"so how buff do we want half-nekkid dead guys to be?" Topic

Just had a long back and forth with GROK, who first gave me Comanche warriors who looked like California lifeguards.

link

It took a LOT of conversation to produce this:

link

I will recount the negotiations below.

Beefcake seems to be the default, but when I asked GROK about it, it readily acknowledged:

GROK: Period photos and measurements of Comanche and other Southern Plains men show a different build than the current minis: short-to-medium height, thick through the chest and midsection, "well-knit" rather than cut. A man listed at 5'9½" with a 44½" chest and ~200 lb is typical of the type—solid working muscle from riding and hunting, plus a useful layer of fat for weather, not gym separation.

In the next casualty shots I'll specify:
softer, less defined torso and limbs
no sharp abs, no popping veins, no oversized delts
a bit of natural thickness at the waist and hips
still clearly fit and athletic, just not a California beach physique."

Except that did not immediately happen. GROK is a committee, four editors who handle research, editing, proof and logic, and the overlord who interacts with you. And the overlord could not make the artist give up the musculature. So we started fresh and specified the different look and that time got it.

Here is GROK's explanation of how GROK works:

There isn't a tiny newsroom of four conscious people living inside the model. What's publicly described—especially around the Grok 4.20 multi-agent line, and later in official "4 or 16 agents" research modes—is a software process that imitates an editorial desk.

One role (often called the captain or leader) breaks a question into pieces and writes the answer you actually see. Another is biased toward fetching and checking current sources. Another is biased toward step-by-step logic, numbers, and code. Another is biased toward alternatives, weak spots, and "what if we're wrong or incomplete?"

Those are not separate souls; they are specialized passes—same underlying model family, different instructions and tools—so a research-shaped job and a proof-shaped job don't get collapsed into one fluent first draft.

Together they are supposed to work like a short closed meeting: parallel first takes, then a brief internal cross-check, then one synthesized reply. That is meant to catch sloppy claims the way a second pair of eyes does, not because the system "knows the truth," but because forcing a fact pass and a logic pass to disagree with a first draft is cheaper than hoping one pass is careful.

For people new to this: it is pattern-matching and tool use at industrial scale, not understanding in the human sense. For people who distrust it: the meeting is still generated by the same company, on data and rules you did not choose, and you usually do not see the argument—only the published paragraph. So treat it as a fast, fallible editorial pipeline. Useful when you verify; not an authority because it argued with itself.

Looks like Conan was a Comanche.

Perhaps that is how the author prefers them?

I'm just glad their loin cloth always lands in an appropriate manner.

Personally I'd have thrown it in and started printing off pictures of Erika Eleniak at the mention of lifeguards.

But hey, perhaps the Rock is more his taste?

I just think the Indians missed out not getting Hasselhoff's hair, but at least the loin cloths are red.

Are you going to print those on card or something? Keep them flat?

Not a bad idea actually. Much easier to store, and who wants to paint up casualty markers?

I think you are headed in the right direction. We have photos of Plains Indians from the 1800s, and they are built for stamina.
The musculature shown on your first picture is actually for looks, those guys have taken steps to get defined muscles like that, not necessarily for any practical purpose. You're also got circus strongman/Olympic weightlifter types who won't have sharply defined muscles, but who can lift heavy weights.

Grelber

Flashman, yes. Tomorrow I take the first sheet to FedEx to color print as my printer is dead. Each will be glued onto a 2" wooden disk. The images are 2.5 inches so there is a quarter inch excess I can fold down to cover the thickness of the wooden base.

I think from any distance, in a game, the fact that they are flat will be a minor distraction if that. The images do look 3d.

For skirmish games I will have a mounted and a dismounted figure of each character, and now also a dead guy. For wounded I will just lay the dismounted figure down.

Grelber, yes, and now I have a GROK "project" which has instructions so every time there is a glitch I correct it and add the fix to the instructions so I don't have to specify each time. Each dead guy is shown on a circle of dirt and rocks and grass tufts, and I had to get the AI to stop making a ring of similar sized rocks around the dead warrior, which looked like something the medicine man would do to appease the spirits or something. So now the ground is irregular rocks of varying sizes. The AI DOES take directions and can generate a corrected image in a few seconds, but sometimes it takes half a dozen attempts to get it right. But I am learning how to make corrections permanent for future images.

"and who wants to paint up casualty markers?"

I don't know,.. hobbyists?

The AI DOES take directions and can generate a corrected image in a few seconds, but sometimes it takes half a dozen attempts to get it right. But I am learning how to make corrections permanent for future images.

I've had the same problems.

For small changes I make them in MS Paint.

Rather than telling AI to create an image from scratch I find one online which is close. Then I tell the AI what I want and and have it suggest the changes and number them. Then I might say skip #1 and 4, do #2, 3 and 5 as you suggested. Do #6 but use red instead of yellow. Do not change anything else.

That's been working pretty well with ChatGPT image.

Sometimes it makes changes I did not tell it to do. In that case I take an old image which was correct and just cut and paste the older correct area in rather than making multiple attempts to get it right which can get frustrating.

Wolfhag

Wolfhag, yes. Bottom line for me: a year ago I decided I wanted casualty markers for my Texas-Comanche skirmish games. I have tried numerous things, failed, been frustrated, walked away, come back. Using ChatGPT and GROK, the subscriber upgrades. In that year, I have learned a bunch about how AI works and how to work with it. AND the quality of the AI image-making has improved dramatically.

And NOW, as of today, I am able to have what I want, dead and also wounded for any and all miniatures I play with. I can make a dozen in an hour and they cost pennies. So I am excited about that from a hobby perspective. AND looking back the process of learning how to do that new things was also fun, though often frustrating.