I’m pretty confident we are well past the dawning of the age of AI. We’re most likely somewhere in the lunch hour or even past noon. Perhaps we’re in the afternoon siesta–that time of day when we’re all taking a nap indoors or in a hammock under the shade, while the AI server farms are busily planting, weeding and harvesting. They don’t take naps.
There’s a plethora (yeah, I consider The Three Amigos to be one of the pinnacles of cinematic history) of books these days on Amazon that seem to be AI-generated. Self-help books, how-to books, all the way through to oddly written fiction books. AI? Perhaps, maybe, possibly, a great deal of them probably. Though, the oddly written fiction books are either the product of (very confused) AI or creative writing graduates who are intent on winning the Booker or something equally “prestigious.”
You gotta be careful these days with “prestigious.” Definitions are definitely changing. Bob Dylan forgot to write that verse. The definitions, they are a-changin’…
I just saw a fairly alarming ad on Youtube for an AI writing program that will apparently take a simple prompt and craft it all the way to a finished book, complete with formatting and illustrations. The ad was focused on non-fiction, but I imagine the program could tackle fiction as well.
We live in a quickly changing world. Some days I wonder if it’s time to go live on a mountain top in Idaho. Off the grid. Dump my cell phone. Buy a library’s worth of physical books. A modern day Stylite monk.
But, that wouldn’t work. I imagine a government inspector would show up soon enough. “Excuse me, do you have a permit for all those books?”
I don’t imagine ever having an AI write my stories, though there are some sensible tools for editing, creating marketing material and things like that. Stories, from a writer’s perspective (and I mean this writer, as in me, myself and I), are almost like an extension of the soul. Or maybe a better way to put it, a shadow cast by the soul. How could you turn that over to an AI and expect it to walk out into the metaphorical sunlight and cast a shadow that genuinely represented you as a soul-ful human being?
What I do intend to use AI for on the actual creative side is for music and video.
Music and AI
I’ve written dozens of songs over the years. I don’t just write the songs, I record them in multi-tracks of voices and instruments. That’s easy enough to do these days with a computer, a decent audio interface and some recording software. It’s easy in terms of inexpensive gear. It’s not easy in terms of writing a good song. And it’s definitely not easy in terms of pulling off a good recording with both professional execution and professional recording quality.
That’s where the AI comes in. There are quite a few options, but I’ve been experimenting with Suno. Suno is both amazing and depressing. Amazing due to quality, depressing due to how it will certainly put some musicians and engineers out of work. Also depressing in that, due to its high quality, it might make kids think twice about putting in the time and effort to develop their own skills.
The ability to create is one of the most unique traits that humans have. It sets us apart from animals and reflects the fact that we are made in the image of God. Anything that discourages creativity should be, in my opinion, suspect, because the reduction of creativity reduces our humanity.
I think…
Video and AI
I used to work in television, many long years ago, so I have a painful appreciation for how much work, time, personnel and money goes into making a show. I have zero interest in getting back into actual show production, but I am intrigued by the option of creating short, one-off productions on my own with AI. Either just for fun or to marry to a song and put that out in the ether (also just for fun).
After some reading and investigating, I’ve decided to try LTX. They have a robust work space in terms of tools and work flow. Flexible controls, fairly specific prompt results, built-in-editing tool (very helpful to have that bundled together). Not a bad price either.
Anywho, once I come up with results good enough for public consumption, I’ll put them up here. And, in the meantime, I have a book to finish. Almost there.
Update: After experimenting with both LTX and Neural Frames, it’s pretty clear that current AI video models only work if you’re able to put excessive time into them and spend large amounts of credits on your projects. Even at a mid-tier subscription plan, you’ll burn through your monthly allotment of credits in a few hours of tinkering. Not worth it at this point. Both of those outfits are at the top of the heap, apparently, but they both leave a lot to desire. For example, if you want to create a video of someone simply driving down the street in a car, you very well might have to hold the AI’s hand in order to convince it that, yes, the driver needs to be seated behind the steering wheel and not just in the passenger seat.
Thus ends my AI video experimenting of early 2026. Perhaps I’ll investigate again in 2027. Depends on the innovation rate, exponential or otherwise.