Listen
Elsewhere Online
Apple Podcasts | Spotify | Amazon Music | YouTube | iHeartRadio | RSS Feed
Rachelle Ayala: Bridging the Past and Future of AI in Writing
Rachelle’s website: https://www.rachelleayala.net/books
Rachelle on Amazon: https://www.amazon.com/stores/Rachelle-Ayala/author/B007DXL5WY
Rachelle Ayala’s journey is nothing short of inspirational, with a background that spans from working on optical character recognition technologies in the early ’90s to embracing AI for crafting compelling narratives today. Her transition from a tech-savvy professional to a romance novelist and then later using AI tools like ChatGPT illustrates the versatile applications of AI across different eras and genres.
AI-Powered Writing: A New Frontier
Rachelle shares her personal experience with AI, from using ChatGPT to brainstorm and flesh out novel ideas to leveraging advanced technologies like Google’s Gemini for more intricate tasks such as enhancing narrative depth and emotional resonance in her stories. The conversation underscores the importance of adaptability, with AI tools continually evolving and offering new possibilities for enhancing narrative structures and storytelling techniques.
Ethical Considerations and Creative Integrity
A significant portion of the episode is dedicated to exploring the ethical dimensions of using AI in writing. The speakers advocate for a balanced approach, emphasizing that while AI can serve as a powerful tool for generating ideas and refining narratives, the essence of storytelling — its heart and soul — remains a distinctly human endeavor. They encourage authors to engage with AI ethically, ensuring that the use of technology aligns with the principles of originality and authenticity.
Community and Collaboration in the AI Era
Rachelle and the hosts further highlight the critical role of community and collaboration among authors navigating the AI landscape. Platforms like the Future Fiction Academy and the AI Writing for Authors group on Facebook are celebrated as vital resources for sharing knowledge, exchanging ideas, and fostering a supportive environment for writers of all backgrounds and expertise levels to explore the potential of AI in their work.
The Way Forward: Embracing AI with Openness and Creativity
By sharing Rachelle Ayala’s experiences and insights, the episode invites writers and readers alike to approach AI with openness, curiosity, and a willingness to experiment, while always prioritizing the creative integrity that defines the art of storytelling. As AI continues to shape the future of publishing, embracing these technologies with an informed and ethical mindset will be crucial for all those who venture into the brave new world of AI-assisted writing.
Full Transcript
[00:00:00] Welcome to Brave New Bookshelf, a podcast that explores the fascinating intersection of AI and authorship. Join hosts, Steph Pajonas and Danica Favorite, as they dive into thought provoking discussions, debunk myths, and highlight the transformative role of AI in the publishing industry.
Steph Pajonas: All right. Hi, everybody. Welcome back to Brave New Bookshelf. I’m Steph Pajonas. I’m the COO and CTO of Future Fiction Academy. And I’m joined by my lovely co host, as always, Danica Favorite. Danica, hi.
Danica Favorite: How are you? I am Danica favorite. I am the Community Manager at Publish Drive, and we help make your publishing process more simple and more efficient. And one of the ways we do that is with AI. And so since Steph and I are both with companies that have passion for AI as well as in our own personal lives, we love AI.
So, we’re here to talk about that with you. Last week, if [00:01:00] you didn’t catch our last episode with Melle Amade, she was amazing. I love her tech background and she has such a great business insight into AI and the use of AI. So if you didn’t catch that, go back and listen to that.
Steph Pajonas: She even name dropped NASA. That was, whoa.
Danica Favorite: Hey, right, NASA. Like, I love that. I mean, one of my favorite parts of the interview was when she was talking about how we’re using AI to get people to the moon and Mars and all of that stuff. And whoa, mind blown. As we have been saying all along, we know that AI is kind of a controversial subject in the author and publishing worlds.
But we also recognize that it is in every part of our lives. So let’s figure out how to harness it for good and use it ethically and responsibly for authors in ways that are going to benefit our careers. So [00:02:00] today we have an amazing author with us who uses AI for her writing and has some really cool stuff.
Steph, you can tell us a little bit about Rachelle and then we’ll have Rachelle introduce herself.
Steph Pajonas: Oh, Rachelle’s fantastic. I’ve met her in the past year or so since I’ve been in the AI space. Really love her and her books. She writes about writing with AI, and she’s also put AI into some of her fiction, too. So that’s really cool as well. I wanted to introduce guys to Rachelle Ayala. She’s here to talk with us today. How you doing, Rachelle?
Rachelle Ayala: Doing great. Thanks for having me.
Steph Pajonas: Please introduce yourself.
Rachelle Ayala: I’m Rachelle Ayala and it is a pen name and I actually do have a technical background. I was there during the days in the mainframes and have a Ph. D. from Cornell University in Applied Math where we did parallel processing.
We now know that you need parallel processing. For A. I. But yeah, I started writing [00:03:00] fiction about 12 years ago. Well, actually, 14 years ago. But before that, I actually worked on AI with optical character recognition back in the early 90s. We were only going from zero to nine.
That was the big deal. You were trying to. Yeah, it was. I worked at NYNEX science and technology. And they wanted to automate reading checks. So when people wrote their checks, they wanted to be able to read the numbers off the checks. And I do remember all the backpropagation and the feature recognition and all this matrix multiplication and going forwards and backwards, adjusting the weights.
It’s so amazing to me that years later, we have LLMs, which are billions of parameters more than our little check writing thing. But I do remember we had a lady she would, she used to sit there and her whole job was an image would flash on, and she was the trainer. The image would flash on [00:04:00] the screen and she would decide if it was a number from 0 through 9. And she’d say, oh, this looks like it could be a 7 or a 1, and then they were looking at the features, and 5s and 6s could be confused, and maybe 9s and 7s, so it was really interesting that all this is bringing back memories of those days.
Steph Pajonas: Wow, that’s super interesting. Especially when you think about where optical character recognition comes from and we started using that in the like 90s and into the 2000s for books. So, it has evolved over time from just numbers to letters.
And then now we’ve got LLMs on top of it.
Danica Favorite: Yeah, it’s to me, it’s amazing, right? Just the advance in the technology and how fast it’s grown. And of course, now we see it just growing exponentially to where before it was so specialized and now this is available to anyone. And so that to me is just absolutely mind blowing.
Rachelle Ayala: I gave it all up to [00:05:00] write romances about 14 years ago. I took early retirement. So, I haven’t been into space. Right? And back about 14 years ago, because we didn’t have the chips that were able to do big, massive amounts of computation. I was a little bit kind of, you know, it was still going, but it wasn’t the explosion that came later with the NVIDIA GPUs.
Danica Favorite: Yeah. So, you know, we have math and computer genius to romance writer. So tell us how you got into that space of combining using AI for your writing, like when and how did that start happening?
Rachelle Ayala: Well, I, just wrote regularly. You know, not by hand. I mean, I typed 10 fingers. I wrote 10 fingers. But I think when Chat GPT first came out in November of 2022. I said, Hmm, a chat bot. I got curious and interested in it, but I had a deadline in December of 2022. [00:06:00] I wrote a novel in 14 days and I was all proud of myself. And why I wrote a little book about it. It’s called, I Wrote a Novel in 14 days. That is the last novel I wrote without AI and I’m a pantser.
So that novel just, You know, I would just go, go, and then every time I wrote a scene, I’d think, Oh, what could happen next? What could happen next? And, since I already internalized the romance story structure, because you do have to have some kind of sense of a story direction.
You don’t want to just be meandering. That’s how I’m able to pants because I do know the romance and I don’t know any other genres, but romance. So I had written a book on love stories, how to write a romance novella in 30 days. And so I do know that, you have the obligatory scenes, you have to have the meet cute, you have to have the conflict, you have to have the lock in, you have to have that moment when you realize you’re in love. So I know those things in the back of my head, but I never put it down in the outline.
And that was the last book I finished. It was just [00:07:00] intuitive. I always knew Elizabeth before because in the whole indie publishing world, she was the famous Jane Austen lady. And so I had known about her and I might’ve started some reading her books, but it’s not really the genre that I read in, but I knew about her. And when I saw a video, I think it was an interview about Elizabeth about AI. I quickly googled her, got on her Gumroad class, because remember, she used to have her own Gumroad classes, and I signed up for those classes, and that’s the Rosebed.
You know, the first class was this. Tell me what a Rosebed is.
Danica Favorite: Yeah. And so for those who are listening, who may not know who Elizabeth is, Elizabeth is Elizabeth Ann West, who is one of the co founders of the Future Fiction Academy, and really wonderful in the space. We will definitely be having Elizabeth on our podcasts here at some point. Wanted to clarify that because we all know Elizabeth.
We love Elizabeth. We’re all great friends with her, but I just wanted to [00:08:00] clarify that for listeners.
Rachelle Ayala: So I’m somebody who came from the indie publishing world and knew about Elizabeth and at that time in January, you get tons of emails in your inboxes from all these people saying, Oh, AI can do this. AI can do that. And you’re just like, Oh, these are the usual suspects because weren’t they the ones selling this funnel thing, or maybe even NFTs, so, but then when I saw Elizabeth, I said, Oh, this is credible. I’m going to follow what she does because I knew she was an author that had actually published books. And so I’m really glad I found her. And then I found Steph. And I also, so I found out about Melle Amade, and so I, so, got into my group, and now I’m here.
Danica Favorite: Nice. We’re so glad that we have you. And so that leads me to our first question, which is, we’ve talked a little bit about it, but how are you approaching AI? What is your approach to AI and the usage of AI as an author?
Rachelle Ayala: Well, I’m actually very flexible for my [00:09:00] CliftonStrengths, my first number one is Adaptability, and the second one is Maximizer.
So, I move where AI moves, right? So when I first started, the only thing that was available was ChatGPT 3. 5 and Sudowrite where you could highlight something and ask it to fill in descriptions or to Enhance , change the mood or tone and so I was working on a book and I think I just immediately started, I think I was already in nine chapters in the book and I didn’t have AI right for me because at that point in time it really couldn’t, I would just use it to enhance some descriptions or maybe give me some ideas for mythological characters And so on.
And then when Elizabeth started talking about story information, so she used to go through this whole thing about, okay, we’re going to layer our prompts. And we’re going to ask AI to help us fill in our story [00:10:00] blueprint. Then I started doing that way. And just today, I was looking at my love by the prompt book because that was the book I put out in March 2023 using chat GPT. It was only chat GPT 3 at the time. And I already had story information in there and I had gone through first you do the premise, then you do the three act structure and the synopsis.
And the thing about AI, because it has this context window, is it’ll remember a little bit of what was before. And back then, the context window was smaller.
And it would give me maybe about five hundred, five to six hundred words, which would be a chapter skeleton from which you can fill in or expand on. So you never thought, oh, well, that’s it, that’s the chapter. And it gave interesting things. And so I was just looking at it today and I said, this stuff actually still works.
But as we go on, we learn more and then we can do more. Our mega prompts got bigger. We understood more about writing [00:11:00] scene briefs with scene beats. Remember we were all like, Oh, we got to make beefier beats. So I think my approach is I keep evolving with the AI. All books after 2023 and onwards have some AI input, but I couldn’t tell you if it’s this much or this much, or if it’s interleaved, or if it’s just thrown out. Each chapter, in fact, could have a different method. It’s a little bit weird now because Claude 3 is here. Now you have to try all three. You have to say, I got to get Sonnet and Haiku. So the work has actually multiply because it’s just not ever the same. Not for me. I’m not somebody who can. You know, okay, this is how I’m gonna write the book, and it’s a cookbook, and I’m gonna go back on that recipe over and over, it doesn’t happen that way, because AI is just changing so fast, and what it can do is amazing.
Danica Favorite: I just kind of laugh about it sometimes because before you’re right, when we all started in this space, it was just Chat GPT. And now it’s like, oh, but what about this model? What about this model? [00:12:00] There’s so many more choices. And it’s really interesting to see how that evolves. I think you’re right. Like it’s so much more work to go through and look at all the different models.
Rachelle Ayala: But, right before the show, I said, let me just review, because I had written this book that, this book, the initial outline was in that Love by the Prompt, it was called Island Heat, and it was about Maya and Luke. And I just realized that, that monkey, Mango, was created by AI.
It just popped up in one of the chapters. I wouldn’t have even thought about it, and now Mango is like one of the most popular characters in that series. Because he’s always so mischievous, and he just is a little mischievous matchmaker, you know. And So, AI will sometimes throw, and this is why I love it when it goes off the rail, because it might throw in something you never thought about.
You’re like, oh, this is great.
Steph Pajonas: Yeah, and he becomes a fan favorite,
right? Then like everybody just wants,
Rachelle Ayala: mango,
Steph Pajonas: the mango Mango, mango, mango.
[00:13:00] I love it. Yeah.
Danica Favorite: Yeah. I think that’s awesome. I think that just the idea of Mango, the monkey, AI was able to pop that in. Obviously sometimes the ideas AI gives, you’re just like no, but I love that you got this really cool thing that you weren’t expecting because I think that’s what makes it so fun are those little experiments.
Rachelle Ayala: Exactly, yeah.
Danica Favorite: That kind of leads me to my next question, because you talked a little bit about some of the evolution and how you’ve evolved with AI and gone with the flow. What does your workflow using AI look like? How does that play into your workflow and how do you use it?
Rachelle Ayala: Okay, so I’m not the person who can give you a recipe, and, I think it just depends on what I’m doing in the day or what I’m thinking about, because I still do a lot of daydreaming about my books and stuff but now, I would just ask Chat GPT or ask Claude what they think about something, so maybe I’m a little bit more dependent on them.
I used to have a friend [00:14:00] that when I lived back in the Bay Area and every Tuesday and Thursday, we’d walk around the lake four miles and I would bounce off every single thing on her. She was like my Chat GPT. Okay, so I would say, Hey, Pat, what do you think if I did this and that? And she said, you know, and then I moved in 2020.
And so in some sense, AI. is taking the place of Pat. I would say, well, what do you think about if I did this or that? Of course I know I’m talking to an LLM, but the thing about the LLM is it’s not human and it might give you something that you totally, I don’t think any human might have thought about.
And you’ve seen this even with Midjourney pictures, right? Images. Ooh, it put this and that together. And that’s what I like about it.
But ultimately just last night I was having a conversation with Chat GPT because I wanted to demonstrate how we layer prompts and go from a premise all the way to a chapter and it wasn’t giving me enough [00:15:00] conflict. It just really wanted the two enemies to lovers to just work together to figure out who their common enemy is. And I kept going back. I said, no, that’s not how an enemy to lovers work. She’s going to have to suspect him of stealing her horse the whole time while he’s saying that he’s going to help her find the horse.
And so I have to go back and I say, no, that’s not enough. And then even when it wrote the outline, in the middle of it, it already had the kidnapper confessing that he had not only stolen the horse, but he wanted to kill the lady. And but in the next chapter, the kidnapper actually tries to kill her. And so if he already confessed they would have nabbed him, right?
So AI still doesn’t have that common sense. That’s why I prefer not to follow their outline I prefer to just to get the cool ideas and the weird stuff that they might come up with and then just design the scene briefs on my own. Then I have several models write and I see if anyone gets the [00:16:00] snarky remark I want, then I pick it out.
So I always say, make the dialogue snarky. And some models are snarkier than others.
Danica Favorite: That’s good to know!
Steph Pajonas: I was going to say that. Claude is really funny sometimes, like really funny. I’ve had it write some dialogue for me of like disgruntled wizards and the things it comes up with will leave me crying with laughter. It’s really funny.
Rachelle Ayala: yeah. Oh,
Danica Favorite: Yeah, and I do want to go back to what you were saying a little bit ago about how, well, I’m not here to give you any kind of formula or anything like that. And I think that’s the point. And that’s why we’re all here is because we want to know how you do it, because there’s always going to be a little gem.
That you get from hearing somebody else or hearing parts of their process. Like, when we were talking in the last episode with Melle, what I loved is she and I are both dictators. And so I was always afraid when I started using AI, I was really frustrated because I wasn’t getting what I needed [00:17:00] from it.
And then I realized, well, that’s because I just want to talk out the story and I want to dictate it. And so then talking to Melle in our last episode, Melle was like, yeah, I just like to dictate. Okay, yeah, that’s exactly my process. And so never feel bad about not having a process that anyone can exactly duplicate or anything like that, because your process is really interesting to me.
And I think our listeners are going to get a lot out of it that this is how Rachelle does it. And if they can get some really good gems or ideas from how you’re doing it, I think that’s really what we’re hoping here. I’m glad you’re sharing and I’m glad you’re telling us a little bit about the different things you do. It’s really interesting.
When you go to write your book, how do you approach it? How do you say, okay, I’m gonna write this book, I’m gonna use AI. What are the pieces and the steps that you like to use? Like you were just saying, you like to bounce ideas off of it, [00:18:00] but when it comes to the outlining, you wanna do that.
Rachelle Ayala: AI does an outline and I feel free to ignore it. I, talked about it in that AI Author’s Journal. I think that was the one book where I tried to force myself to go through the “so called” AI writing process where you do the whole top down waterfall. If you’ve done software design in the old model was the waterfall, right?
You first you do your requirements, then design, then coding, and testing. But I was always one of those agile people. You just do a little of this, a little of that, and then you go back, change the design. So even in software engineering, I was an agile programmer and not a waterfall, but I tried to force myself to go through the waterfall.
I wrote it in the AI Author’s Journal and I had the outline, but then as soon as the story started, it just, did not go the way to outline. And honestly, I think mine was better [00:19:00] because AI would take the more conventional path. So what they wanted was to have this, they love to have the secret baby, you know?
So if you ever say you want a second chance romance, AI will always say, Oh, she had a son or a daughter and she didn’t tell him about. But I didn’t, that was not what I wanted. That was not how the story would be more interesting because we don’t need a child in between to bring them back together. So I would go through the waterfall steps, feel free to ignore the outline. And then from the very first chapter, it might deviate.
Steph, you’re a lot like me. I think you just start writing. And then when you get to a point, you say, Hmm, I wonder what AI thinks about this.
And sometimes it’s not even that you need to know, it’s just the same as having Pat walk with me around the lake. I wonder what Pat thinks about this. And I copy and paste. Alright, what do you think? What do you think happens next? And also I like to ask, what do you think happened in this chapter?
Because sometimes, if it’s interpreting it not the way I thought, [00:20:00] then maybe I’m not conveying it right. So for me, it’s a very organic process. And I always leave the parameters, the temperature to 1, and the top P to 1, because I actually want it to get more creative or to, think outside of the box.
So I’m a very not control type. So that’s how I approach AI. And now with Rexy, it’s like a power tool where now I can get four or five LLMs to tell me what they think.
Steph Pajonas: I’m very similar in the fact that I’ll be thinking, Oh, I’m going to outline this book. I’m going to be prepared. I’m going to be prepared. I’m going to go in there and I’m not going to write myself into a corner and literally like two chapters in. I’m, I’m way off. The outline, I’m all the way out in right field and even still, I’m definitely outlining more that I have these tools and still, I can’t stick to them.
So I, I understand completely.
Danica Favorite: And for those of [00:21:00] you listening who are like, oh, we have all of these pantsers. I am a plotter. Like a plotster. Like I start kind of panting, but by the end of chapter two, if I don’t have an outline to work off of, I am just. Not a happy camper. And I cannot continue to write until that outline is done.
So there is
Rachelle Ayala: See, we’re all different. So,
Danica Favorite: Exactly. Exactly. And so for those of you who are listening, who are like, okay, you’ve had all these pantsers talking. There are plotters. It’s okay. You are covered. And that’s why we keep encouraging people. There really is something for everyone and there’s a little gem that everyone can get.
And even as a plotter, I’m listening to Rachelle going, Oh yeah, I can totally see that. I love that idea of going to the LLM and saying, Hey, what do you think of this? And I think that is super cool.
Rachelle Ayala: I’m really excited about the really long context window ones that can actually keep the story straight now, a lot of them advertise. Well, Steph, you already [00:22:00] know, like they can say I have 200, 000 tokens, but it’s just picks and chooses because they’re using RAG to fix information that stands out.
And in fiction, everything’s supposed to stand out, but Google’s Gemini 1. 5 Pro , actually, I can upload the work in progress, and I can ask it for, to critique it, and I even asked it, do I have enough romantic tension, and show me the things where you think I have enough, or where I could add more, and it will actually tell me which things, like, that time when they went to the museum, and they did this and that, so this is almost like having a human critique partner, whereas I don’t think Claude can do that. I’m not saying anything bad about it, but I think we’ve all uploaded long documents, asked it about it, and it was kind of really summarized information.
Danica Favorite: Yeah. I think that that’s true. And I think that, and I’m sure Steph is going to agree with me on this, that each model has its own strengths and weaknesses. And so some of it really [00:23:00] is learning which tasks you want the different models to do and why and how and all of that. And, but like you, I have been super impressed with Gemini.
I’m hoping maybe this weekend to get even more time to dig into it because the more I play with it, the more it’s just blowing my mind with how good it is at the tasks I want it to do. So, I’d love that you pulled that out. Cause I’ve a couple episodes now been singing Gemini’s praises.
Steph Pajonas: They just released a notebook feature, which will probably be a few weeks old by the time this episode goes live, but the notebooks feature allows you to bring in your Google Documents into the notebook and then that becomes the RAG that you can search on. I tested it last night with my current work in progress.
I brought it in, I asked it questions about my work in progress, and the analysis was really great. It became sort of like a story blueprint so that you could be like, what are Garrick’s [00:24:00] eyes color? Tell me about what is happening in the first chapter of this book.
It’s really good at all of those things. Every single leap forward is bringing more for us to use. It’s great.
Danica Favorite: Yeah. I’m very excited about that. So I have not played with notebook yet, but now that I’ve heard your feedback as well, I think I’m definitely prioritizing that because I read about it and I was very excited and I haven’t had time to sit down with it yet, but I think that’s my next priority because the possibilities are amazing.
Even though we’re talking about Gemini and how excited we are for it, there are a lot of tools out there. So, we’re, we’re fangirling over it now, but I think as you are going to see
Rachelle Ayala: subject to change, right?
Danica Favorite: Yeah, exactly, exactly. And that was a good point Melle made in our last episode is that really pay attention to who is making those tools and who’s behind them because a lot of the things we’re excited about and even listeners who are finding this a year [00:25:00] from now Gemini might not even be here. I doubt it because obviously Google’s behind it, but you just don’t know.
And so always take that word of caution. But that brings us to our last question. And what is your favorite tool or at least the tool du jour? What are you really loving right now?
Rachelle Ayala: It has to be Rexy WordBanger because I’m a programmer, I understand everything about data stores, static variables dynamic things, and I just love how WordBanger allows you to use your content management system to you can store your character information, your synopsis, your motivations, your static settings and, and you call that story information. At the same time, you can take the information for your same briefs, like what you’re writing on right now and instead of copying, pasting it in these boxes have it in the data store. And then you can programmatically pull these pieces [00:26:00] of information out and assemble your prompt and then have it sent to the LLM and have the response come back and then stored back into your Notion. And so basically automates a lot.
Danica Favorite: I think that is a super cool feature for those listening who don’t know what Rexy WordBanger is. Steph, I know this is your wheelhouse, so could you give us just a short explanation of Rexy and what WordBanger is, how it works, all of that, so that people understand a little bit about it.
Steph Pajonas: Sure, just real quickly Rexy is a prompting tool that we developed at the Future Fiction Academy for our students. We wanted to give them the ability to prompt the AIs without having any other coding or prompting getting in the way. Some of the other tools out there they take your prompt. And then on the backend, they’re doing some prompt massaging basically, and like putting in different variables. And then they’re sending that to the AI, and then you get back a response.
So we wanted everything to be very transparent. We wanted [00:27:00] to teach the students how to, prompt properly and how to ask for things in a way that will get them good, rich answers back from the AI.
And so we developed Rexy, which would give you the option to store all of your stuff in Notion, which is what Rachelle was talking about, like, store all of your information and then send things to AIs and get back responses and then store all of those responses in Notion as well.
WordBanger though is our cool tool that we came up with that allows you to sequence prompts, allows you to prompt something, get a response and then prompt again, and you can continue to use the response in your prompt, or you can choose to set it aside. It allows you to keep going and going and going and going so that you can move on from one prompt to the next, to the next, to be building whatever you’re building. You could write a book that way, right?
Rachelle Ayala: And it’s very flexible. You can use it as a plotter where you could just put your [00:28:00] outline in and then just tell it go through each one of the outlines and write a scene brief and then it’ll lay it all out. You move the scene briefs into the Notion slots, and then you say now go through and just take each s cene brief and write a chapter.
So plotters can use it.
Or if you’re a pantser you could say, okay, here’s the previous scene, and here’s what I want to do. Because you do have to put in some of your input. You have to say I want this to happen and, Oh, it’d be cool for that. And then you say, write me a scene brief. And now that for a scene that continues from the scene that I just told you about, and then I told you what I think should happen. Or you can even ask it what you want it to happen.
So it’s not like you’re tied to a specific method because the flexibility is in the way that WordBanger allows you to put a prompt first, then the output from the previous prompt, and then the second prompt, so that you can sandwich the output from the previous prompt between [00:29:00] your instructions, which could be pulling in things from your content management.
You could have different writing style for each character. So you could say, oh, this scene from this point of view, and his tone of voice is blah, and it will pull that in. The problem with Chat GPT or Claude, is it’s going to remember the other character’s writing style, because its content window just slides. But you have no control over what’s in its content window. So let’s say you wrote a scene from the female point of view, and her voice is very different. And the next thing you want to write from the male point of view, but because her voice was still up in the context window, it bleeds in.
What Rexy does, each time it sends an insta, it’s a fresh prompt that goes in. So even if you were filling in with story information, it’s not that it remembered it from the last time you sent it, it’s that you’re packing it in for this prompt, And so you can very precisely control what the LLM is seeing, and that’s what’s really cool.
And that’s why it [00:30:00] beats the long context window chat because the long context window chat will bleed in. So you wanted to know some of the story, right? You want the previous thing, but you don’t want the voice. Let’s say you don’t want the voice of that previous character infecting the current thing you’re working on.
And you can do that with Rexy, but not cutting and pasting it into a chat bot.
Steph Pajonas: One of the nice things about Rexy is the fact that it gives the author, the creative, all the control, right? Just like she was saying, you don’t have control over the context window in Chat GPT. But when you’re sending a fresh prompt through Rexy to any one of these AI tools, you get the chance to say, I want this, this, and this story information to be a part of this scene. And it won’t be poisoned by anything else that came before that. So that gives you the control to do the kinds of things that you want to do.
Rachelle Ayala: It would be cool if from one point of view, you had Claude Sonnet write it. And from another point of view, you have Mistral or something write it. [00:31:00] That could be pretty cool. Like you do those dual first person point of views and use a different LLM for each cause they have different voices, even if you try to tweak it.
Steph Pajonas: They do. Mm hmm.
Danica Favorite: Yeah, that would be super cool. That would be amazing. Actually, I want to play with that now. And, and one of the things I just want to say again, this is for some of the listeners who are like, okay, so we know Steph has a really big tech background. Rachelle’s got a tech background. Melle has a tech background.
I do not have a tech background. And I think it’s really important to know that this is also something for non techie people. And I know some of the terms are probably confusing to people who may not necessarily be familiar with all of them or familiar with the tech stuff. Trust me, I get it. I was there.
I am still some of the terms. I’m like, okay, quick Google, what does that mean? But also the thing I love about Future Fiction Academy is that for learning how to do all of this stuff, they have some really wonderful videos. They also have some free [00:32:00] videos. So make sure that you are looking for that.
The other thing too, that is great is Steph has the AI for Authors group on Facebook and I’m one of her moderators and we’re there to answer questions anytime. Obviously, if you’re like, okay, what did they mean by this when they were talking about this in the Brave New Bookshelf, feel free to drop your questions in the comment section and we can answer it there.
Or like I said, we do have our Facebook group. We have a Facebook page for Brave New Bookshelf. There are a lot of really good resources. And again, if you’re not a techie person, that’s okay. Cause I’m not, and it’s really interesting and fun to learn. You don’t have to be a techie person to master it.
So don’t feel intimidated because it’s really a lot easier than it sounds sometimes.
Rachelle Ayala: Right. And as writers, you’re already pretty intuitive and you are very good at communicating. And so Chat GPT and Claude, they were designed to be conversational. [00:33:00] So don’t be shy with them, just talk to them and converse with them, and you will be able to influence how they’re going to give results to you, because they’re actually programmed to want to please you and to give you what you want. Well, to a limit, because they don’t the not safe for work stuff, but you don’t have to have a tech background.
And in fact, sometimes having the verbal abilities to tease out the stuff from the LLM is important too, because sometimes say, I’m sorry, I can’t do this. A tech person may just say, well, why can’t you do this and try to change their prompt. But. a more intuitive person might just say, well, what if we are writing fiction and I have a problem that I want to explore this area and it’ll just help. Okay, sure, fine. We’ll discuss this.
Steph Pajonas: It can be convinced sometimes to, to help us, even if it doesn’t want to in the very beginning.
Rachelle Ayala: I’m surprised that my book, Love by the Prompt, that’s over a year old, [00:34:00] is still selling because it does not require tech background, it’s just Chat GPT, and it’s just here’s the premise, here we’re putting in the story information, we’re asking it to write a scene . And once people dip their toes in there and they figure out, Hey, I can do this too. I’m not saying Rexy is easier , but the people in the Future Fiction Academy, everyone jumps in to help. Steph ,Elizabeth, and Stacey and everybody, of course they help, but other members also jump in to help too.
Steph Pajonas: Yeah. It’s a great community. We’re, we’re really blessed to have everybody in there helping at pretty much all times.
Danica Favorite: Yeah, I think it’s really great. I think that there is a great community of supportive authors who want to help with AI. So, like I said, they’ve got the Future Fiction Academy. We have our AI for Authors group and really feel free to reach out because the thing that I have found is that the people willing to help, if [00:35:00] you just ask, it’s actually, sometimes I feel like you get more help than you can even handle because people are really generous with their time.
So I love that about the community and I really hope that people listening are encouraged by that and can find some comfort in that community as well.
Steph Pajonas: Yeah, AI Writing for Authors on Facebook is definitely very helpful to people. They can come and ask questions. Rachelle has been in there for a long time as well. We need to remember that the community here can support each other, especially when you’re learning and you’re trying to figure things out, so. That’s why I’m really glad that we’re able to do this here too with the podcast.
Danica Favorite: Yeah. Me too. Like I think this is really going to be something very helpful to people. And that is always our goal is to be helpful and educational and see where we can guide people. So before we close out for today, Rachelle, do you have any final thoughts or anything else you wanted to [00:36:00] share about AI and publishing or something we haven’t asked you or touched on that you would like for people to know?
Rachelle Ayala: I think one of the things everybody’s going to see from this podcast is there’s no right way to use AI and that everybody can use it to their comfort level. And I know there are people who have different opinions about how AI should be used. But I think any use that’s ethical is fine.
Some people are not comfortable in certain areas and that’s okay too. And I think by having this podcast, we could show people all the possibilities of how AI can assist you. But at the same time, you do realize that you’re the ones putting in the prompt, right? And so that’s coming out of you. And so you’re using AI rightly, or you’re doing something sketchy by putting something in the prompt that you shouldn’t. It really is on the human to decide how they’re going to use AI.
And then the other part that I was more thinking about just lately is how much your [00:37:00] prompts affect the output. So. I used to think, Oh, like, even in my Love by the Prompt, I gave a whole list of prompts. I put them out in a on a website so people can copy and paste them. But now lately, I thought, Wait, you know, some people say, Oh, AI is not creative. I asked it for a high school reunion romance, and it came back with a secret baby.
Well, that’s because you’re asking the same prompt that everybody else is asking. So. It’s good to look at other people’s prompts as a guideline, but I think the more you put of yourself into the prompt, then you’re going to get something different and unique and that comes from you.
So, that’s just a thought I’ve lately been having.
Danica Favorite: Yeah, I completely agree. I’m like, yep, that is so true. You have to take that prompt and put your spin on it to get something that is more you.
So, Rachelle, how can people find you online if they want to follow you or find your Love by the Prompt book [00:38:00] or any of your other books?
What’s the best way for them to find you online?
Rachelle Ayala: I have a website that’s not that well maintained. It’s Rachelle, R A C H E L L E A Y A L A dot net. And if you go slash books, it’ll have a whole list of my almost a hundred books. You could just look up my Amazon authors page, I guess. That’s probably the best place to find me because I’m not very good at email marketing or funneling or any of those things. I just like to write and I like to help people.
So look me up on Amazon or if you have Facebook, then AI Writing for Authors group, I’m in there or FFA just send me on discord.
Steph Pajonas: Excellent. Thank you so much for coming today. It was so fun to talk about all these things because, we’re all creatives and it’s really, really cool to see what other creatives are doing with AI.
Rachelle Ayala: Well, thank you so much for having me, and I totally enjoy FFA and the Facebook group that you put together, Steph. Thank [00:39:00] you so much, and thank you, Danica, for all of what you do for publishing.
Danica Favorite: Yeah, you’re welcome. This is a passion project for both of us and something that we really want to be out there to support authors. And, and like you said, not every piece of AI is for everyone and that’s okay. We’re just here to share information so that people can make educated choices about what might work for them, how they might use it.
And if there’s something that’s not for them, if you’re listening to this and you’re like, yeah, no, I don’t want to do that. That’s okay. You don’t have to. But maybe the next episode you’re going to get a little gem where you’re like, oh my goodness, I really want to do that. And we want to support you as much as we can throughout that whole journey.
Rachelle Ayala: Totally agree. Thank you.
Steph Pajonas: Thank you so much for coming, Rachelle. We really appreciate you spending some time with us here today. So that’s it for today’s episode of Brave New Bookshelf. So on behalf of Danica and me and Rachelle, we’re going to wave bye. And we’ll [00:40:00] see you guys all next time. Thanks for coming.
Rachelle Ayala: Thank you.
Danica Favorite: Bye.
Thanks for joining us on the Brave New Bookshelf. Be sure to like and subscribe to us on YouTube and your favorite podcast app. You can also visit us at BraveNewBookshelf. com, sign up for our newsletter, and get all the show notes.