I’ve been writing on the Web since 1995.

Some years, I was prolifically writing.
I’d post and post plus distribute my content across social media platforms and more. But over time, my writings became less and less.

Why is that?
Some of my pauses were due to the writer’s block that writers are prone to from time to time. Other pauses were more related to having too tight of a schedule with little room left for writing.  But most times, it was because whatever time I did have to write, it was often spent writing for clients (and their campaigns). Somehow, across the years, I guess I kind of lost a bit of my own writing mojo.

And now, sigh, AI is here.
Unless you’ve been sleeping under a total rock, generative AI is all the rage these days, especially for writing.

While AI tools have their uses, their increasingly powerful abilities are, in large part, made possible by two basic but key components:

  • insanely huge amounts of processing power (which most of us have absolutely zero control over)
  • the training(s) behind their large language models (LLMs)

The second bullet point is where we all come in.
While most of us have no control over the galactic surges of processing needed to power today’s AIs, we do have *some* measure of “control” (no matter how slight) over how the LLMs get trained.

Allow me to explain …

  • For an AI to work, it needs an LLM to train it.
  • To train a single LLM, zillions and trillions of words and images (and more!) are voraciously needed.

In this stead, AI companies will suck up every syllable and image pixel from the Internet (for starters) with insatiable and bottomless appetites, and even this vacuuming up of the world’s content many times over is hardly ever enough so they have to hit up numerous other offline sources just to train one single LLM.

In such content-gobbling quests, ethics and copyright often go out the window.
Search up this topic and you’ll readily find countless lawsuit after lawsuit (filed by some of the world’s most celebrated authors and creatives) against these AI companies for copyright violations well underway.

I’m not really sure how I, nor we, can solve this gargantuan, growing, and real challenge.
With these contexts and expanding AI climate, I’m much more mindful today of what I post because the moment I hit the “Publish” button — either here and pretty much on any other platform — I’m essentially ringing the content dinner bell for AI LLMs to come on over and gorge on all my mental contents, produced from my organic brain and allegedly protected by copyright.

You, too, face these same obstacles; in fact, we all do.

For these reasons, I myself have to think (more than) twice about what I’m willing to share or post online.
These reasons also inspired my decision to pull my content back a bit, as much as I reasonably can, when writing online because knowing that some AI’s LLM is going to suck it all up and monetize it without me (and your content without you, too) feels violating and discouraging.

That all said, I’m a visual communicator and I love to write and produce art.
So I can’t and won’t stop writing or designing but I don’t feel I can do so as freely as I once did. And while I realize that PDFs and ePubs are not protected formats, at least I can attempt to (package and) monetize my own creative contents in parallel to the LLMs just usurping it (often without regard to my own, and your own, livelihood(s)).

Measured efforts are better than none
Moving forward, I’ve decided to keep on writing (and sharing some my art) but I’ll only share just enough to express an idea or get a quick point across. This measured attempt to limit and safeguard my content, I realize, may be feeble — AI companies will continue to content-pillage without regard nor bounds; but some effort is, to my mind, better than none.

Until the next post,

MAYRA YADIR

Visual Artist • Creative Writer • Indie Publisher

See author's posts