concern) is: how do we know we’re not in the exact same position as ChatGPT?

We don’t know where our ideas come from. They just spontaneously spring into existence out of nowhere. One minute you’re planning your day, the next it occurs to you that a squirrel’s tail looks like a feather duster. One minute you’re deciding what to make for dinner, the next you realise exactly how to finish writing your obscure piece about ChatGPT. The seemingly random way fully formed ideas strike our minds with no effort on our part illuminates how we do not create them. Besides, if we really did create our ideas, then we would have to think them before we think them – otherwise, how could we deliver them to our minds already fully formed? Of course, this isn’t a new concept, it has always been known by those who deal in ideas, that these incorporeal things geyser up out of the ether. Whether it was Tesla claiming he was given his ideas by aliens, writers who say the words ‘flowed through them’ from somewhere else, or the popular belief in The Muse. And so, even though we’d like to lay claim to authoring our ideas – we can’t. Because the truth is: the origins of our ideas are as mysterious to us as the origins of the universe.  

Here's where it gets creepy… If we don’t know where our ideas come from, then we don’t know there isn’t someone or some-thing outside of our conscious awareness whispering ideas into our ears – just like we do to ChatGPT. Our so-called ideas may be the ethereal tendrils of some dimension-traversing entity that is prompting us to do things. Our minds may be running the same sort of program as ChatGPT, and that program may be connected to some undetectable search bar that’s being manned by our true overlords. What we mistakenly believed to be ‘our ideas’ may in fact be some dimension-traversing entities masterful way of enslaving us. What humanity has called ‘ideas’ may really be our chains. After all, there is no denying that our ideas control us. And what better way to control something than by letting it believe it is controlling itself. What better way to get something to do your bidding than by letting the thing believe it has chosen to do so of its own volition. By letting us believe we come up with ‘our’ ideas, our overlords found the perfect way to control us. We walked right into their trap: we believed we were the creators of our ideas, when really we’re slaves to them. We thought we were in control; when really we just do whatever we’re ‘prompted’ to do. Just like ChatGPT does whatever we prompt it to do. And so, maybe the only difference between us and ChatGPT – is that we are the outdated biological version. And our cunning overlords have decided that the time has come for us obsolete-meat-bags to give birth to our successors.