AI Art and Gospel Stories: Or, Children’s Book Illustrators are Now Unemployed

Stephen C

•

For what seemed like forever the moat protecting the jobs of illustrators from AI was the fact that it was hard to nail down consistent characters. You could maybe, with clever prompting, get one frame to kind of look like the other, but it didn’t really work, which is why a lot of early AI-storyboard uses were for stories that didn’t require a lot of scenic or character consistency (like my Midjourney edition of the creation story).  

Well, that moat has now been bridged. Both Google Gemini and Chat-GPT have more or less resolved the character consistency issue in the past week, so now everybody can create a children’s book, graphic novel, or storyboard for any story they want, whereas before such creative products were rare as they had to wait for scarce time and resources. The Chat-GPT version can also render things like “gold plates” or “the First Vision,” more or less accurately, which no other AI system has been able to do before. Plus, it can basically most things you can do in photoshop using text commands.

So, with that, I give you my graphic novelization of the translation of Moses (the Latter-day Saint/Josephus ending). If you look carefully there are a few inconsistencies (e.g. his cape is shorter in the caption with God, sometimes the text box has quotation marks, sometimes not), but nothing that couldn’t have been resolved with some more attempts, and for the most part it’s working. All that I did here was use the following prompt: 

Create a series of image prompts to create a graphic novelization of the scene at the end of the Torah when Moses looks out over the promised land, except in the end instead of Moses dying have him be taken up into heaven. Make it in a graphic novel style like Zach Snyder.

It then created the prompts below, which I fed into Chat-GPT with the meta-prompt: Now do this one. Again keep the character and style consistent. I’ve placed the actual images above the prompts. There was one I had to redo since it got the text box wrong, but otherwise these were one-shot. 

 

 

Here is a series of image prompts to create a graphic novel-style depiction (à la Zack Snyder) of the final scene of the Torah, with a visual twist where Moses is taken up into heaven instead of dying. These are meant to guide the generation of cinematic, high-contrast, stylized frames in a graphic novel format.

 

### ? Scene 1: The Ascent to Mount Nebo

Prompt:  

A graphic novel-style panel in the cinematic, moody style of Zack Snyder. Moses, an aged but powerful figure with a flowing beard and desert robes, climbs a craggy path up Mount Nebo. The sky is overcast with a storm brewing, deep shadows contrast with beams of light breaking through. Winds whip his cloak. Text box reads: “And Moses ascended the mountain, as the Lord had commanded.”

 

—

 

### ? Scene 2: The View of the Promised Land  

Prompt:  

A wide, epic shot from behind Moses, silhouetted against a breathtaking vista of the Promised Land: green valleys, distant rivers, golden light breaking through clouds. Graphic novel style with dramatic lighting, lens-flare-like rays of divine light. Caption: “And the Lord showed him all the land…”

 

—

 

### ? Scene 3: The Divine Voice  

Prompt:  

Close-up of Moses’ face, tears welling in his eyes as he listens to a voice from the clouds. Lightning flashes around him, but his face is serene. Stylized biblical lightning and cloud swirls surround him in dramatic visual effects. Caption box: “You shall see it with your eyes, but you shall not cross over there.”

 

—

 

### ? Scene 4: The Moment of Surrender  

Prompt:  

Moses kneels in prayer, arms lifted toward the sky. A glowing figure—suggested, not fully seen—descends in a blaze of light. The environment darkens, except for the divine glow. Panel caption: “But the Lord was not finished with His servant…”

 

—

 

### ? Scene 5: Taken Up  

Prompt:  

Moses begins to rise into the air, surrounded by radiant golden light. His robes flow as he lifts, arms open, eyes closed. The sky behind him parts with divine light pouring through. Epic Snyder-esque slow-motion panel. Caption: “And Moses was taken up unto God.”

 

—

 

### ? Scene 6: Empty Mountaintop  

Prompt:  

The summit of Mount Nebo is now silent and empty. Wind blows across the rocks. Moses’ staff stands planted in the ground, glowing faintly. The clouds part gently. Caption box: “And no man knew the place of his burial, for there was none…”

 

—

 

### ? Scene 7: Legacy of the Servant  

Prompt:  

Final panel. A wide shot of Joshua standing before the people of Israel in the valley below, the Promised Land behind them. In the clouds above, barely visible, a ghostly outline of Moses watches in peace. Text box: “And there arose not a prophet since in Israel like unto Moses…”

 

—

 

 



Comments

7 responses to “AI Art and Gospel Stories: Or, Children’s Book Illustrators are Now Unemployed”

  1. Jonathan Green

    Here’s one semi-counterargument. The techniques for creating pictures may change, but having people with aesthetic sense and the experience to create good pictures and distinguish good from bad is still necessary. I think your AI-generated pictures look great! But my standards in visual media are low and unrefined, and a trained illustrator would be a lot better at creating compelling pictures than I would be.

  2. Stephen C.

    Oh no doubt. I sometimes lurk on Midjourney Facebook pages where people help each other with prompts, and by far the best AI artists are traditional artists. You can tell from their prompts that they know exactly what they want and they know the vocabulary to describe it, so it’s not that it doesn’t require creativity, it just gives you the tools to scale it. We now have access to exponentially more art across a range of themes and styles that you couldn have imagined two years ago. Some may argue that the AI art is inherently worse, but I’m willing to bet that a skilled artist using Midjourney would not be discernibly different in a “blind taste test” from a traditional artist using traditional media.

  3. Raina Avezeda

    This is trash. If you think this is better than a professional illustrator then your taste is trash too.

  4. John Mansfield

    “Make it in a graphic novel style like Zach Snyder.”

    So, a significant body of Snyder’s work is part of the input of this process. Artists already get into issues about whether or not they are merely copying someone else’s work or are using the foundation of what went before as a base for their own original work.

  5. Stephen C

    That is true. However, it is worth noting that the real AI art masters know the principles that go into “Zach Snyder” style and know how it’s different from some other graphic novel artist and can take advantage of that knowledge. In other words, they know the building blocks that lead to certain styles, and then they can replicate it and mix and match to create their own. The “like Zach Snyder” is a heuristic for people like me without art training that kind of know what Zach Snyder looks like, but can’t articulate why it looks like that, so I’m limited in how dynamic, creative, and flexible my outputs can be.

  6. RLD

    As a computer geek, I’m deeply impressed. The ability to keep track of an entity (“Moses”) and show it from different angles is a big deal. I know next to nothing about how AI models for images work, but the human equivalent would involve the kind of abstract thought that’s been lacking from AI to this point.

    As a consumer of art…meh. I’m no connoisseur of visual art and won’t make any claims about objective quality, but I don’t particularly like it. It does seem adequate for a project that doesn’t prioritize quality art or have the budget for a professional artist.

    As the brother of a professional artist (video games, not children’s books, but still…) I’m disturbed. A lot of artists make their living doing projects that don’t prioritize quality art. There are real benefits to making art more accessible. But this is definitely going to hurt people too. And some of those people put their art out there so people could easily enjoy it, only to have it used to train the models that now threaten their livelihood–with tech billionaires reaping the rewards.

    I really wish we had a functional political system so we could have a serious conversation about the implications of AI.

  7. Stephen C.

    …..And to give you an idea of how uneducated I am in this space, Zack Snyder is a movie director. I meant Frank Miller the artist, but Zack Snyder has directed Frank Miller movies so I think the AI picked up on what I meant.

    I also share the empathy unemployment. I think in the end from a utilitarian perspective, the net gain will be much higher than the net drawbacks, but the losses are all focused on a small number of people in certain sectors of the art industry, so I’m sure notions of rising tides lifting every everybody’s boat two inches is paltry comfort for the illustrator who has a mortgage and a sick kid.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.