Skip to content

Episode 9 – Balancing Creativity and Technology with Cameron Sutter from Plottr

Listen

Elsewhere Online

Apple Podcasts | Spotify | Amazon Music | YouTube | iHeartRadio | RSS Feed

In this episode, hosts Steph Pajonas and Danica Favorite engage in a thought-provoking discussion with Cameron Sutter, the innovative mind behind Plottr and the blog Real Human Writers. Together, they explore the transformative role of AI in the publishing industry, debunk myths, and highlight practical applications for writers.

Introduction to Cameron Sutter

Cameron Sutter is a multifaceted individual who wears many hats. He is the creator of Plottr, a visual storytelling tool that has revolutionized the way writers organize their stories. Additionally, Cameron runs the blog Real Human Writers, where he offers balanced perspectives on the use of AI in writing. With a background in software engineering and a deep interest in machine learning, Cameron brings a unique blend of technical expertise and creative insight to the table.

The Genesis and Evolution of Plottr

Plottr began as a personal project for Cameron, designed to help him organize his own writing. Over the years, it has evolved into a robust tool that serves a wide range of writers. Plottr offers a visual way to organize storytelling, allowing users to see the threads of their story come together on a colorful, user-friendly timeline. It also doubles as a Series Bible, helping writers keep track of intricate details across multiple books.

One of the standout features of Plottr is its array of templates, such as Romancing the Beat and the Hero’s Journey and many others. These templates guide writers through different story structures, making it easier to break through creative blocks and achieve key milestones in their narratives.

Integrating AI into Plottr

Cameron and his team have been exploring how to integrate AI into Plottr in a meaningful way. Their focus is on making AI a valuable tool rather than a mere toy. One of the primary uses they are considering is brainstorming. The idea is to leverage AI to generate character ideas, scene suggestions, and plot points based on the context of the user’s story. This could be particularly useful for writers who hit a creative block and need a nudge to move forward.

Balancing Perspectives on AI

Cameron has a balanced perspective on AI. He acknowledges the divisiveness surrounding AI in the writing community but emphasizes that it’s possible to use AI ethically and responsibly. Cameron explains that the current large language models (LLMs) are not true AI but rather sophisticated tools that can assist writers without replacing their creativity.

Addressing Ethical Concerns

Cameron has been vocal about the ethical implications of AI in writing. He proposes the idea of training an LLM solely on novels that authors have given explicit permission to use. This approach could create a more ethically sound dataset and potentially yield better results for writers. Both Steph and Danica agree that this is a promising direction, although it requires careful consideration of the source data and training methods.

Cameron’s Workflow and Favorite Tools

While Cameron’s primary focus has been on developing Plottr, he and his team use AI extensively for various tasks. One of his favorite tools is OpenAI’s GPT-3 playground, which allows users to tweak hyperparameters and understand the underlying mechanics of AI-generated text. Another exciting tool he’s exploring is Searchie, which offers AI-powered search capabilities for educational content, making it easier for users to find specific information within videos.

Cameron Sutter’s insights offer a balanced and nuanced view of how AI can be integrated into the writing process. His work with Plottr and Real Human Writers serves as a valuable resource for authors looking to navigate the evolving landscape of AI in publishing. Whether you’re a seasoned writer or just starting out, there’s much to learn from Cameron’s approach to combining technology and creativity.

For more information on Plottr, visit Plottr.com. To stay updated on Cameron’s thoughts on AI and writing, check out his blog at Real Human Writers.

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: Okay. Hi, everybody. Welcome back to the Brave New Bookshelf. Once again, I’m your co host. I’m Steph Pajonas, COO and CTO of Future Fiction Academy, where we teach authors how to use AI in every part of our process. We’re having a great time talking to all of our people who’ve been coming on for interviews and Danica especially has been so great about calling and emailing and doing all the stuff that she needs to do to bring these people on.

So Danica, hi, how are you?

Danica Favorite: Hey, I’m Danica Favorite. I’m the community manager at PublishDrive where we help people distribute their e books, audio books, and print on demand to the largest worldwide audience from one simple, little dashboard. And [00:01:00] also, I can announce this because by the time this goes live, it will be true, but we’ve also got a super cool new AI metadata tool. So be sure to check that out. Our links are all in our information.

Today I’m excited because we’ve got Cameron Sutter here from Plottr, and he also has a really great blog called Real Human Authors. Is that right? Real Human Authors?

Cameron Sutter: Real human writers.

Danica Favorite: Real Human Writers. Thank you.

Cameron Sutter: I doubted myself though. I might have to check that later.

Danica Favorite: I wrote it down and then I literally changed the page. Yes, Real Human Writers. So, we’re here with Cameron Sutter from Real Human Writers and Plottr. Two super cool spaces, and so excited to have him. Yeah, Steph said I do all the inviting and finding people, but let me tell you something, she’s the tech wizard behind.

As we have said multiple times, I am not techie. Do not ask me to do tech things because I don’t understand them, but I’m learning with AI. And so I love that Steph and I have this great [00:02:00] partnership. I can do one thing that I’m really good at, and she can do the other thing that she’s really good at. And so, as always, I think that’s why we’re such great partners. And that’s something I hope our listeners take from that because you don’t have to do everything. You just partner up with someone that can be a good partner for you.

Love the fact that we get to have Cameron and his perspective on AI. So, Cameron, why don’t you introduce yourself? Tell us a little bit about what you do, what you write about. I’d love to hear about both Plottr and Real Human Writers. I know that’s a lot, but feel free to share because I do want our listeners to hear about the different tools that can help them and make them more successful as authors.

Cameron Sutter: All right. Well, thanks so much for having me. I have heard so much about both of you before. Meeting you today. I, I feel like we’ve probably talked at some point, but I know everybody I talked to is like, Oh yeah, you gotta, gotta talk to these two. They know what’s going on with AI. And I just happened to meet the connection.

So thanks so much for having me. Really excited to be here.

So yeah, I’ve been [00:03:00] working on Plottr for actually, it’s been 10 years now that I started Plottr just in my basement for my own writing. And In 2020 is when I left my job and started doing this full time. And so I’ve been serving writers for a while.

I do the code myself, which it’s not to brag. It’s to give a little bit of a background to me because I am a software engineer and I’ve been following machine learning and AI for years and there’s so much to know about this that I cannot say that I’m an expert. It is very complicated, but I’m trying to learn as much as I can about like the technical aspects of it.

So about a year ago, maybe 14, 15 months, probably the hype really started for this. It was, I think it was when GPT 3. 5 came out that because it was so quick between three and 3. 5 and it was such a big jump that I think the world just exploded.

Like in six months. This is going to take over the world, you know? And so that’s when it really got on our radar for Plottr. Okay, we need to start doing something quickly about this or figuring out if this is useful. [00:04:00] And so we’ve spent the last more than a year trying to figure out how this is going to be useful for writers and listening to a lot of writers, talking to a lot of writers. It’s been a very divisive thing for writers. And so it’s been awesome to hear both sides and to feel the emotions of both sides and then to use it for ourselves, trying to figure out how it’s useful.

And so I’m leading to why I started writing about it. The Real Human Writers is because I felt like a lot of things these days, it was too divisive. And there’s other paths that you can take with AI, or at least other ways to think about it. And having some of the background of what’s going on in underneath, I wanted to be a voice for that other way that you don’t have to be hateful either way, you know, you can love all people that use or don’t use AI and you can still use AI and enjoy the benefits of it, but hopefully understanding it will help you.

Danica Favorite: You brought up some really good perspectives and actually really the reason why we wanted to have you and like, when we were coming up with our lists of dream guests, you made that list [00:05:00] because, yeah, because when you talk about AI, like, definitely you’re friendly towards AI, but you do put out both perspectives and you make it really clear that if you don’t use AI, that’s okay.

If you do use AI, that’s also okay. And you really try to blend the two. And I think that’s what we really appreciate about you because there are some times I’m like, he wrote this and I don’t agree with that. And it’s okay. And then there are times I’m like reading what you say. I’m like, yes, that’s exactly how I feel.

And I love that because it just shows that we need a variety of perspectives. It shouldn’t just be one voice out there. We want all the voices to feel like it’s safe to express how they feel and understand that there isn’t a right or wrong way. There’s an ethical way and we talk about being ethical a lot, but as long as you’re being ethical, we really want to embrace everyone at wherever stage they’re at.

So thank you for putting out that perspective.

Cameron Sutter: It’s kind of put me in the firing line [00:06:00] for both sides. I, I’m not fully in the hate AI camp. And so the people that do don’t like me and I’m not fully in the use AI for everything camp. And so I get some hate from that side too, but that’s just who I am as a person. So it’s like where I feel naturally, I think just always in the middle.

Danica Favorite: yeah, yeah. I’m that way as a person as well. I don’t know if you’ve done the CliftonStrengths or done any of the stuff with Becca Syme, but my personality is high restorative and so I suspect you might have a little bit of that because typically as high restorative people, we just kind of sit in the middle and we hear everyone’s perspectives and we want everyone to get along.

Cameron Sutter: sounds like me. Yup.

Danica Favorite: Yep. Yep. So, why don’t you tell us a little bit about Plottr because I think that Plottr is a really unique tool out there that could help a lot of writers. Last time we had someone on to talk about creating Story Bibles and things like that with their tools, and that really helps writers. But I know Plottr also has some really great tools for helping writers organize their writing [00:07:00] and writing workflow. So I would love for you to talk about that and tell us about what Plottr can do for writers.

Cameron Sutter: The main idea of Plottr is that it’s a visual way to organize your storytelling. And so you can see the threads of your story kind of weave together on this colorful, hopefully friendly and easy to use. We call it the timeline, but it’s kind of like a spreadsheet that just a little more bubbly and easy to use and more colorful, but it’s being able to visually see your story come together and see the threads of it just individually and, and reason about those really changes the game, at least for me. And I, for a lot of people that we talked to it really changes how you write your story and how you think about story.

And then it also became your Series Bible, and I didn’t start this or that wasn’t the original idea for this, but just getting feedback from people and figuring out how I use it and stuff.

I was so disorganized and so having everything for the whole series in one place has been game changing for a lot of people and for myself too.

So [00:08:00] it’s kind of those two things. It’s visually seeing your story and it’s a Series Bible to keep everything together.

And then, Oh, one thing I got to mention is the templates. We’ve got tons of templates, like Romancing the Beat the Hero’s Journey. Almost 40 now, actually, that are just like different story structures or different ways that you can think about story and some resonate differently with different people. But using those just helps you to learn how to write, but also guides you if you already know what you’re doing. Just kind of guides you and gives you nudges about what you need to write next or whatever. .

Steph Pajonas: I will say that I’ve used Plottr and I’m not much of a plotting person. in general because I’ve always pantsed my stories, but I tried it for one of my novels because I was like, this is really cool idea to be able to see your story in front of you and little bubbles and you can move them around.

I really did like the templates. I thought that they were fantastic because you know, you, sometimes you just need a little direction. You may understand your story or what, you know, Kind of what you want to do. But then the templates [00:09:00] help you break through any of those blocks you may be having, especially about like certain milestones in your story you need to be reaching.

I really liked that. So I’ve used it. And then the next book that I tried to plot, I couldn’t do it. hate when that happens. So I’m really excited to see that you’re continuing to evolve the application because that there are a lot of people who need that organization especially people who love to see things visually, you know, like they’re really into mind maps. They love like having boards with post it notes on them. It’s a way of looking at a story. And I think that that was what made it really special.

Cameron Sutter: And we found it’s also great for pantsers. I don’t know if it ended up helping you, but not really. So, for pantsers. Being able to go back after you’ve written the first draft and move things around and edit it, because that still happens in your story. So we found that’s great. And then the Series Bible aspect of it, when you’re on book three, book seven, book 10, it’s going to be so hard to remember those details of that one character.

What was his eye [00:10:00] color? Or that one animal, we’ve had so many people tell us that Oh, I changed eye colors between books and my readers wrote to me and like, Oh, why’d you change his eye color? I love when he had green eyes, you know, and it’s just hard to remember those details.

Steph Pajonas: It is. It’s hard to remember all the details. I’m working on a fourth book in a series right now, and I’m constantly having to go open up Scrivener, search for the character, find out what, I couldn’t remember the last name was. And I didn’t write it down in my notes. There was just a first name and not a last name. So that sort of thing, being able to catalog all that is really helpful.

Danica Favorite: I was just going to say, yeah, I have a number of manuscripts where the first draft is, instead of the character’s name, it says what’s his name, because I have to go back and look it up. So, right now, Plottr doesn’t have any AI in it . I was curious if you are using any AI in Plottr and if there were plans to do so. And if you’re not comfortable sharing that, that’s totally okay. I’m just super curious.

Cameron Sutter: Yeah, I could probably talk your ear off about it. So, maybe not didn’t [00:11:00] want to ask that . So we spent like the last year trying to experiment with it and see where it would be useful in Plottr. Because it’s great at some things, not great at others. And so our big thing we’ve realized lately is we want it to be a tool, not a toy, and so we’re trying to figure out where is it going to be a valuable tool for people that they’ll use consistently and not just a toy that’s like, Oh, that was fun. And then never use it again.

And so one of the things we feel like AI would be great for in terms of how Plottr already works is brainstorming.

So giving you ideas for characters, like give me five character ideas that already uses the context of your story and knows that so that you don’t have to write a prompt every single time. And so it’ll know your genre because you’ve already put this in.

It’s a fairy tale. Here’s the premise of my story. Here’s the genre. What’s the other one? I’m thinking of here’s the theme of my story, things like that. It will already know those. It’ll know the backstory of the characters and things like that because you’ve already put those in and so it’ll have a lot of the context already for you.

And so when it gives [00:12:00] you five ideas, hopefully the idea is at least that there’ll be within the genre or at least be useful ideas for you. And so it’ll give you five ideas and you can mix and match or choose which ones. And also things like scene ideas to give me five scene ideas for the inciting incident or something. I’m kind of stuck here. What should be the darkest night of the soul? Those kinds of things.

So, that’s one way that we’re planning to put it in there. Actually, it’s already in there. We’re already built it into Plottr, but we’re in the stage where we’re just about ready to get beta testers to try it out and see if it’s actually useful.

But I wanted to ask you what do you think of that? As far as like a use for AI within Plottr and I’ve got some other uses, but what do you, what do you think about that? Would that be a useful thing? Would it be more of a toy?

Steph Pajonas: Definitely useful. I mean, this is where I fall down as a pantser trying to plot is that sometimes I don’t realize the next piece of the story until I’m in the story, right? And then there are times when I’m in the [00:13:00] story and I’m like, ah, I know something has to happen in this next chapter.

Something, something pivotal has to happen. And this is where AI comes in really handy because I can feed it what I’ve already gotten, tell it the, just like you said, the genre, the theme, the main characters, the setting, all of that information is already there. And then it can just give me five ideas for the next chapter of the story.

I can tell it, Oh, I want some conflict in here. I kind of want this or that to happen. And it can generate from there. And then sometimes even if all of those ideas are a complete waste, It

can sometimes get my brain moving. And I’m like, Oh, I like that. But if I put a little twist on it like this, it’ll work for me.

So I could definitely see it being helpful in Plottr because I think that’s where a lot of people get stuck, right? Sometimes they get stuck in the story, they don’t know what’s going to happen next, and that can actually be very detrimental [00:14:00] to their forward momentum. They’re writing, they’re working, and then all of a sudden they hit this brick wall, and if they can’t get past it, Sometimes they’ll give up and we don’t want them to give up. We want to hear their stories.

So I like the brainstorming., I feel like brainstorming is probably the thing that I actually use AI for the most. Like I’m churning ideas, I’m working on it and it’s keeping me in the chair. It’s keeping me writing. So I say thumbs up to that one. I would definitely go for that.

What about you, Danica?

Danica Favorite: I definitely also give that a thumbs up. What’s interesting is I think most of the guests that we’ve had and talked to are using AI for brainstorming. And first of all, I do feel called out by Cameron’s comment that he wants it to be a tool, not a toy. I definitely have way too many writing toys.

But yeah, I think it’s kind of the same thing. I’ve probably got. Several different stories. Jill Cooper, if you’re listening, please close your ears, but that I haven’t finished. And part of it [00:15:00] is because I am a plotter and I have the entire plot laid out for some of these books, but there’s always that one linchpin scene. That I don’t know what’s going to happen and I can’t move forward, even though I know the other stuff. I’m so stuck and AI has been really helpful with some of that, but having an AI that understands the context of my story to be able to plot with them is really great because. So many times I’ll try to feed it into like Chat GPT or whatever and the stuff it gives me is just garbage It’s not true to what I want to do.

And I love Steph and I’ve talked about this before where 2am you wake up, you want to brainstorm something. I cannot call a friend and say, Hey, What do you think about this? My friends are not going to be happy with me. So I like having that brainstorming ability. I think that’s also the important thing. to note about brainstorming with AI is that yes, they give you ideas. I’ve never taken the entire [00:16:00] idea my AI brainstorming gives me. I always tweak it and make it my own because I’m picky. There’s stuff I don’t like. And that’s the same thing as when you’re brainstorming with a friend, right?

Like they’ll say, Hey, why don’t you have him take over the city? And you’re like no, I don’t like that. Ooh, but what if his friend takes over the city? It’s that same concept. So love the idea of having it be able to understand the context of my story.

Cameron Sutter: and with friends brainstorming sometimes you feel somewhat obligated to run with their idea or at least consider it. Whereas with an AI, you’re just like, Nope, that’s garbage, throw it away. You know.

Danica Favorite: Yes, I’m glad you brought that up because I haven’t been able to put that thought to words, but it is so true. I feel like sometimes you’re brainstorming with friends and they think this idea that they have for your story is the most brilliant thing ever and you think it’s the most stupid idea ever, but you can’t tell them that. So, with the AI you can. I was just saying in the AI for Writers group yesterday that I have told the AI go home you’re drunk.

Steph Pajonas: Like, no, [00:17:00] Claude. No, I really don’t want to do that. I have typed in before. I don’t like that idea. Nope. Sorry. Move along.

Cameron Sutter: One thing that I found that’s really interesting is when we’re trying to get it to brainstorm name ideas or names for characters, not necessarily just like a list of names, but give me five characters. The names are very repetitive. Have you found that? It’s very hard to find anything, but like there’s always a Luna. There’s Johns and Mary’s very common, you know.

Steph Pajonas: For science fiction stories, I feel like it’s always Elara. Elara. I’m like, why does this name keep cropping up all the time? And then Danica and I have said, Said a few podcasts so far that the favorite town for AI is Willow Creek. It is always setting stories in Willow Creek. I got to find this Willow Creek because all the good stories are being

Danica Favorite: I had a story idea for Willow Creek, and now I can’t use it because of AI. But yeah, that’s why this bookcase right here, this shelf right there there’s probably about five or six different baby name books that I still religiously use [00:18:00] because AI is still not great at it. I don’t know why we’re not training data on baby name books. We need to do that. Maybe it’s a copyright issue or whatever. We won’t go into that, but yeah.

Cameron Sutter: It’s actually something that one thing that we’re trying to figure out on the Plottr team of how do we make this better? And that’s one of the ideas we had of like just training on baby names or something like that so that it will have a bigger list of names to pull from.

Danica Favorite: Yeah. If you could get some ethically sourced baby name data, I’m down.

Steph Pajonas: Yeah, yeah, you should figure that out. I think that’s a winner right there.

Danica Favorite: So yeah, let’s get into the questions that we ask everyone. I think we’ve got some good feedback. Hopefully you’ve got some good ideas for Plottr. If not, you know where Steph and I are now. We’re friends now because you’ve been on the podcast, which means just get in touch with us. We’ll give you all of our opinions.

Cameron Sutter: I’ll be coming to you. Believe it.

Danica Favorite: Yes. So first question, we’ve kind of talked about this, but I’d love to hear how you’re approaching AI and publishing. I know your blog talks a lot about that on Substack. So tell us [00:19:00] what your approach to AI is. That way we can get that perspective and hear how your thoughts have evolved with that.

Cameron Sutter: Yeah. So I think it’s important to make the distinction that what we have right now, these large language models, LLMs are not true AI. Because when it comes to, I see some good reactions there because when it comes to true AI, honestly, I’m a little more I’ll just say I’m a little pessimistic. It’s scary.

But when it’s these LLMs for writing, I’m very optimistic. I think they create incredible tools for us to use. So I’ve been taking the tact of trying to teach that these are not true AI. That these are just language models, they sound like a human, sort of, sometimes, but they’re not a human, they’re not really AI, there’s no reasoning going on behind the scenes.

The example I like to give is the little riddle of the farmer and the fox, the grain and the, and the chicken where you have to get it across the river and the boat’s only big enough for two of [00:20:00] those. And so you have to figure out the puzzle and get them back and forth without one of them eating the others.

And so if you ask AI that, but you vary it in different ways, that helps you to see that there’s no actual reasoning going on, because a lot of times it’ll just give you the scripted answer that everybody has learned but if you change it so that the boat is big enough for all of them to go across. One time out of 10, it’ll give you, Oh, just put them all on the boat. But the other times it won’t, it’ll give you the answer is if there was only spots for two on the boat. So there’s no reasoning going on.

That kind of sent me down this path of like, Okay, if I can explain why there’s no reasoning going on, I think people won’t be scared as much. And so I try to teach about what’s going on behind the scenes. And I use metaphors like, it’s like Google Maps, and it’s just a GPS system for words. And so if you think of a map and the coordinates on it, you have latitude and longitude, which are just two data points to figure out where a spot is on the map. And with that, you can find a path from one place to the other [00:21:00] using Google Maps.

So AI is in a very dumbed down, simplistic way. It’s kind of making a 3d map of words where the points, the latitude and longitude are probabilities. So cat and dog are more probable to be close to each other, whereas cat and wall are less probable that they’ll be close together.

And so the distances between words are probabilities of them being used. That makes not a 2d map, but there’s lots of different dimensions of these words. And so it’s like a multidimensional map and that’s kind of hard to visualize. So I just say it’s a 3D map of words.

So if you imagine like the world or the globe, you have latitude and longitude and then altitude, and that’ll show you in a 3D space where you are in the world.

And the LLM is taking the last word of your prompt and trying to figure out what the path through the word should be. When I explain that, it seems to demystify it for people and then they’re like, oh, that’s not so scary.

Maybe I don’t have to worry about it being better writer than I [00:22:00] am and taking over my job and things like that. When I talk about it like that it brings people over to the happy side. But I can’t say that for sure. But it does seem to disarm people’s fears a little bit.

Danica Favorite: Yeah. I think it’s a great analogy. It’s one that it is nice because it’s just a different way of explaining something that Steph and I, we were just in PMs today going, Oh, why is this so hard for people to understand? And so I love the more analogies and ways we have for people to understand how LLMs work is really important because you’re exactly right. It isn’t true AI. It’s just an LLM. But I think the AI buzzword is what we’re all used to.

So I’m glad that that is where you’re coming from, especially because you do have that computer programming background. And so, you know, you have that education behind you to say, yeah, this isn’t the thing we should be afraid of.

Cameron Sutter: Yeah, definitely. And so that means that you’re free to use it as a writer. Like if you were afraid of it before, it’s not true AI, it’s okay. It’s okay to use it. It’s going to be a good tool in your [00:23:00] toolbox, but it also feels like, and I wrote about this recently and I, This was after I wrote this, I think that’s when you approached me, Danica, but I was saying it feels like the sentiment has kind of changed in the writing world, at least a little bit, or at least with some people that before is like very hateful, very divisive, and maybe those people are being quieter lately or something, but it feels like there’s more of a, yeah, use AI for whatever it makes sense for in your process, but it doesn’t have to take over your creativity. If you love a certain part of it, do that part. You know, you can keep doing that, but there are parts that AI is going to do great for you. And why not use that? It’s going to make the parts you don’t like easier. And the parts that you do like, you can do those more often.

Danica Favorite: Yeah. I will say I’m very pleasantly surprised by how I’m seeing that tide shift. Although today, for whatever reason, The hate in my Facebook feed was really strong. No one commented on my post, but it was just people’s posts. And I’m like, guys, come on. So I think it’s kind of it ebbs and [00:24:00] flows.

And I know again, this is going to be a little bit later than the announcement that was made today about News Corp and Chat GPT coming up with a deal to train on News Corp data and people were flipping out on it. And Steph and I are both sitting back like guys, you still don’t get it.

Cameron Sutter: There’s, there’s definitely still some hatred. Yeah.

Steph Pajonas: Yeah, and this, week particularly was strong because of the new deal between OpenAI and News Corp. And then the hubbub over Scarlett Johansson saying that, she felt that her voice had been used for the assistant. So, I think that there were two fairly negative news stories this week.

And so there’s a little bit more negativity. And then when that tends to settle down a little bit, people, people feel a little bit more positive. So I just think it just happens to be this particular week. That’s a little bit, little negative.

Danica Favorite: Yeah, it’s, it’s whatever is in the news cycle going on. And as we were talking in a conversation earlier [00:25:00] this week someone had said about, you know, well, it’s just about the sensational headlines and absolutely The headlines that are, Oh, no, AI is going to take over the world. Those are going to get the most clicks.

And so of course, that’s what is out there for people to click on. We do have to remember that sometimes the news comes out and there’s this instinctual, I better freak out. And then once you process the information and really take a look at what does the story actually mean, you can calm down and say, yeah, it’s all good. We’ll all be fine. It’s not going to ruin our lives.

Cameron Sutter: And actually, that kind of brings up the whole thing about the News Corp and OpenAI. That brings up an idea that that I’m kind of playing with. And I’ve been asking people around to see what they think, if this is valuable, if it’s worth building. But I was thinking that if we were to take one of the open source LLMs and train it solely and only on novels that people have given permission to train on. [00:26:00] And so they had to come to us and say, yes, we agree to you using our book and here it is. And then we train this LLM on those types of novels and we’d probably make one. For each genre or something like that, I feel like that would be a much more ethical and a lot more pure of a data set. And it might make a big difference in what it generates. What is your guys thought on something like that?

Steph Pajonas: Well, if you’re going to train a brand new LLM, you definitely need a whole lot more data than, a few people’s novels, right? You’re gonna need billions worth of words in order for that to train. But like you said, there are these open source models out there. There’s Lama, from Meta. And then there’s Mistral from the French company. And there are just, there are a bunch of them.

And you can make a fine tuning data set that you can fine tune those models and then. Possibly get something that works a little bit better for novel writers. There are a few of them that have been showing up [00:27:00] on Open Router. ai, which is the place that I go for all the latest and greatest new open source models that are out there. They’ve got a few that have been trained on role play data. So if you’re into the role playing universe, they’ve got a bunch of them that are trained like that. They have one that was, it was called like Cinematika, I think. That one was trained on a bunch of screenplays, I believe. It was interesting to prompt that and see the kind of responses back . So there are definitely people out there that are doing this. I’m totally open for giving my books. I actually been like, somebody take my books, feel free. I’ve got like 40 of them and train a data set. That’d be totally cool with that. So it’s just one of those things where I think it’s going to get also easier to train these open source data sets over the next year or two.

And. I think that it’s open and ripe for experimentation. So go for it if you can. Definitely.

Cameron Sutter: Cool. And is that though, like [00:28:00] those open source datasets, I don’t know this. And so I’m trying to figure out from you two experts, they’re already trained on a ton of data, right? And so it wouldn’t really truly be starting from zero and they know nothing. And so I’m sure people would still have a problem and say that they’re not ethical, right?

Steph Pajonas: Right. So there are some people out there that are just, that are very much against any of the LLMs that are trained on the data that’s out there right now, because some of them feel that the source material was improperly used. So, Metal Llama. I believe it’s been trained on like a lot of Facebook data.

So when we become Facebook users, and we’re using it for free, right? That means that our data is what Facebook values and what they use to, to train their models. So for me, that feels ethical. Like I knew when I was signing up and I was hitting that terms of service that they were going to be using my data to do lots of things, including, selling ads, doing market research, all that kind of stuff.

So for me, that was like, I had a feeling [00:29:00] that was coming. Right. So I think that you’re going to probably want to do some background research on the different open source models that are available, see what they use to train their base model find the ones that you feel are the most ethically sourced and then run from there.

And I think that Over time, too, that some of these will be built on synthetic data, which is basically outputs from LLMs. And then that could be very interesting, too, to take a synthetic data set that was used to train an LLM and then put a fine tuning on top of that, that is all novels and whatnot, and see what comes out of that.

That could be really interesting as well.

I just find this whole field to be extremely interesting and I’m constantly reading up on it and finding out what they’re doing and what they’re working towards. So I think there’s a lot of room for experimentation here. And I’m just gonna have to sit down and look at what they’re talking about and what they’re using for their training and decide which you feel is best for you and your customers.

Cameron Sutter: Yeah. Interesting.

Danica Favorite: It is very fascinating. And I [00:30:00] would say if anyone’s listening and they are curious about some of the fine tuning stuff the Future Fiction Academy does have some really great classes on fine tuning. Steph kind of started that whole fine tune movement, because as you see, she’s super geeked out over it.

And it really is a process, but she’s gotten some really good results.

Steph Pajonas: Yeah. Me and Elizabeth have done fine tuning OpenAI GPT 3. 5. We’ve had some fun with it. She was definitely the first out of the gate and I was running after her, trying to keep up,

But we’ve together, we’ve trained plenty of them now and we’ve got a good idea of how it can be used. It’s pretty fun.

Danica Favorite: Super fun. let’s get to our next question about your workflow. What does your workflow with a I look like?

Cameron Sutter: So in my own writing, I actually, this is one thing that I’m very disappointed about, and I’m trying to rectify this. I made this great tool for writers to be able to improve their lives. And then I didn’t write as much because of it, because I was running the business and making the tool. But in the last year or so, I’ve really tried to get back into [00:31:00] writing. And so right now I’m currently writing a nonfiction book about AI for writers. And so I don’t use it that much in that. I will use it for the cover and blurbs and things like that. But the fiction stories that I’m getting back to as soon as I’m done writing this book, at least I’m hoping we’ll see, we’ll see what my process is like.

I’m sure I’ll use it a lot for brainstorming and things like that. But yeah, as far as the writing, I’m unfortunately not at a point where I use it that much. But my team uses it a lot for all the things that we do. And so it’s more like secondhand AI usage. I’m like, Hey, we need this, can somebody make this for me?

And then my team is like, Hey, we got this awesome thing. And it’s just what we needed. So, I’m kind of the wrong one to ask about my AI workflow at the

Danica Favorite: No, you’re not. That’s actually the perfect answer. And that’s what I told you before we started is there aren’t any wrong answers because I want people to see that there are a variety of perspectives. There are a variety of use cases like for you personally, you’re not there yet because you’re doing other things. You’re thinking about it, but you’re not sure what that’s going to look like. And that’s [00:32:00] valid because I think that there’s going to be some listeners who say, Oh, yeah, that’s me too. And that is the whole point for us is we want people to have that experience of realizing that there isn’t, again, there’s no one true way. There is whatever way works best for you, and the fact that you have thought about it and you’re working through what you need to work through in your way, that’s exactly the right answer.

Cameron Sutter: I passed.

Danica Favorite: Okay. So, last question, and again, we’ll see what your answer is on this one, is what is your favorite AI tool? Do you have a favorite tool?

Cameron Sutter: Yeah. So it’s actually the, Chat GPT, the playground, the tool that they have because you can see what it’s thinking behind the scenes and like how it gives you access to the hyper parameters. But for those that don’t know that I call them the dials of creativity. Like you can dial in how the LLM responds. And so you get to play with that and see what the different responses for, [00:33:00] if you dial it up or down and, and just to, I really like trying to figure out what it’s doing and why it’s doing it, especially because I’m trying to build a tool for writers. And so I’m trying to figure out how it works so I can build a good tool for writers. So that’s my favorite use of it right now. It’s figuring out how it works. And so the playground .

And then there’s this other one that I discovered recently that we’re going to use on the Plottr team. So we’re doing a lot of educational content on Plottr. And there’s this tool called searchy that will let you make this like Netflix like experience of your video content for people to come in and consume your content. And they have this AI powered search that is the coolest thing because so you ask it, okay, What do I do in act two? And it will scan through all the videos, or it already has scanned through all the videos that you’ve put into this thing. And it will give you a summary of what those videos said about what you’ve taught about act two. And then not only that, but it will give you a reference to where it found that in the videos. and it will link you to the minute of that video. So if it’s like a 45 minute video, you probably don’t have [00:34:00] time for that in the moment, but you can watch that minute where they teach that one thing that you wanted to know.

And I feel like that’s going to change the game for educating yourself in general, about writing when we were going to release this next month. And so it’ll have this whole, we’re calling it the vault, this whole vault of educational content that you’ll be able to search and just find the minute of video that you need, which is just amazing to me.

I love that.

Danica Favorite: That is amazing. I need to know the name of that tool because I haven’t heard of it before.

Steph Pajonas: Same.

Cameron Sutter: Yes. I will email you afterwards. It’s called searchy, but I’ll email you about it.

Danica Favorite: Awesome. Thank you. Yeah, we’ll put that in the notes as well.

Cameron Sutter: So those are my favorite tools with AI.

Danica Favorite: Like I said the whole point of this is not to have right answers. It’s to give a variety of answers . Although I do love that you’re playing in playground because I think there’s a lot of comfort in it. And I know this is something that FFA talks about a lot.

And I’ve learned a lot from them about this is that knowing what’s happening behind the scenes and why is so important [00:35:00] because. It really does show you, okay, this is what’s going on. And is it the prompt or the AI that messed up? When you have those tools, a lot of times it becomes really obvious. Okay, this was a bad prompt or the AI messed up, or this was just a bad dice roll. As Elizabeth Ann West likes to say, it was just a dice roll. I’ll just do it again and having that access to that back end and seeing what’s happening gives that perspective of, okay, cool. Because it’s all a learning process and if you view this whole journey into AI and working with the LLMs, if you view it as a learning process, not just for you, but for the machine. And that whole growth that comes of it from everyone, I think it just is so much more rewarding. And frankly, in my opinion, it’s a lot more fun.

Cameron Sutter: Yeah. And Elizabeth and I talk a lot. She’s taught me so much about how AI works and the way she describes is just, it’s so good it’s [00:36:00] amazing.

Steph Pajonas: All right. So, Let us know about how people can find Plottr and get updates about what you’re going to be doing with Plottr in the future.

Cameron Sutter: Our website is Plottr. com and we have accounts on all the socials. It’s either like on Facebook, I think it’s Plottr app on YouTube it’s Plottr and so forth. And so if you go to Plottr. com slash blog, I think you can subscribe to our newsletter to find out about things, but yeah, our plans right now for AI we’re almost in the beta stage and both of you, if you want to be in our beta group, we’ve got a small group of people that we want to play with it. So if you’d be okay with that, we’d love to add you to that group. I’m getting nods. Okay, great. And and so once we, Get feedback from people that it’s like, is this useful? And so once we get that, or like what would be useful. So when we get that we’re hoping later this year, we’ll have something. It’s not our main focus right now.

It’s something that we think is valuable, but we’re also not trying to get on the hype train and get something out there so we can make a ton of money as soon as we can. We’re trying to make something that’s useful.

I’m also writing on [00:37:00] it’s real human writers dot sub stack. Dot com, right? So you can find out about what I’m writing, but I’m not really talking too much about Plottr there but also on our blog Plottr dot com slash blog. I have a few articles about AI and you’ll have our timelines and things like that.

Cameron Sutter: And so if you follow us, you’ll find out what we’re doing with AI and I’m trying to be more vocal about it and talk about our reasoning and also our timeline and things like that. So

Steph Pajonas: Perfect.

Thank you so much. All right. Well, we’re going to wrap up this interview then. We’ve learned a ton about Plottr and what Cameron has been using AI for. And I think that a lot of our listeners are going to find some synergy here with him and what he’s been doing. Because a lot of you are listening because you’re curious about AI and you may not have even added it to your process yet.

So don’t think that you’re alone. There are definitely lots of people out there who are learning and adding it to their process as they go along. All right. I’m excited to move along with learning more about [00:38:00] AI, just like Cameron and Danica are, and we’re going to be doing lots more of that in coming episodes of the Brave New Bookshelf.

So everybody, thank you so much for tuning in and we will see you all in the next episode.

Bye.

Danica Favorite: Bye. Thank

Cameron Sutter: Thanks for having me.

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.

Leave a Reply

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

Brave New Bookshelf