By Kait Carson

Have you been following the news about AI developments? Do you use AI in your work? Is AI part of your pantheon of creativity?

My answers are simple. Yes, sort of, and no.

The sort of pertains to my use of ProWritingAid for spelling and comma assistance. I switched schools in the fourth grade, and the change from phonics to rote spelling methods meant no word ever looked properly spelled again. As for commas, I tried. I really tried. ProWritingAid is one hundred percent responsible for my editors retaining their hair.

As for writers who dump their completed manuscripts into Claude or ChatGPT, or any other AI program for a preliminary edit, I can only shake my head in disbelief. These are the same programs that participated in wholesale pirating of published works, copyright notwithstanding. What in heaven’s name makes you think they’re no longer harvesting your words (a/k/a training) simply because they’re being forced to pony up damages? That seems naïve, but your mileage and tolerance limits may vary, and I’ve been told the developmental and line edits are exceptional. They should be. They’ve used outstanding models for their frame of reference.

Nor do I understand why authors would use AI to ‘write’ novels. Part of the joy of being a novelist is manipulating words, crafting them and then polishing them until they shine, holding them to the light and watching the rainbow prism they emit cover your laptop. Publishers are requesting authors affirmatively state if they used AI to produce their manuscripts, but there is no reliable ‘test’ and the system is honor-based. AI may produce technically perfect words and punctuation, but it can’t reproduce nuances or the author’s heart. At least not yet.

Wikipedia license free

As a member of the Baby Boomer generation, my distrust of AI may be traceable to the movie 2001: A Space Odyssey, and HAL the computer. That didn’t end well as I recall, and Keir Dullea was so cute. Not fair. It may also seem that the movie had provided a glimpse into the future.

Last month, The New York Times and other publications carried stories about a new AI model, Mythos. Anthropic, the company responsible for Mythos’ development, paused the rollout citing cybersecurity concerns. Mythos, it seems, is excellent at finding and exploiting software ‘bugs.’ It also seems it’s capable of exploiting these bugs on its own. While the company has paused the rollout, the story left open the possibility that the model might ignore the pause. Interesting concept.

If that’s the case, how can authors, or anyone, protect themselves from the insatiable appetite of AI? In the spirit of everything old becoming new again, we might want to consider taking a page from our past and penning our stories with….pen and ink. Come to think of it, that would also resolve the question of AI as an author. At least until robots learn cursive, followed by pounding away at manual typewriters. Right, not much of a leap. It’s a brave new world, and a scary one.