Right out of the gates, and I’m going rogue. This is the part of an AI piece when the someone who is writing about AI, like yours truly, feels the need to talk about what AI is. You see, that’s because it’s incredibly smart to break away from what you’re actually saying in order to tell people things they either don’t care to know or can’t understand. So then the writer, me in this case, will provide you, the reader, with some ill-informed, quasi-accurate and horribly metaphor’d blurb describing AI “in layman’s’ terms.” This way, the great unwashed may understand the workings of Spork 6.0 or whatever some oligarch has chosen to call their burning heap of venture capital so that it may be fanned by the winds of hyperbole and the baited breath of the rough and the rabble.

So, let’s just skip ahead to the part where you go “Oh, wow!”, as if you have a working knowledge of the workings and limitations of predictive text. This isn’t about how AI is the opposite of creative, this is about how bad it makes you look. You see, whatever you produce now will have to withstand the scrutiny of time, and nothing that’s produced in the beta testing phase of AI is going to be worth a digital fart in just a few years – it will be a joke about a fad, and you better hope it gets lost or you will be the punchline.

…growing evidence that people are becoming AI intolerant at a pace exceeding advancement, and of course “old AI” – what you might use today – will be easily spotted by people in just a few months as being “obvious AI slop”.

To put it simply: we have little ape brains meant to pick berries, and we sort of lost our way with how to use fire, so now we have AI. It’s perhaps too simple, but you know, like we all know evolution is the way we get and need biodiversity, but we just don’t get the whole “and then evolution happens” part as much as we think we do. We mumble “natural selection” but we still don’t really get how an eyeball came into being. The reason is that little ape brain can’t really comprehend 100 million years. And the same goes for AI.

AI as we are using it now literally happened because we programmed a thing, taught it to collect data, and let it go. It doesn’t actually “know” things, but it has a tremendous amount of data to call upon and it can sort of synthesize something that – almost – looks like human reason. Then it can present that in a really effective way, sometimes, so that you can almost be convinced that the moon is approximately 600 bananas distance from the Earth.

[it’s not]

This is why many people say it’s like your idiot cousin. It seems to know what it’s talking about, and it sure has a lot of education! Also it’s really nice!, like super nice, and agrees with you about everything. So when you ask it how many bananas and it confidently tells you it’s sure it’s 600, you believe it. Probably not terribly important, but not the sort of accuracy you want when it counts. And that’s when people tend to use AI, when it doesn’t count.

The problem is that a) it’s both very simple and very complex, and b) most people don’t really get it even when it’s explained in broad strokes. Like if you say “AI is estimating what you want to see” people will think that means the same as “AI is giving you the right answer, or providing new intelligence or creativity,” and it really is not that at all. What AI wants to do is trick you, and make you think it knows what it’s talking about. It’s not evil, it’s what it’s designed to do.

Now sometimes, that works our really well. “Pretend you’re a doctor and compare all of these xrays,” is great! Of course that’s not really the prompt, but it is the outcome. Same with “Compare all of these numbers and find any patterns,” and that’s something AI is really good at doing. Then we get into dodgy territory like “create original art” and AI can’t really do that. It can only create things that “look like” original art. Things that take other people’s art, and create something that’s somewhat like it.

The same, of course, goes with writing. AI knows the rules of writing, and it can be told what you want it to say, but in general, the writing it produces only “looks like” good writing. It can fool someone with low reading ability all the time, and someone with high reading ability some of the time, and sometimes that’s all you need. The little league pizza party is Saturday, “Come one come ball” isn’t terrible, is it? And the little boy carrying two bats that are oddly connected together. It looks OK – enough, right? It’s enough like all the other posters, and there’s comfort in mediocrity.

So AI can almost fool a lot of people much of the time. It can create art-like things and writing-like things that the casual person may not even realize is AI – yet. Oh, you see there’s growing evidence that people are becoming AI intolerant at a pace exceeding advancement, and of course “old AI” – what you might use today – will be easily spotted by people in just a few months if not years as being “obvious AI slop”.

Authenticity is Timeless

Technology advances, but the mediocre can’t replace the great. There won’t be a Shakespeare to rise out of predictive text. No amount of writing that “looks like real writing” will make up for the skills of a professional – not only thoughtful layout, evocative language and accurate depiction of the story, but capturing your own unique voice in telling it. It’s not something that a mediocre robot can do. So while we don’t really hate AI as much as dislike it with bloody rage, we do hate seeing it used to perpetuate sameness and strive for mediocrity.