Conundrums
in which our hero still doesn't use AI
Dear Readers,
Earlier today, I posted an illustration of my character Willem Shaxbeard (an alternate universe version of The Bard) to the Daz Studios forums for a bit of feedback. An unhelpful commenter, someone who obviously hadn’t read most of the previous posts in the thread, asked why I didn’t just use AI. I didn’t tell the troll to fuck off, though I thought about it. Instead, I said that I was thankful for the feedback but that I had no plans to use generative AI now or in the future.
Those of you who have been with me for a while won’t be surprised by this. What you might be surprised by is how much I sometimes want to use generative AI, but then don’t.
A week ago, I was struggling with one panel of issue #6 of The Blood of Seven Queens for an entire day. One panel. One. Fucking. PANEL. And the reason was that I didn’t feel confident in my ability to illustrate a crowd to flesh out the scene. I kept smashing my brain against the computer screen—one smash after another after another, until there was a thin film of jellied brain cells coating my monitor. Then, sometime just after midnight, I wiped my gray matter off with a paper towel, finally gave in, and just used a different camera angle.
Now, in the early going of that process, I thought about how easy it would be to take the image I’d already worked hard to make, run Photoshop’s “generative fill” on the small area where I wanted the crowd, and ask the algorithm for a bit of help. I stuck to my guns, however—my values—and made a mess of my MacBook instead.
And my head, of course. It’s better now. I spooned everything back in there.
In the end, is the result any better? I don’t know. But when you already use 3D modeling in your process, a tool that some folks view as somehow less-than, there is an incredible desire not to be seen as even more of a lazy hack than you already are.
At any rate, I wish we lived in a world where the ethical problems with generative AI didn’t exist and where a small use of it here or there wasn’t considered just as bad as generating a whole piece from a text-based prompt. Alas, we don’t live in that kind of world. And so, for now, I’ll continue to avoid AI at all costs.
And eventually, I will scour the Internet and watch a whole bunch of videos on how to illustrate big crowds of people—since Writer Chris, that big dummy, keeps writing them into the scripts.
Yours,
Chris
P.S. Not to bury the lede, but the end result of how hard these first six pages of issue #6 have been is that I’m delaying the release until February. I just can’t hit the original November target without going insane.
Or, well, without going more insane than I already am.



I totally get the tug of war between wanting to use generative ai while also wanting nothing to do with it. I read that Adobe’s Firefly generative model is only trained on licensed and public domain content, but who knows how true that is. Difficult waters to wade in. While I have admittedly dabbled with it out of curiosity, I stand by you in your totally understandable stance.