I’ve been developing ChosenPath.ai for storytellers who have great ideas but find traditional IF tools overwhelming. With Chosen Path, you don’t need to write every path and manage complex logic. Instead, you only control the creative vision - the Story World - and let AI handle everything else.
What you define
Your World’s rules and atmosphere
Compelling characters with secrets and motivations
Manages multiplayer if you want collaborative stories
Who this works for
Dreamers who come up with great premises but have no skill and desire to code.
Architects who design compelling worlds but don’t have the time to write stories in them.
Satirists who see the absurd in the world but have no design skills to create a game around it.
The Worldbuilding Academy has beginner-friendly video lectures that walk you through the entire process. There’s a small community on Discord who can help you as well.
Everything is free during beta. No catch, no ads, no data harvesting.
What stories would you tell if you didn’t have to write every single branch? Looking forward to seeing what this community creates when the technical barriers are lowered.
I think you might have chosen your community poorly, lol. You logged into the “suffer as a writer for personal enjoyment and no profit incentive” website.
I for one cannot wait for the end of your beta period, when presumably I will have the opportunity to pay to read words that someone could not be bothered to write themselves.
Lol, I get you guys! Fortunately, the traditional way of writing IF is still available. If I want a really strong IF experience, I myself would play the finalists of recent competitions and not enter a random world in Chosen Path.
I believe that there is also a space for experiments with new premises, though. For niche ideas, and for short experiences that can be created quickly. For IF that incorporates multiplayer. And for an AI tool that helps traditional IF writers test different approaches for their stories.
Chosen Path can do these things. It’s not supposed to replace hand-crafted interactive fiction.
My personal belief about the value of “ideas” and “premises” is that the quality of an artwork ultimately depends on execution, and an “idea” on its own can only be good or bad insofar as it’s an idea that lends itself well towards being executed. A good idea is one that “basically writes itself,” while even the worst ideas could hypothetically be turned into a great work of art by someone with a lot of skill.
All of which is to say: if all you have are “great premises” but no skill or desire to execute on them, you have essentially nothing of value to contribute. I think I literally do not believe that there exist “dreamers who come up with great premises” but have no skill and desire to execute on those ideas in some appropriate medium. The skill and desire to execute is an actual necessary component of developing quality premises.
I believe that there is also a space for experiments with new premises, though. For niche ideas, and for short experiences that can be created quickly. For IF that incorporates multiplayer.
Yeah, I’m afraid you’re unlikely to find a receptive audience here. You’re welcome to advertise IF-creation tools here, including LLM-powered ones, but this is specifically a community built around writing in an old-school and generally unprofitable style—and as a result, it tends to be generally uninterested in AI.
But of course, there are exceptions, and your post isn’t breaking any rules. Don’t take this as an official moderator warning or anything! Just a suggestion to temper your expectations, since this may not be the audience you’re looking for.
A couple months ago, I wrote an article that did some numbers that seems relevant here.
AI-generated stories read like “AI slop” because LLMs can’t tell good stories, for the exact same reason that LLMs can’t tell good jokes: LLMs are trying to minimize surprise.
If my argument is right, the whole idea of “get the LLM to generate 30,000 words for you” won’t succeed in making words worth reading, because the LLM will be generating 30,000 words by identifying the words most likely to come next, which minimizes the likelihood of surprising the reader.
LLMs are getting better and better at minimizing surprise, which is making them worse and worse at writing stories. The words it generates are boring, by design.
LLMs work better for well-defined problems, where surprises are unwanted. But in fiction, surprises are essential.
So, yeah, I think this is the wrong approach entirely. Turning a premise into 30,000 unsurprising words doesn’t even help you test/validate whether the premise was good.
If someone was looking to do commercial work why would they ever use it. Frankly there are other tools that do the same thing so his only stand out feature is the a.i . No one wants to pay a.i written work so it’s the huge flaw in this model .
Then the examples are bad . I was trying to be nice but the examples he used embodied all the negative stereotypes about ami into one . I was kinda shocked he used them as examples to sell ppl on this tool which is why I said for him to delete them.
Lol that’s exactly why I wouldnt use this tool . Also you don’t own the copyright for something that is 100% with a.i with no human intervention at all .
Alot of ppl will totally skip over an indie game with a.i art so if I did use a.i art in itself it only be a prototype for my own eyes . By the time it got to the public that would be gone .
Now to be fair for I do think a.i is gonna improve drastically in the next five years . So I can’t say this attitude will stay half the “content “ I see on Facebook , YouTube and Instagram is a.i and it seems the general public is very bad at telling even when it’s obvious. So it’s gonna be scary once it’s perfected to remove all the flaws and even I can’t tell .
Ive already seen a few a.i videos and pics that have almost fooled me .
This is mostly correct but a.i text is bad because it has a hard time maintaining context .
Because a.i can generate text that definitely reads better than the average self published novel, most fanfiction , all Japanese light novels, Chinese web novels ,romantasy novels and ya novels . The issue is the text does not maintain context and starts falling apart rapidly . It has hard time maintaining context from scene to scene . Once this problem is gotten rid of I cant say the writing will always be slop. The a.i music already is at at a level where I say it’s no better or worse than a lot of the generic pop songs I’ve heard on the radio.
In fact someone did a.i 50s motown version of 50 cent music and several other rappers and from what I’ve seen the general public have been loving these songs . The motown version sounded significantly better than the original and the singer which was suppose to be 50 cent voice with a motown feel actually sounded like he he had some some soul. I had thought a cover band really made the track meet not I found it was all a.i .
So I’m saying while a.i does produce slop it’s not like someone who very skilled wouldn’t be able to fool ppl. A.i created literature won a prestigious literary prize and a.i art won a prestigious art prize and neither of the critics in either case where able to tell it was a I it was only reveleaed after winning . I see ppl do a.i slop experiments all the time with art and funny enough ppl often pick the real art usually made by some amateur as slop. I feel like we are in place where this is the worst a.i will ever be so I say don’t count on it remaining slop or being slop forever. I already see the general public getting fooled by the most basic a.i content on YouTube ,Facebook and Instagram .
Going through just a few points raised in this thread:
“LLMs can’t surprise.” I don’t think this is true. They do still struggle to surprise in a way that works nicely in a novella-length story arc (at least on my platform), but it’s getting better with improved systems and newer models. Plus, many IF experiences do not need real surprises in the first place. Many satiric worlds, simulations, and slice-of-life stories work well without that element.
“Premises without execution are worthless.” On the platform, you can define elaborate settings, game mechanics, and instructions for writing and image generation. You can just say “Create a cyberpunk world of cybernetically enhanced piggies”. That, I agree, will tend to be shallow. You can iterate from there, though, and add a lot of clever ideas and dynamics that will guide the stories. I believe that that is a form of execution.
“Stereotypical examples.” Fair point. I myself am not a storyteller or worldbuilder. I created the examples so people get a sense of the flexibility of the system. They should be replaced by more creative worlds. Happy to feature better, user-generated worlds (if they want that).
“Creators don’t hold copyright.” It’s true that creators don’t own the text that the engine creates. Nobody does; it’s in the commons. They do hold the copyright to the world definitions and instructions, though.
As a social entrepreneur, I’m mostly interested in how AI-powered IF can be used in education, coaching, etc. I also think that AI will play some role in creating IF for entertainment, though. So the task is to figure out what works and what doesn’t, and what is useful and what isn’t. In order to do that, we need to experiment. Chosen Path is such an experiment. It’s a side-project that can be used for free. I invite people to play with it.
As an idea person myself, I come up with a ton, mostly bad, but due to sheer volume, manage to land on a few decent ones. That said, ideas really aren’t worth much at all. Execution is everything.
. I was trying to be nice but the examples he used embodied all the negative stereotypes about ami into one . I was kinda shocked he used them as examples to sell ppl on this tool which is why I said for him to delete them.