I’ve been keeping this blog for over a decade now, posting something I’ve written most every week. Over those years, blogs have faded from prominence on the internet — the steady pivot to video that was already taking place a decade ago has firmly ensconced with video essays and whatever it is that’s raging on TikTok’s algorithm. Blogs are quaint, writing isn’t a way to get views or shares.
Enter generative AI.
Large language models like ChatGPT have burst into the mainstream in recent years, promising a new, easier way to churn out stuff. New art, new articles, new blog posts, all with the typing in of a prompt. It takes the work out of making stuff. There are an array of issues with generative AI, from its questionable source provenance to its tendency to just make stuff up. And that stuff it makes up is, more often than not, just not very good. All the same, proponents of those LLMs revel in how much easier it is to make stuff now. No longer will writers, artists, and other creatives suffer under the tyranny of the blank page: it can all be done with a snap of the fingers.
So why then do I bother with this blog? If you want someone to yammer on about capitalism in Star Wars or The Lord of The Rings as an anti-war text or deciding what race to classify yourself as surely one could just ask their LLM of choice and have it spit out a halfway coherent screed. Granted, its logic may not flow and it may cite nonexistent sources, but for informal writing like this it’s just the vibes that matter, right?1 Just think, if I utilized an LLM I could start cranking these out, posting a rant essay every day, replete with excellent SEO to maximize readership. Hell, I could even make a buck off of this blog.
But… why?
I don’t keep this blog because I’m trying to make a lot of stuff. Having a language model make these posts for me is confusing the result with the process. I write this for the sake of it, not for the finished thing. I write this because I’m exercising a muscle, keeping my brain in the habit of thinking closely about the media I consume. I write this because I like it.
I used Grammarly to double-check my grammar until it started telling me how to rephrase chunks of text, rewriting my post to conform to whatever the glorified autofill machine deemed normal. But that’s not what I want my writing to be, nor is it what I want your writing to be. These posts are my thoughts explored and written out and when I read something by someone else I want it to be their thoughts explored and written out. Writing (and art, and storytelling, and so much of life) is communication — I want to hear what the other person is saying, not what an algorithm thinks they want to say. I want their idiosyncrasies and turns of phrase.
When Apple Intelligence showed up on my phone it would suggest responses to text messages. Get out of here; even if the suggested message was near identical to what I was going to write, I don’t want it to text my parents for me. I’m texting them because I’m keeping in touch, because I want to connect. I don’t want to outsource that and I don’t want someone else’s messages to me written by a machine either.
So, why.
I write this blog to think out ideas in my head (some more fully formed than others) and you read this blog because, for some reason, you’ve decided you find those ideas mildly interesting. Generating it via a Large Language Model takes the fun and the soul out of it, reducing these blog posts to just content smeared across the internet. I write — and read, and play video games, and look at art, and all those beautiful acts of existence — to connect with others who share this world, not the algorithm’s calculated text.
That and, at this point, it’s just inertia that has me typing something out every Saturday.
- Absolutely not. I am being facetious. I put effort into trying to get things accurate on this blog. ↩︎
Essays, Not Rants! 