One of the most important aspects of our approach to building AI features is ensuring we use your data ethically. This article might get boring or technical, but some people really want to know these details.
“Your data” feels so sterile, so generic. Instead of calling it your data, let’s call it your stories, your creations, your ideas, your brainchild … children … brainchildren? Maybe let’s call it your “story data.”
We’re determined to be good stewards of your story data because we are also caring for our own at the same time.
How AI in Plottr Handles Your Story Data
Now, when you use the forthcoming AI features in Plottr, we make it very clear that you’re about to use AI, so there’s no way it’s going to happen without you knowing it.
And when you do use one of the AI features, Plottr makes a request to one of our servers and that request is carrying some of your story’s data.
I can detail what that entails (characters, scenes, etc.), but for now, it’s only important to know that some of your creation is flying through the Internet to our servers (whether you have Plottr Pro —which is all in the cloud — or not)
Our servers get a prompt ready with your story data and send that to the LLM model (the AI).
Will Plottr Use My Story Data to Train AI?
When you talk to an LLM this way — from computer to computer instead of chatting with it like ChatGPT — your data isn’t stored in the LLM and it isn’t used to train the LLM at all.
That may change and we have no control over AI platforms’ policies, but for now, here’s what OpenAI says about using their API to use ChatGPT (an API just means two computers talking to each other). And that’s the way we are talking to ChatGPT, through the API.
So once we get a response back from ChatGPT with the results of your prompt, we store some statistics about it (how many words in the prompt, how many words in the answer, etc.) and then we send it directly back to the device that you were working on.
We Don’t Store Your Story Data Beyond Plottr Pro
We don’t store the prompt, nor the response, nor any of your story data. If you’re on Plottr Pro, your story data is safely stored in our cloud anyway, but if not, it’s only temporarily on our servers and then discarded.
We don’t analyze your story nor the LLM’s response. We don’t use your story data to train an LLM of our own. We don’t have an LLM of our own. We are just using the available ones out there like OpenAI, Claude, etc.
What About Risks to My Story Data?
There are always risks with digital data, and it’s a compromise that we’ve all accepted by using computers and the Internet. That’s nothing new for Plottr AI and your data.
It’s like when humans invented fire, or the concept of double-edged swords. Fire has improved the world for millennia, but it also burns people every day. And we all know what happens with double-edged swords 😏
Luckily we’ve built the AI features into Plottr in such a way that you don’t have to use them at all if you don’t want to. They’re not forced upon you, and it’s very clear when you’re using them.
So to summarize:
- Plottr doesn’t have its own LLM
- We’re relying on existing LLMs
- We’re using the API so LLM models don’t train themselves with your story data (but we have no control over that long-term)
- Your story data is briefly transmitted on our servers but then gets discarded once you get your AI-generated output (unless you’re on Plottr Pro)
- We’re not analyzing or reading your story data
On that third point: We have no say over OpenAI’s policy around how it stores data, but if it changes in the future we’ll let Plottr users know so you can keep using AI features fully informed (or opt not to use them, as you prefer).
Like I said at the outset of this post, this wasn’t going to be the most fun or interesting, but it’s important for everyone to know how we’re treating your data and how it will be used for upcoming AI features, if you choose to use them.
We’re trying hard to be good stewards of your story data because we are also caring for our own story data and it’s precious to us, just like it is to you.
Cameron Sutter,
Plottr Founder and CEO
About Cameron
Cameron Sutter is a sci-fi/fantasy author and the inventor of Plottr – the popular visual story planning software. He’s escaped death by explosion, rock slide, disease, and car accident. He loves doing funny accents for his kids, but believes his life’s mission is to serve writers. He lives near Oklahoma City with his wife, six kids, and too many pets.
2 thoughts on “Plottr AI and Your Data”
All existing LLMs were trained from stolen data in the first place. There is no ethical use of any current model, plain and simple. If you were actually worried about this, you would keep your product far away from them.
As it stands, I’m sorry that i zany recommend plottr to anyone, now, and will have to retract my recommendations from the past. I’m glad i only ever got the stand alone version, and i won’t be updating it so long as there’s any avenue open to an LLM.
i don’t know why it’s so difficult to understand that OpenAI and all thier ilk consistently lie and steal data from every source they can, but that is literally what they have done from thier inception. There are *zero* real user protections for anything that touches thier model. You’re deluding yourself if you think any differently.
God this is disappointing.
Thanks so much for expressing your concerns about AI tools and LLM models in general. You’re not wrong and I totally hear you. We’re not taking a hard stance for or against LLMs honestly. It’s a murky, complex subject.
First let me just say that these features will be completely optional and turned off by default. You will have to purposefully engage the LLM features and you’ll know when it’s about to be used. There won’t be any surprises there, and if you don’t click the button each time you want to use the LLM feature, our system never calls out to an LLM model.
Secondly, your data is never used to train any model because we’re using the API.
Lastly, you’re probably not wrong about what OpenAI has done. What if we use Claude? And what about Google and Facebook that have been selling your data for years? I have the same hangups with them. I don’t trust them, I don’t like them, but to operate certain things in my life, I kind of have to use them.
That’s not to be argumentative. I just know it’s a not a cut-and-dry topic and we’re trying to navigate this world-changing technology the best way we can.
I know we’ll make mistakes. That’s how we learn. There’s no other way unfortunately. Hopefully you all will give us some grace as we find our way and stumble along the path.