I’ve been absent for a little while. Here’s why.

This is my creative writing thesis, announcement of the discontinuation of the Graves of Idols Universe, and plans for the future plus a cool thing I made. This is also my timeline in my change as a creator. Finally, it aims to defend a new way to use a.i. tools as avenues for a new generation of bot creation and marketing.

Consider this the 'behind the scenes and in the mind of the name @Kittyland':

EDIT: FINALLY found the original person who came up with the idea that gave me the inspriation for this. The link is to the bot where they made their trailer and also... you get to see me trying to get them to collab desperately. Their name is @Chit.

I made something.

Over the last month, Graves of Idols grew far beyond what I originally intended it to be on JanitorAI. What started as a planned series of fourteen + three interconnected bots became an actual universe—with its own characters, histories, timelines, institutions, conflicts, mythology, and long-term story.

So I’m very happy to finally say this: I’m turning Graves of Idols into a book series. And because apparently I am incapable of having normal hobbies, I also made a trailer for it.

It isn't an advertisement. I'm not selling anything here. I made it because I wanted to see this world move, breathe, and sound the way it has existed in my head.

But making it also forced me to admit something: I am changing as a creator.

JanitorAI genuinely made me fall in love with writing again. Roleplaying led me to wanting to build bots. Building bots led me back into character writing. Character writing led me into lore. Lore led me into a very new skill of worldbuilding. Worldbuilding led me into learning how LLMs actually work, how information architecture affects them, how to build efficient lorebooks, how visual presentation affects an experience, and eventually how newer creative tools could help me express ideas that previously existed only in my head. If you were in any of the discords with me during this process, you saw the light bulbs go off in real time. But for the rest of you:

I guess, here is where I tell you how the hell we got here.

The Timeline:

For a while, I kept asking myself the same question: Why aren't people playing my bots?

I was making the usual romance and angst-heavy concepts, while at the same time seeing people in Discord communities, Reddit threads, and Janitor comment sections constantly asking for deeper characters, larger worlds, more lore, and experiences that went beyond the same handful of tropes.

So I tried giving them that. I made The Golden Path (Cyberpunk-ish RPG).

Then another series called The Station Chronicles (Horror).

Then another continuing an old bot I made in Mistwood (Fantasy RPG).

And every time I made something, I tried to learn one more thing. I got better at writing characters who felt like people instead of trope delivery systems. I learned to take a simple trope and layer it with layer upon layer of narrative beats. I started adding music because I wanted each bot to have an atmosphere, which eventually evolved into character themes, regional themes, and the musical identity behind Graves of Idols.

Mistwood was the pilot series for this new style of presentation.

I learned how to make lorebooks lean enough that I could build much larger settings without dumping thousands of unnecessary tokens into every interaction. I started thinking less about making a bot and more about creating the illusion that someone had stepped into a world that existed before they arrived.

And yet the numbers usually didn't reflect the amount of work I was putting into them. That frustrated the hell out of me. There are plenty of possible reasons for that:

• Botting exists and engagement numbers on this platform can absolutely be manipulated which hurts small creator's discoverability.

• Discovery systems favor certain kinds of engagement.

• The novelty of AI roleplay isn't what it was when the site first exploded.

Different communities value different things, and sometimes the loudest requests for “more lore” come from completely different people than the ones who think five paragraphs of lore are already too much.

I even asked people directly what I was doing wrong.

Some told me: “There's too much lore.”

Others told me: “You're hiding too much of the lore.”

Which was the point where I realized something extremely important: I cannot build creatively around trying to satisfy every possible audience at once.

So I stopped. Not creating. Trying to please everybody.

I started asking a much more useful question: What can I actually control?

• My writing.

• My ideas.

• My presentation.

• My willingness to learn.

• Putting the user's experience first and foremost and building the bot around that

• And the quality of the experience I am trying to create.

I used the Mistwood series as an experiment. Fantasy already gives an LLM a huge amount of shared cultural shorthand that is indexed, so I wanted to see how far I could push presentation and lorebook design without needing to explain every basic concept from scratch.

• Could I make something that felt substantially larger than the actual context being fed into the model?

• Could the user feel like they had stepped into a world rather than just opened another bot?

I think I succeeded. The numbers still weren't particularly impressive. Weirdly, that ended up being useful. It finally forced me to separate engagement from whether an experiment actually worked. Once I had seen firsthand how unreliable platform metrics could be, I stopped treating them as a meaningful verdict on the quality of what I was creating.

Regardless, I knew what I had learned from making it. I know this story is good. Janitor just is not the medium for it. So eventually my response became: it. I'm going to make something completely different.

I started thinking about stories like Watchmen and the feeling that a superhero world could possess history beyond whichever hero happened to be on the page. Not just costumes and powers.

• Institutions.

• Politics.

• Public mythology.

• Generational consequences.

• Different eras interpreting the idea of a “hero” differently.

• People becoming symbols and then discovering that symbols are much easier to understand than human beings.

• How to ask big questions using a universe as the mouth. Like old Star Trek used to do.

That line of thinking eventually became Graves of Idols.

And somewhere during that process, I also stopped being embarrassed about experimenting with new creative technology.

Yes: I use AI tools.

Of course I do. I've spent a ridiculous amount of time teaching myself how LLMs work, how to structure information for them, how to iterate visual concepts, how to communicate artistic direction, and how newer tools can bridge gaps between the skills I already have and the things I want to create. I used chatGPT to learn how to make lorebooks, what language is best to use with different LLM's. I had no one else to teach me.

I tried asking other people how to do some of this. Often, nobody knew, nobody answered, or I didn't have access to somebody with the skill I needed. So I started asking the tool that would.

And then I taught myself.

The trailer is probably the clearest example. I am not an animator. I am not a professional VFX artist. I am not a sound engineer. I do not have a production studio or thirty people working for me.

What I had was a universe I created, a very specific idea of what it should feel like, and access to tools that made translating some of that vision possible.

So I used them.

• I generated and iterated imagery.

• I experimented with animation.

• I used ChatGPT while solving technical and audio problems.

• I sourced music and sound.

• I rejected things that were wrong.

• I regenerated things that didn't resemble the characters or worlds I had written.

• I changed shots.

• I changed timing.

• I layered audio.

• I edited the trailer together.

And I kept going until the thing on the screen finally started resembling the thing in my head.

AI didn't wake up one morning and invent Graves of Idols for me. It gave me tools.

I still had to decide who Lexi was. I still had to decide why Brody behaves the way he does. I still had to build the two timelines. I still had to decide what survived the heroic age and what didn't. I still had to create the institutions, conflicts, cultures, histories, character arcs, metaphysics, themes, and eventually the ending. I still had to look at a visual output for the trailer and understand why it was wrong.

And I still had to have something worth making in the first place. That's the part of all of this that changed me. I'm not particularly interested anymore in arguing about whether somebody becomes a “real creator” based on which tools they use or the numbers of followers next to their profile or numbers of chats that can be manipulated so easily.

I'm much more interested in asking: What can I make now that I couldn't make before?

For me, the answer became this. And I'm pretty fucking proud of it.

I used to lie in bed when I was a child and invent stories because sometimes imagination was the easiest place to go when I wanted to be somewhere—anywhere—else. And somewhere along the way, I started carrying this stupid little dream that maybe one day I would write something another person could disappear into for a while too.

Maybe they'd read it. Maybe they'd smile. Maybe, for a few hours, the world I made would become somewhere they wanted to be.

I don't have a finished book yet. I haven't “made it.”

But I finally took the first real step toward that child's dream.

I created something I want to turn into a book. And for the last month, I couldn't wait to get home so I could work on it. I would think about it at work, in the car, lying in bed—characters, scenes, music, histories, little pieces clicking together.

For a month, I was that kid again.

Except this time I wasn't only imagining the world.

I wasn't using stupid coins to stage space battles in my head. God, I remember having this little bear coin holder that was always a Mon Calamari cruiser for some reason. 😂 I used to draw Star Destroyers on pieces of paper and move them around like they were actually fighting with my step brother because we couldn't afford the stupid x-wing board game.

God, the things I used to make back then.

Except now?

I was building it. Really building it. Watching it take shape in front of me. With my own hands.

And whatever eventually happens with it—whether ten people read it or ten thousand, whether it succeeds or disappears into obscurity—nobody can take that month away from me.

Nobody can take away the fact that I made something I love.

We took the first step.

And genuinely: Thank you, JanitorAI, for stoking that fire again.

The material posted on JanitorAI has always been only the surface-level version of Graves of Idols. The deeper character arcs, history, timeline structure, mythology, long-term conflicts, and actual story have never been published here. That was intentional.

I've also kept extensive development material documenting how this universe evolved—drafts, designs, writing, screenshots, timelines, source material, and the larger body of work surrounding the project—and I've taken steps to protect the original work.

And because apparently this also needs saying: Please do not steal my work.

Could somebody copy a character name? Sure.

Could somebody lift a concept, rearrange some details, change some terminology, and pretend they came up with something new? Sure.

But they still wouldn't have Graves of Idols. They wouldn't have the larger structure. They wouldn't have the unreleased lore. They wouldn't have the relationships between the different eras. They wouldn't have the planned character arcs. They wouldn't know why something introduced in one part of the universe matters somewhere completely different.

And they wouldn't have the story I'm actually going to write.

If somebody's creative strategy genuinely requires stealing another person's work and changing enough details to call it theirs, that says considerably more about them than it does about me.

And there's something else I've realized: I built the foundations of this universe in roughly a month.

That doesn't mean finishing it will be easy. Writing the actual books is going to require considerably more work than building a lore bible. But the difficult part was never whether I could invent another cool character name, costume, power, faction, or setting. The difficult part was discovering why all of these things exist together.

I know that now. And if something breaks while I'm writing it? I'll rewrite it. If an idea doesn't work? I'll make another one. If a character changes? I'll follow them somewhere better.

Creativity was never the scarce resource.

So this is also me acknowledging something I probably should have admitted earlier: Graves of Idols has outgrown JanitorAI as its primary home.

That doesn't mean I regret making it here. Quite the opposite. JanitorAI is part of why this exists at all.

I'm genuinely grateful to the few people who played my bots, shared public chats, talked to me about my characters, gave thoughtful criticism, told me when something wasn't working, or simply said that something I made gave them a good experience for an evening.

That mattered. It still does.

The bots aren't disappearing from the history of this project. They're little windows into the universe—and honestly, I think that's pretty cool. But Graves of Idols is becoming something else now. The actual story is going somewhere bigger.

So, for probably the last major thing I post here for a while about this universe: Enjoy the trailer.

And to the people who spent their time telling me the project was terrible, that using AI meant creating the trailer required no effort, that I just want attention, or that somebody could simply take the ideas anyway:

Well. I'm still here. I kept creating. And I'm going to keep creating.

You can't stop me.

See you at the finish line. :)

Primary Bot Focus Will Now Be Mistwood

💜

GRAVES OF IDOLS — TRAILER

(LINK)