On Large Language Models

The dry, dull, flotsam of artificial intelligence

Yesterday I read a superb essay by

Erik Hoel

about ‘writers’ using Ai to ‘write.’

You can find it here.

It concludes with the following powerful paragraphs:

“As with any new technology, we all have to take a stand, and mine is that I will never feed you AI content slop under my name, be it via a chatbot-wrapper with my photo on it on some other platform, nor here on Substack, nor in my books. My writing will always be the authentic me, the one that is sometimes wrong, the one that occasionally fucks up links, the one who makes mistakes, the one who once wrote “yolk” when he meant “yoke” and everyone got in a pedantic huff, the one that never manages to write a post under 1,500 words; but the one that is trying, really trying, to produce stuff that’s authentic and artistically and intellectually worth your time. I believe the broadcast-based connection between the consciousness of the writer and the reader is a sacred trust, I’m honored it exists between me and so many people, and I don’t plan on swapping it out with an artificial replacement.

So with that tale told, consider this my annual plea to become a paying subscriber of The Intrinsic Perspective to support this choice—and if not to me, then to your other favorite writers/artists/thinkers/creators. Because there are going to be wild, absolutely wild, financial incentives to use AI to scale content creation in ways that slip easily, thoughtlessly, into the harmful and illusory. Most of the big outlets will do it, if they haven’t started already. Subscriptions and connections to real humans you trust are the only way to combat it. We will be what remains of the internet when this is all over.”

That really spoke to me, because as one who strives to write, and to write about what I believe to be important things, I am very concerned about words strung together by Large Language Models being passed off as writing.

Photo by Sergey Zolkin on Unsplash

I have, in the past, used artificial intelligence to do some really specific things.

• I’ve used it as a search engine. Not to replace Google, but to quickly summarize the results.

• When Substack first made it available for image generation to illustrate posts here, I used it to generate a handful of images for my essays, and indeed I generated the little evergreen trees graphic you see in your browser tab above with it.

• I’ve used it to ask myself questions.

I haven’t used it much at all, but I have used it for those things. And I think that those uses are OK. Because I don’t claim to be a Search Engine. I don’t claim to be a painter (although I’ve often wished that I possessed that skill.) And I don’t claim to be a crazy guy wandering the streets while yelling questions to himself.

But, I do claim to be a writer.

And as such, I do believe it would be immoral for me to let Ai do the writing that I was claiming authorship of. For me to pass off Ai ‘writing’ as my own, and sell it to you, for either your money or attention would be fraud.

I have never written with the assistance of artificial intelligence, and I never will.

For if I did, I wouldn’t be a writer.

The vast majority of my writing is over on Emeth, and it’s all about Freemasonry. A really strictly defined niche.

Other people write about Freemasonry too, and that’s awesome. We need more voices within our Ancient Craft, not fewer.

But, a couple prolific Masons don’t write their own stuff. They use Ai, and they just try to pass it off as their own unique creation. Honestly, that angers me. It angers me because they are offering nothing new to the Craft. Whatever they ‘write’ has been written before, because that is how Large Language Models work, by repackaging previous writings. And it angers me because they are claiming authorship of works that they did not author. That is strikingly dishonest, and unbecoming of a Mason.

The thing is, I knew that their work was Ai generated long before I ever took the time to have it analyzed by machine.

Because it is enshittified. Ai can spit out facts, but it is dreadfully dull. Boring. It can’t be anything else because it is created by computer. Human writing is interesting because humans are interesting. Humans have tremendous skills, and horrible weaknesses, and those come through in writing, and they are interesting. Our greatness and our errors, both lend interest. Computers can’t match that, they are simply bland.

People who claim to ‘write’ while regurgitating Ai slop can’t be stopped, and I wouldn’t advocate that any institution try to stop them. They can do themselves, I’ll do me. And I’ll win in the end because I’m far from perfect, I fuck up from time to time, and that is, if nothing else, interesting. That is my edge over the machine.

But we should recognize that these fake writers clog things up. They flood the internet with their content, making the real stuff harder to find. So, they do actual harm.

And this problem will continue to grow in the future. We are in the really early days of it now. What we can each do to fight it is to read actual writers, and to amplify them as we are able. Share their work so that they have some hope of being found in the flood of Ai generated garbage.

You honor me with your time and attention whenever you read my words.

In return I pledge to you that they will always be my words. I will not pump out Ai generated flotsam and put my name on it.