On AI, Authorship, and the Future of Publishing

AI is one of the most talked-about—and misunderstood—tools in publishing right now.

In this episode of Talking Book Publishing, we sit down with Julie Trelstad to explore how authors can use AI as a creative partner, not a replacement. From writing workflows to marketing strategy and author rights, this conversation breaks down what’s actually useful—and what to avoid.

Julie shares how she uses AI in her own process, why strategy matters more than tools, and what’s coming next as AI reshapes the publishing landscape.

In this episode, we cover:

• How to use AI as a writing assistant without losing your voice
• Why most authors struggle with discovery—not writing
• The difference between ethical and unethical AI use
• How AI can streamline book marketing (if you know what you’re doing)
• Why authors need a strategy before tools
• A new system that could allow authors to protect and monetize their work in AI

If you’re writing, publishing, or marketing a book right now, this episode will help you understand where AI actually fits—and how to use it to your advantage.

Connect With Us

Kathleen: kathleenkaiser.com

Adanna: maddannastudio.com

Podcast listeners receive a special 20% off all ProBookLaunch products. Use code TBPLISTENER20 at checkout.

Have a topic or guest suggestion?
Email us at podcast@talkingbookpublishing.today

Join the conversation on Instagram: @writerspubsnet

Transcript

Copy

Kathleen Kaiser (00:09)
Hello everyone and welcome to Talking Book Publishing. I'm Kathleen Kaiser along with my co-host, Adanna Moriarty. And today we have a guest that I think has a very interesting view of what's going on with AI and authors. And not just in your creative part, but in the other kind of tools that you can use it because it is a tool So Julie welcome.

We're very happy to have you.

Julie Trelstad (00:36)
Thank

you.

Adanna Moriarty (00:36)
Julie, how do you say your last name? Okay, Julie Trelstad

Julie Trelstad (00:39)
/ˈtrɛl.stæd/

Kathleen Kaiser (00:41)
Great. you know, this is such an interesting topic and there's such debate inside the writing community on when to use it and when not. And lot of agents, they don't even want you using Grammarly anymore. They ask if you've used this or you've used that when you're, submitting books. So how do you feel about it from a perspective?

Julie Trelstad (01:02)
I think, you know, like I'm a writer too. And in creative life, what I have found is that it can be an amazing organizer of your work and can act as like, it can keep track of all your characters. It can help you achieve continuity. It can help you integrate comments into a draft. You can have conversations with it saying, this working or not? I work with Claude AI when I do my writing and I've given it explicit instructions that it is not allowed to write any prose.

that it must only act as an assistant. And recently I found you can either use the Claude Documents feature or the ChatGPT Canvas feature to keep a running log of what it does with you. like if I ever decide to sell my work and I probably will self-publish, but if I decide to sell it, I will have a complete log of all the interactions I had with AI and very specifically

what the AI did with my work or what we did together in a working session. And I've set this up for several other authors too, but I think like the running documents feature is great. You can give a rule to the AI is every time we finish a session, let's have a summary of what you did and what I did.

Kathleen Kaiser (02:12)
Very good. So you're drafting in Claude.

Julie Trelstad (02:15)
I'm not drafting in Claude, I'm drafting in Microsoft Word. And Claude is sitting in on the side and I'm saying, okay, I will upload my finished chapters for instance. And I say like, I'm at this point in the outline and this is what I'm thinking about writing in this chapter. And I'll work with Claude maybe and just say, these are the general beats we're gonna write based on where I wanna go and what I've already done.

And then I will go off and write. And when I'm done with it, I'll go back and say, does this match? made these creative changes. I even wholesale changed character names. And then it went back and made a revision memo for me that is also keeping live. when I'm ready to go back over the whole draft, all the decisions I've made will be recorded.

Adanna Moriarty (03:00)
It's kind of like working with a developmental editor as you're writing. Is that sort of what it

Julie Trelstad (03:05)
Exactly.

sort of think like that old like, what is it? It wasn't Sleepy Hollow. they, there's this, in the old days, there used to be these actual desks where the writer sat on one side and the editor sat on the other side of the desk and they used to pass the paper right and back and forth as the writing process is going. And that's how I think of it.

Kathleen Kaiser (03:24)
Hmm.

Adanna Moriarty (03:24)
I actually really like that. I mean, it's a smart way to utilize something cause I mean, when you're writing and you're creative, you can kind of get off into tangents and, maybe drop the story or just lose it along the way. So, I kind of like that to, keep yourself in check and make sure that, you're hitting points that keep the story moving along and keep it interesting.

without actually using AI to alter what's coming out of you.

Julie Trelstad (03:52)
Exactly, exactly. And I will give it a draft when I'm finished and I'll say like, is there enough interiority? Do you understand what the protagonist is feeling? And it will give me like a list of critiques and say, well, in this part, I really can't see it. You should add some more detail here. And I think it is very much like working with a developmental editor because I know what questions to ask too. Like I know what I want the final output to be.

Adanna Moriarty (04:15)
So was this just experiment? How did you figure out how to get it to behave like this so you get exactly what you need from it instead of it overriding you Was it just sort of, experiment and try and go back and forth until you find the prompts that work?

Julie Trelstad (04:32)
Kind of. I've taken a number of courses in AI and I recently took this thing. It was like 30 days of AI challenge. And if you all want the link to that, it's a free course you can take. I've learned some techniques. And one of the main things I learned is like, if you have an idea of what you want to do with the AI, have a long conversation with the AI and tell it to write its own rules.

So I spent some time in one session with a chat telling the AI exactly how I wanted it to behave. And Claude allows you to have, when you have individual projects, and the same is true of chat GPTs, custom GPTs, or Gemini's Gems, is that you can create a set of instructions. So I worked with my AI to create the set of instructions that I used in that project. So it always behaves that way.

Adanna Moriarty (05:20)
I work with ChatGPT I I prefer ChatGPT I've played with all of them. I started a while ago and you know, it was like trial and error. And I mean, I feel like I get really good output out of it. And I think that the people who complain about it don't know how to use it.

Julie Trelstad (05:26)
I like them all.

Yeah, it's just like, you know, a relationship. If you don't tell people how to treat you, they're just going to treat you like generic person.

Kathleen Kaiser (05:48)
I also think it's also how we relate to things. Some people, it's like the Mac and the PC type of people. Your brains are wired differently and you can get this to do it, but you can't get that one to do it. Okay. And I think with it, since there's so many good options out there now, find the one that understands you and can give you back what you're telling it to do. It can't be a mind reader. You have to have a pretty clear idea of what you're doing.

But if you give it to it correctly and maybe tweak it a little bit, then it'll do it. And I find it really, really a fabulous tool.

Julie Trelstad (06:26)
I think so too, but the more you train it and customize it, the better. I think, and just keep asking yourself, I actually did another thing, it can create a word manuscript. And I do that more for marketing materials. Claude now has a skill that can actually output a word doc. And I was, okay. Yeah. And I was frustrated with it because it didn't, it didn't do,

Adanna Moriarty (06:43)
ChatGPT does that too.

Julie Trelstad (06:48)
H1, H2, H3 heads, know, the way you like as a publisher and you do produce, it just created a flat document with things that were bolded and underlined. And if you're a production editor, you wouldn't understand my frustration.

Adanna Moriarty (07:00)
ChatGPT does headers and subheads and.

Julie Trelstad (07:03)
Well,

Claude didn't, so I taught it how to do it. Yeah, you can actually, I said, you have to do this. And so there's, you create a piece of code and then you put it in your own version of the software. It's called a Claude skill. And now it gives me beautifully formatted documents, even in the fonts that I like it to be in.

Kathleen Kaiser (07:07)
Good.

Adanna Moriarty (07:24)
I mean, I feel like they all vary anyway in what they're really good at. Like perplexity is really good as like for web designers, like when you need code and stuff like that, like it's output for code is much better than ChatGPT's as I've never tried Claude for code. But I mean, I do think that they, you know, they vary in skill for what you actually need to, cause I use perplexity for that's the only thing I use it for is when I need good code.

Julie Trelstad (07:53)
Excellent.

Kathleen Kaiser (07:54)
Yeah, well, Adanna is a web designer too, so she needs a lot of coding for things. ⁓

Julie Trelstad (07:59)
Yeah, I've been experimenting with lovable, but that's a whole other level.

Kathleen Kaiser (08:03)
just want to learn how to make the cute little videos. That's the next thing I'm learning. I actually have a client who is coming up with some really cool videos and he had never done them before. And he said he took a weekend and he kept doing it. And what he came up with was really good. And he says, now I'm going to refine it. And I said, you got to show me how to do this. I really, like it. And they're not cartoonish looking. They actually look like real people in real locations.

Julie Trelstad (08:26)
amazing. Yeah.

Adanna Moriarty (08:28)
I

mean.

Kathleen Kaiser (08:28)
It's

really, really cool. And the voiceover does not sound mechanical at all. They're really getting good at the voice.

Julie Trelstad (08:34)
And see, when you talk about using AI in this way, isn't this very creative though? mean, to get that output, have to be creative and like have a vision to make it do exactly the things you want it to do.

Adanna Moriarty (08:46)
I totally agree with that. And this is my argument when people are like, you know, AI is going to take over or AI this or AI that. And to me, it just tells me that they haven't really delved into the process of trying to make anything with AI. Because if you did, you would realize that whatever is out there is still coming from other human beings. It's just people who have learned how to manipulate whatever AI they're using, whether it's ChatGPT or midjourney or Claude or, there's now

plethora of out there to make videos and all kinds of stuff. I mean, you can't just give it something and say, make me a video on. You have to be really specific. And you have to give it locations and all of these things. And then refine, refine, refine until you get it to where you want. And I think that's really important when we talk about using AI is that, there's still work that goes into it.

hours, know, hours and hours of work.

Julie Trelstad (09:44)
Exactly. It's absolutely not the push button and done. Although, you know, I think there are people who are doing things fully automatic in a way that's not good for the rest of us, but that are copy, copycats, piracy, a lot of that stuff is happening too. But I think, you know, on the other side, there's some really interesting creative stuff.

Adanna Moriarty (09:55)
I agree.

So I like to always address ethics when we have someone on that, is knowledgeable about AI and the process of it. I mean, I do think that ⁓ you just segue perfectly into it. I do think that there's, an ethical way to use it and a non-ethical way to use it. what do you think about that? when you're looking at it, other people and yourself, how do we keep ourselves ethical while we're using AI?

Julie Trelstad (10:30)
Yeah, I think the same way you would keep yourself ethical using other people's materials. when you talk about unethical, I think about a case last year where, you know, I had somebody I was working with and she published her book with Penguin Random House. And then within 10 days, there were six copies of a book on the same subject by a similarly named author with AI generated covers.

that were priced lower than her book. And I think it is completely unethical to copy other people's work. Wasn't there a case last year where a romance writer left a prompt in her book, basically saying, make this passage more like my competitive author? I heard about this story and it arrived in the book. And how embarrassing for her.

Adanna Moriarty (11:15)
How did I get past editors? mean, my

Julie Trelstad (11:18)
Well

It's one of these rapid release romance writers. If you're intent, I'm going to knock this person off. I mean, I don't think you're thinking that. But I think if that is your goal, if you want to like, copy and make it better, and you're seeing all these book marketing scams, say where people are using AI to like fish authors and, prey on them.

⁓ I think that that's not good. It's just like, think if in your heart what you're doing is good. I don't know if that's a good ethical thing.

Adanna Moriarty (11:45)
No, I do think that's a great I mean, I feel like the last time we had someone on with AI, they told us a story similar to the, New York Times bestseller, and then a slew of books that come out that are similar and like almost the same name and kind of undercut their sales. To me, I think if you're using it to create, right, like you're like, I want to write

a romance book about blah, blah, write it for me. And that's the involvement of your creativity. That's wrong. That's my opinion. Like write a book, you know, use it how you use it. I think that's great. But if you're using it to, pump out complete works without much involvement from yourself, I don't think that's super ethical use of AI personally.

Julie Trelstad (12:29)
Exactly. I have an AI policy on my website and, you know, just to describe how I use it. And I think that's probably worth, you know, anyone who's using AI, just even in your diary, just think about like, I'll do this, but I won't do that.

Adanna Moriarty (12:42)
Right.

Kathleen Kaiser (12:43)
Well, it's also when you're doing it this way and you're pumping it out, I think that's one of the reasons when Amazon changed to A10 last year, their algorithm, it's going out to see, do you have a website? Do you have a social media presence? They're trying to find the scamsters because they got deluged. mean, in '24, early '25, it was a mess on Amazon with people doing like what Adanna said.

Write me a romance, historical romance with these kind of characters. Here's their name. You know, they're just throwing junk out there. ⁓

Adanna Moriarty (13:19)
And

the dupes of best sellers, you know, like, it was just inundating the market. And I mean, I think that's why they updated to A10 was to try to weed out, fictional authors from the real authors,

Kathleen Kaiser (13:22)
That's sellers, right?

Adanna Moriarty (13:36)
Us that are authors, I feel like, we work hard on our social media presence. Like we make sure that we have a website, we do all these things. And then somebody comes along and can thwart it by using AI the bad way.

Kathleen Kaiser (13:49)
Yeah, well, there's scamsters and everything in this business, especially since authors are so isolated. You know, they usually work alone with their computer in a room. They don't have always a community. It's one of the things with the Writers and Publishers Network, we put out scam alerts. And in fact, next month, we're going to have on Talking Book Publishing, we're going to have Victoria from Writers Beware. ⁓

Julie Trelstad (14:11)
Oh great, she's awesome.

Kathleen Kaiser (14:12)
She and I and Desiree Duffy did a thing at Bookfest about a year and a half ago about it because that's when the AI letters were starting to just descend and now they have all of them and they're using real authors names like Rebecca Yarrow. I get the emails from a lot of our clients. The email comes to me that's on their website and now I've gotten three from her today for three different authors and they're using real authors names.

I mean, it's really, really bad. And it's all AI generated. I think all they do is read the description on Amazon or wherever, and then they write the piece up. And it's very friendly and very first person and as a fellow author and all that stuff. And it's like, oh, God, it's a nightmare in a lot of ways. I'm really so sorry for authors that are more isolated.

Julie Trelstad (15:07)
For sure, yeah. it's because of that, that's probably why these scams are happening.

Adanna Moriarty (15:13)
And authors are vulnerable because when somebody's like, I'm interested in your work, we're like, really? You are? All of us are. All of us. Even the top billed guys. You're like, I wrote? For real? Are you a real person? So I mean, I think we're vulnerable to it anyway because we spend so much time.

Julie Trelstad (15:27)
Definitely there.

Adanna Moriarty (15:36)
with our work and it's like making babies, right? And then somebody's like, I'm really interested in that. And it's easy to fall victim to it, I think. I realized that we were so eager to talk to Julie about AI that we didn't let her actually tell us all the things that she does. Because she's a writer, she's a marketer, she has this cool thing where she's helping.

authors with AI and rights and all of that. I was just thinking maybe we could talk, we've talked about the writer part of Julie, maybe we could talk about marketing and then like segue into, you know, the protection bit.

Julie Trelstad (16:16)
Yeah, I'd love to give you like, like all of this is I'd love to talk to anyone about AI and writing right now. It's so much fun. I've had a really interesting career in publishing. I've been in the industry for more than 30 years and I like joined at the moment that desktop publishing arrived. Like so my first job we had paste up galleys on the wall and I've been around for like

the first round of eBooks and CD-ROM and working with authors when they had to start doing social media and now AI. So I've always been really interested and been in that place where I'm like the next person doing the technology. And I started really getting involved with others in marketing when I was the digital rights director for three years at Writers House, which is a large literary agency in New York City. I was hired there to run their self-publishing program at the beginning of the Kindle.

version and we turned all the Sweet Valley High books into ebooks. were like 130 of them. And, but what I discovered there is like the agency was a lot less interested in self-publishing than it was in helping their authors build these digital platforms. And it really became clear to me that the, industry had very much shifted. So the industry shifted when I started in the 1990s and I was publishing books.

⁓ Barnes and Noble stores were opening like every month there was a new huge opening and we were selling print books like crazy. I worked in home improvement books and literally everything we created, we would automatically sell 4,000 copies to Barnes and Noble, which was just such a different world and environment than it is now. And the reason that is is that, and I think traditional publishing has maintained that connection to the retailers.

and to retail support. But at the same time, as you all know, like 90 % of books are bought online, on Amazon or anywhere else. So that online presence isn't very well covered by publishers. They aren't great digital marketers for the most part. And that piece of the marketing is left up to authors. And so like in my time in the industry, it's completely shifted. But authors, I honestly think

publishers could do a better job and not push as much down on authors. But this is where we are because some authors are really good at it because some authors are really good at it, we all have to be really good at it. And so I think now is just like, you my feeling, you know, it's complicated. My feeling is that no author should hire anybody to help them with marketing until they understand what they're doing. And I know like you do a lot of education with your clients, like authors,

really need to know how books are marketed and sold, how readers discover them. And once they understand that, then they can build their platform or create a platform where they're interfacing with great marketing groups and influencers and building their own communities online. I don't think it's anything different than other people do, but it's really more of a mindset shift that I think that

Authors need to think of themselves in the digital space as like one note of connection where they need to build a web around them. They don't need to build a giant audience. They need to build the connections to the people who have the bigger audiences. And that's the only way as an individual, you can really do that without becoming a full-time social media star.

Adanna Moriarty (19:31)
I totally agree with that. mean, that is part of what I do when I have a new author website client. I mean, we always do strategy too. So I never just build a website. We work a little bit on branding and we work on strategy because what's the point of a website if you have no strategy? And I think that it's also part of the reason why I built our GPT that Kathleen and I, we created a business around, which is Pro Book Launch.

because that gives you strategy that you can use marketing material. mean, really good marketing material. Cause I mean, we spent over a year training it, know, rewriting the training and making sure that it put out what we would put out. We tested it on clients. Like we did all of that to make sure that it wasn't reading generic AI, that it was, you know, it would be the same that if we wrote it, because we did that for years, like where you just.

have to sit down and write all the copy for everything, I mean, website copy, back cover copy, press release copy, social media all of it. And I think that, I mean, marketing for the individual author, whether they have a big deal with Simon & Schuster or whatever, or they're self-publishing, if they don't understand marketing strategy, their books are not gonna sell.

Julie Trelstad (20:50)
Absolutely. when I've worked with a number of, I work with as many traditionally published authors as self-publishing authors. I have something called the Authors Platform Studio, which is a group coaching program where people can come in and learn. Actually, I help them write their own GPTs. And then like the GPT I've created actually came out entirely from my years as an acquiring editor. And I always saw like the main problem. I always feel like authors have

is nailing their positioning. And usually that's earlier on before they get to the marketing is like, who are they as an author and how do they fit into the larger ecosystem? So I have a GPT called the little black dress that actually looks good, know, gets the thing that looks good on every author that really helps them be super specific in who they are. So when they're making pitches, either if they're going out for querying or if they are like, you know, creating marketing materials next.

as well as to help them identify their closest competitors and how they're alike and how they're different. So they can tell that story.

Adanna Moriarty (21:50)
Yeah, mean, know, comps are so important. I think comps are so important. just in even being able to see, because I mean, I think we find often people have no idea what niche they're like. Like, think there's something or they want to be or they're like, this isn't a genre, it's for everyone. And we're like, that's never the case, like, ever.

Julie Trelstad (22:05)
Yeah.

Adanna Moriarty (22:15)
You know, so being able to like, you know, run your book and see what like actual real comps come up and why is such a piece of the puzzle that I don't know, I just.

Julie Trelstad (22:28)
Well, as an acquiring editor, always had to bring the comps and I always had to prove to my publisher that this book was exactly like this book that was bestseller last year, but it's totally different. And then like, where's this Venn diagram? Where is it like the comps, but different? So.

Kathleen Kaiser (22:43)
Well, and I think too, one of the things we do with Pro Book Launch is that because I'm a white person, why do we do this? So we haven't tell them, this is why you use this. And people go, well, why do I need comps? It's because if you're going to be advertising, you need to know if you're specifically going after what type of titles, or if this is a someone who's very much in your niche, you want to be up on the page where they are.

I was working with a client a couple years ago and it was with Simon and Schuster and they had one gal who her old job was to pitch the comp and buy the digital ads on Amazon. So they were in the first three slots. Again, know, they're both against the other, you know, the comp title. And I learned a lot from her. was really kind of interesting, but it was like, Whoa, you know, they're when you can get that micro

you've got a chance of selling because I know like on Amazon when I'm looking at something I say, what else is here? I always just glance. I never look at extras. I just look at the ones there. And I have found really good books that way.

Julie Trelstad (23:52)
Exactly. it's just like, it's kind of interesting, like a lot of people don't buy any books at all in every year, but the people who do buy books buy a lot of books and they don't hesitate, they don't even hesitate. They're like, oh, that's kind of interesting. Bye. So if you can have your book showing up with those others and you know, strategy, I always promote too. It's like, who are your 25 most competitive authors or even

people who talk about books like yours, the influencers, those are the people you need to be talking to on social media. Just having the conversation with them, being seen in public in this group of authors.

Adanna Moriarty (24:26)
I always tell clients to like back to that strategy to, know, if you get good comps, you can then go look them up on social media and see what they're doing on their social media that's working for them. And so you can mimic it, not identical, but you can take their concepts in your genre, like say you're a thriller writer and, you know, match what is working across the board.

and thriller authors. Because what a thriller author does is going to be totally different than what a romance or a horror author does. They're just the way that you structure your social media and even the personal things that you would post on your social media matter when you're a genre. When you can pull comps and look at that and really put yourself in a little niche and work on it.

I so I mean I think it's also important to be able to and it's hard. Months and months of months of like work if you don't utilize AI.

Julie Trelstad (25:30)
Exactly, I was just going to say, it is awesome for doing this.

Kathleen Kaiser (25:34)
Yeah, does it.

Julie Trelstad (25:35)
It's awesome. know, it can find, can pull up the research, it can find the social media handles, it can, you can even set up following and like it's, yeah, it makes so much of this so much faster. You can even like do screenshots of all these other authors and load them into chat GPT and say, describe this style. And then you can take that description and take it into an image generator and create your images or your videos.

So like, I think that AI can really help you, you know, do this a lot.

Adanna Moriarty (26:06)
faster.

Kathleen Kaiser (26:07)
Yeah, that's what I keep saying. Marketing is a science. It's a tool. if you can, AI is the most incredible tool for marketers. I read something the other day, because I get a lot of the emails from AI, different promoters and stuff. And it said, all those people that are worried that AI is going to take your job. No, it's not. It's going to be the person who's learned AI. They're going to take your job.

Adanna Moriarty (26:31)
Well, I was just going to say that like as, you know, as a tool for marketers, like we know what we're looking for already, right? Like, cause we, have knowledge of it. So if you're a person who has no knowledge of marketing and you jump into chat, she'd be T or Claude or whatever, and you say, I'm an author, write me marketing material. have zero idea if the output is good marketing material because you don't already have the eye for it. Like,

we were trained for it. I do think there's a difference. I mean, you can learn it, but if you're just like, ⁓ I listened to this podcast and they all said AI is good for marketing, but you have zero ideas about marketing, it's not going to work for you.

Kathleen Kaiser (27:18)
No, no,

Julie Trelstad (27:18)
Yeah, exactly. So it's really fun. And I like to create little tools like, you know, an email marketing builder, GPT, which I've just been playing with it, you know, is like, think about what you do, anything that you do that's repetitive, and you can do it over and over again, you can train an AI to do.

Adanna Moriarty (27:35)
I built one for our podcast. Because I was like, you know, I do the same thing all the time. Like, you know, I do the website description, I write a post for the guests, I write a post, I write several posts for us, like, you know, I was doing all this stuff. So I just took everything that I did, and previous ones that I'd written, and like uploaded it into, you know, like, to help create the training.

Kathleen Kaiser (27:37)
Yes.

Julie Trelstad (27:38)
person.

Adanna Moriarty (28:01)
And then I made a GPT that does all of our marketing material. In our voice, with our brand, with our strategy, like it's just, you know.

Julie Trelstad (28:04)
Yeah, I like it.

That's, yeah, again, back to the creative. I do want to talk about AI, right? It's even though I know it's a complete segue.

Kathleen Kaiser (28:17)
I thought we should move on to that. has been really, I'm enjoying this conversation so much.

Julie Trelstad (28:22)
I

Adanna Moriarty (28:23)
I could talk about this forever. Like we could probably do like a three hour podcast to talk about AI. Because I love the complexity of it. know, like I really, the good and the bad, like I like to discuss all the ways that you can use it for the good and also like dive into like the nitty gritty bad. you know, I don't know. I could talk about AI forever. But yes, I agree that we should discuss, you know, the

the thing that you're doing with AI rights, because I would like to know more.

Kathleen Kaiser (28:53)
Yeah, I get one thing I'd like to add in those. I think we need to look at AI as an opportunity. absolutely. I like to add it. And you have to get a lot of people don't grasp opportunities and they miss so much. And when you're an author, and especially if you're an indie author, you've got to look at every tool that will help you. And then you have to learn how to do it. You can't sit there and think you're going to ask the right questions.

You need to take a few, like you said, some courses in different things. The 30 day, whatever that is.

Julie Trelstad (29:24)
I'll get you a link to that. It's a free course. He offers it every three months. It is mind-blowing and simple. So I think that's a really good one.

Kathleen Kaiser (29:31)
that link on our website because I think it'd be great. why don't you tell us about this rights thing? That sounds really.

Julie Trelstad (29:37)
To talk about the dreadful and the opportunity all in one, it gives you that opportunity because I think like the main thing that authors are resenting right now, when we all found out that the large language models were actually made up of scraped author content that they took millions of books. And really interesting story I heard about that from some of the people who worked on the development teams were like, apparently Chet's GPT was a thing like three years before they released it.

It wasn't until they started giving it fiction, like as a diet to eat, that it could start talking to people. That it actually needed that the stories that came out of fiction to make the language models conversational, which I find so true, right? Isn't that? Of course. It's not about

Adanna Moriarty (30:24)
I didn't know that. That's interesting. it needed our creativity to be able to talk to us.

Julie Trelstad (30:31)
It did. It needed creativity and it needed that dialogue and human because like before that they were just training it on technical manuals and memos and like any printed thing they could find and it just was not coming off as human and once they gave it fiction it's just like hello.

Adanna Moriarty (30:47)
I am a feeling big. We always joke that it was really trained on romance novels because it's better now. It's better now. like, like two years ago, everything that it wrote, like it used every adjective, adjective it could find, like, which is very romancey. So that was always our kind of joke was.

Kathleen Kaiser (31:07)
And

the exclamation

Adanna Moriarty (31:09)
I think you're

Julie Trelstad (31:10)
think not

wrong because, you know, AI just works on a probability model. So it always is just delivering the most likely next thing that would come from a chain. And if its database was made up of books and the largest by far e-publishing genre is romance, there is an excellent chance that it was fed a diet high in romance content.

Kathleen Kaiser (31:34)
Yeah, it was. I'm convinced of it.

Adanna Moriarty (31:37)
⁓ totally. mean, well, I mean, because they're the most books are romance. But I mean, my goodness, like when we first started playing with it, like when I really first started playing with it, I was like, you know, you'd have to take it and edit it so deeply because it was like, wow, I, I don't need these many exclamation points and adjectives. Like, let's comment.

Julie Trelstad (32:01)
It's like, yeah, my system instructions say no L-Y words.

Kathleen Kaiser (32:05)
But let's go back to the right.

Julie Trelstad (32:07)
Yeah,

so back to the right. So this huge amount of scraping happened. And as a result, lawsuits, know, Anthropic had to pay $1.1 something billion to authors for the things that it scrapes. There's still an open AI case going through. And there's also sort of pushback on that Anthropic settlement saying that that's not nearly enough.

and authors and the authors guild are saying like, this is UOS big time. And these AI companies are like, they're getting a lot of cash and like, okay, so what? No big, no big deal. Of course, AI industry is bigger than publishing by what a factor of 10 already and it's a brand new industry. But so there's a feeling though, like if these AI companies had gone to say Random House when they were doing their development and said, hey, we wanna use your authors work for this thing.

Of course Random House would have said no. mean, first of all, publishers would have no idea what those rights were and they would have never said yes. They just wouldn't have. However, now we know. Now we all understand what the situation is. And in Europe in 2017, actually, they passed a law requiring any kind of software development or AI companies if they were using copyrighted works that they had to license it and pay for it.

And the problem with that is like, in order to do that, still, you actually probably have to write to the William Morris Agency, right? The copyright clearance, their databases and stuff, you still have to like, basically send a letter and see if you can get that permission. So there's still not an easy, seamless way to do it. And a rights agency also probably would not know what to charge for this because like, what is it even worth?

⁓ It's like all new territory. We know how to sell foreign rights. We know how to sell television and film rights and options. But how do you sell rights to an AI? So to solve the problem in Europe, a fellow by the name of Titusz Pan created something called an ISCC code. So that is the International Standard Content Code. And it's like an ISBN number.

But instead of being something you add after the book is created, it's actually, this fingerprint is a digital fingerprint made out of the book's contents. So it's not like the whole copy of the book, but it can be used to identify this one particular work electronically. And so if that work is used electronically or the rights, it will be able to identify it and give you legal standing.

So last year, I've been working with this company called Streetlib, which is an international ebook distribution company. And their founder, Giacomo D'Angelo got together with Titusz and they decided to create, they were like, my God, what is ChatGPT going to do the book publishing industry and what can we do about it? And so they created, they took the ISCC as this idea of like a digital identifier for books.

Although the ISCC can also identify a song, a podcast, an article, like any kind of content can be given a fingerprint, not just books. But AMLIT was designed to be a book-specific registry, the same way the Bowker ISBN database is. So what it does, it allows you to create an ISCC for your book in a very streamlined way, and then create a machine-readable code that tells AI companies.

Very geeky term for this are the text and data mining rights, also known as TDM. And what that means is there's four kinds of rights we've identified. Well, one is no rights at all. You may not touch this book and it may not be used in any AI application. And that's kind of the state of the art right now. Any book that's getting an identifier is usually starts at that point. The second one is model training.

So next time somebody comes up with the next generation of open AI and they need to train a model, yes, I'm in or no, I'm out. But if an author opts to be in it, they would generally be paid a one-time fee because it's like a one-time occurrence. And then there are two other rights that are super interesting. One of them is research results. So can your book be returned as a search result?

And the other one is, your book be used in a generative AI? So in other words, can your book be used in a chat bot And for those, you would get recurring revenues based on how much of your book is used down to the three sentence level. So it basically pays you out like Spotify would, like with a token. And you can get residual, yeah, go ahead.

Adanna Moriarty (36:41)
was going to say it kind of sounds like, you know, when we first started with like streaming music, like LimeWire and stuff like that, like how it shook up the music industry because like nobody, none of the artists were getting paid because like you could upload a CD and somebody, you know, somewhere else in Florida could download your CD. So it was like bypassing the music industry, which created things like Spotify and stuff to make sure that the

musicians were still getting paid.

Kathleen Kaiser (37:10)
Yeah, the Napster started the real steel.

Adanna Moriarty (37:13)
Yeah, napster and lime wire. I couldn't think of the other one.

Julie Trelstad (37:16)
my gosh, yes, exactly. It's that kind of thing. And AMLIT though is open source. It's not like, you know, I've seen like these other things like created by humans and cashmere, they're like creating proprietary standards around AI rights. What we're trying to do is create like an open source, just like an ISBN. Everyone can get one. You can be a publisher, you can be an author, you can be a content creator, and you can participate in this.

Adanna Moriarty (37:40)
So is it?

Julie Trelstad (37:42)
I'm

Adanna Moriarty (37:43)
Is it, I mean, is it free? it paid? it, can you have already published work? Like, how does that all work?

Julie Trelstad (37:51)
All of the above, mean, number one, we are brand new. Like this company was started just the end of 2025. So we're in our first phase. Because we are a partner company with Streetlib, if you have your books distributed through Streetlib, which is similar to Draft2Digital, only it's more Europe based. Like we have more Europe distribution on it, even though we also have Amazon and the usual suspects. If you...

upload your book into the StreetLib system using their Pro Plan. I think that's 299 one-time payment. You can get an F-Free Amlet for all of them. So that automatically gives you the ISCC identifier and gets you to the opt-out level so that you at least have that identifier. If you want to license your book to an AI, that's coming in the future.

And there will be some fees associated with that. It's usually going to be like a percentage for the platform. And then there are also going to be some fees for if you want your work notarized. But even with the free version, you get a timestamp and you have legal basis of claim. And what's super interesting, like once you get an AMLIT for your book, we actually have on the application,

If you see somebody who may have copied your book, can check it against your file, against your digital fingerprint and know immediately, is this like an 80 % or more match? Is it a copy of my book or not? Which gives you legal standing and proof that the book is yours. So that is accessible to anyone now. And we're working on rolling it out more widely, you know, over the course of 2026.

Kathleen Kaiser (39:31)
Okay, so I have a question. Say we're a nonfiction author. Okay, and you actually would like to be quoted in articles or different things. How does that work with this?

Julie Trelstad (39:41)
Well, you would just like so that's kind of the next level. I just say like this licensing out, you know, pricing and that whole infrastructure still being built. If somebody does get to the AMLIT and they find your book, the AMLIT database is machine readable by AI systems or by like anybody who's building. So if somebody would say building a news app and they plugged into our our API, our connection.

and that app developer searched or their AI searched for content like yours, a non-fiction author that they wanted, it would then return a result and tell that person who they had to contact to license the rights for. So they would say, contact Kathleen.

Kathleen Kaiser (40:22)
Yeah, so that would mean like ⁓ say ⁓ an author is, you know, whatever their topic is and they're known about it. So they a licensed quote because a lot of, especially with college and college books, mean, professors putting out books, they're always quoting in different things like that. would streamline that system.

Julie Trelstad (40:43)
Absolutely. And I think by the end of the year, we'll start to see automatic payment for that. So that's like, we're building it in levels. First, it's like, do not touch. And that's activated and fully working. And then being able to turn on these other rights is coming soon. The thinking is, and this is a hypothesis at this point, that those rights are going to be based on the list price of the book. So if a textbook is $35,

however many tokens that's broken into however many chunks, each of those tokens and say a chat bot brings up 30 tokens and there's 100 in the book, they would get paid $10 for that usage for instance. So it would be based on the usage based on the list price. It's our hypothesis. and different obviously like training rights might be like 10 X the cover price or something like that.

or even more, just depending on the value of the book. But I mean, what we're seeing, what we're really doing at AMLIT is just creating an infrastructure and we're working with other licensing agencies who are going to build like these deal making platforms on top of it. And even authors will be able to make their own deals.

Kathleen Kaiser (41:49)
Yeah, that's really interesting because that is something if you're especially a nonfiction author, you're sort of an authority or expert in a certain area that to be able to get your work out further. read a lot of nonfiction. And sometimes I'll see a name and the quote that's in a book. It was like, there was one book that came out last May, Pronto, and she I think a quarter of her book was all of her

Julie Trelstad (42:02)
Exactly.

Kathleen Kaiser (42:16)
indexing and everything that she had done. mean, it was an amazing book. But I found it's like, well, that's really interesting. So then I went to find out who was this person. You know, so if they got money from it being in her book, it could also drive you to be finding them.

Julie Trelstad (42:30)
Exactly. And there's also thinking like I told you, like search results are one of the options. ⁓ We think that some people are going to allow search results for free as long as there's an attribution and a link to go buy the book in that search result for discovery reasons. Which I think is super interesting because I think more people would be inclined to include the free but advertising in their app, right? And I can tell you something even more horrifying and also hopeful.

Kathleen Kaiser (42:56)
Yeah.

Julie Trelstad (42:56)
⁓ it's like one thing we're seeing, which is so bizarre. And I don't know if you were going to say this, like people are discovering books in chat GPT. They're using their chat bots to find books and to find content. And there's a finite number of humans on the earth who read books. but every chat bot and AI or AI agent can read thousands of books an hour. And if you are sending out a request as a human.

to send your AI chatbot out to find you all the information on the latest advances in bicycling. That chatbot's going to go out and check out all these sources and read it, which then will pay the authors for the content that that chatbot brings in. What I'm trying to say is the amount of

reading that can happen of a book is going to increase by more than a thousand fold in the next couple of years. It's because for every one human who reads, their chat bot can read a thousand times more information and pull it back and synthesize it for them, which is just kind of mind boggling, but also suggests that this sort of like licensing by little bits and pieces could be a huge opportunity for authors.

Adanna Moriarty (44:07)
Yeah, it sounds like it. Well, I definitely would like to add a chunk of this part of the conversation into the show notes. mean, I don't know if you have like a paragraph or something that you can send me so we can add it onto the webpage, like outside of the episode description, because I do think that it's really interesting. And I do think that people are going to be really want to know about

Kathleen Kaiser (44:10)
That's really amazing.

Adanna Moriarty (44:35)
you know, protecting rights as authors in the age of AI, because I think that's the place that makes people so afraid of it is that they're just, you know, like they're going off this information that we know that it was sort of just eating books, right? And so I mean,

Julie Trelstad (44:54)
Eating and then throwing them up,

Adanna Moriarty (44:56)
Right, like the hungry, hungry hippos, you know, just ram, ram, ram. So, and I know for me, like, there's always this thing where I'm like, okay, I use AI, and how do I balance it? And I think that and making sure that also my work is protected and, you know, all of that kind of stuff, because it is scary. But

As a previous guest on the show said, I mean, they probably are not going to want to steal your work until you're a New York Times bestseller. most authors don't actually have to worry about it.

Julie Trelstad (45:25)
for that that copying scheme. Yeah.

Adanna Moriarty (45:27)
Yeah.

But yeah, no, I think that this is such an important part of the conversation and thank you for bringing it to us. Like, I really appreciate it.

Kathleen Kaiser (45:38)
Yeah, mean, authors always want to know where they can make additional money from their book. You know, I mean, most authors don't make any money off their books or minuscule. But to be able to do that, it would be, you know, it's just another revenue stream. Why are it? Why not?

Julie Trelstad (45:55)
Yeah, and like I think, you know, just in that same vein, authors should be looking at all the rights venues, the foreign rights, the film and television, the language rights. And there are just so many different ways to sell a book that we don't always talk about that can compound your investment in your time and your creative work.

Kathleen Kaiser (46:17)
Yeah, and I think selling a book is the part that most people have no clue. Absolutely no clue. Amazon's it.

Adanna Moriarty (46:22)
I mean, yeah,

Julie Trelstad (46:23)
Thank you.

Adanna Moriarty (46:24)
hard. Yeah. mean, selling a book is hard. Yeah. I mean, it's hard for us and we, we know tricks. I mean, I think it's hard, you know,

Julie Trelstad (46:33)
think the business model is stacked against us. you know, just my, all my years in the industry is like our modern current book industry is based off of what happened in the aftermath of World War II when paperbacks became the thing. And, but in order to have a profitable paperback, know, hardcover books were kind of bespoke. They didn't expect to sell a gazillion copies. Paperbacks were expected to sell a gazillion copies. Boom.

And so it's a mass market business. And even though a lot of, you know, a lot of books will sell fewer copies, the economics, I mean, the price, the cover price of a book has not risen anywhere close to inflation. Like books are far cheaper than, thank you, I'm glad. But it's bad for authors. You know, it's just like the compensation doesn't match. And like,

Kathleen Kaiser (47:20)
Yeah, we.

Julie Trelstad (47:25)
Okay, that worked if there were only four publishers and they were printing massive amounts of books and that and they could earn profit off of those pennies, pennies, pennies. And they had guaranteed places to sell them and all the newsstands and there weren't things undercutting it. So the fundamental economics of book publishing, there's no supply and demand. There's way too much supply and way too little demand unless now we count these chatbots that are going to thousand time the demand.

So, you know, it is like a structural inherent problem. So even saying it's hard, that's true. The odds are truly stacked against you. There's certainly things you can do and set your expectations in the right place. you know, know that this industry is really, it's a hard one. It's a puzzle for sure.

Kathleen Kaiser (48:08)
It's like the music industry. came out of that. Even at the record labels, we'd have 15 releases that month. We figured out the three we'd push because we knew we would sell those. And then figure out the second subset.

Julie Trelstad (48:22)
Exactly

what the big publishers are doing. They're sending 99 authors and they're promoting one. And a lot of those folks who don't get that A-list promotion then never see sales. It's a chicken and an egg. They didn't get sales because they weren't the one who was anointed.

Kathleen Kaiser (48:37)
But that's the truth of business. You take the product that sells. Everybody's, it's a money market. You want to make money. And if that's what you're trying to do, you've got to figure out how to get yourself in front of that audience to be able for people to buy you. Discovery is a big part of the whole.

Julie Trelstad (48:56)
and AI and social media make it possible.

Kathleen Kaiser (48:59)
Yes,

they do. Yes, they do. That's why they're so essential.

Adanna Moriarty (49:03)
I, ⁓ for one, think this has been a great conversation. I would like to have Julie back and talk more about all of this, because I think that there's more conversation here to have, obviously. We're at an hour and we could keep going.

Julie Trelstad (49:17)
You

Kathleen Kaiser (49:17)
So thank you, Julie, for joining us. We will definitely have you back. This has been a fun conversation.

Julie Trelstad (49:23)
No, I love talking about this and love enjoy the conversation with you all too. I know you're playing in the same plant sandbox.

Kathleen Kaiser (49:30)
Well, your enthusiasm is what I like because you make it understandable. People need to be able to relate to it and grasp it because it's out there to help them. It's really out there to help them. So we'll be doing that. so that's it for this ish episode, excuse me, of Talking Book Publishing. Thank you for joining us. And we'll see you in two weeks with our next episode.

Julie Trelstad (49:35)
Thank you.