091: AI in Book Writing + Publishing

 

A few weeks ago, I solicited thoughts and feelings from my community about AI entering the book writing industry. I share the concerns and general sentiments I received on TikTok and Instagram, and I also talk through my own complicated feelings as a writer and editor (and someone whose job could be very much impacted by AI in the next ten years.)

----------

Mentioned in this episode:

https://www.newyorker.com/culture/infinite-scroll/my-ai-writing-robot

https://www.vulture.com/2023/07/sarah-silverman-openai-chatgpt-lawsuit.html

https://www.businessinsider.com/openai-copyright-lawsuit-authors-chatgpt-trained-on-books-2023-7

- Click here for ways to work with me + a free character profile template: www.thekatiewolf.com/info

- The last Tuesday of the month is a Q&A episode! Submit your questions for me HERE.


- TikTok: https://www.tiktok.com/@katiewolfwrites

- Instagram: https://www.instagram.com/katiewolfwrites

 

 

AI in Book writing + publishing

Hello, welcome to the podcast. Welcome to what I think is going to be an interesting episode about AI. I am actually re recording this episode. I had an episode fully ready to go a couple of weeks ago, and then some new stuff came out about chat GBT, about AI, about authors. And I felt like I wanted to redo this episode, because part of this episode is me just sharing my thoughts and initial concerns. But then I also polled people on Instagram and said, hey, tell me your thoughts. So I'm also going to be sharing from other people, what their thoughts about are about AI and writing and publishing.

And my feelings have changed since these articles were published since this news came out. And I just yeah, I wanted to give more of an updated take on this. This is not a deep dive into AI. That's not what this podcast is. I mean, you know, if you've listened to my podcast before, they're shorter episodes, we're not covering like every angle of this, this is an extremely complex issue. And I don't pretend to be a subject matter expert on AI, on chat GPT, on tools available in the AI space for writers like none of that, if that's what you're looking for. I'm sure that there are other resources, maybe probably other podcasts that do that. This is just my thoughts and concerns. And then again, sharing some concerns and thoughts and impressions from other writers in just in the writing community.

So I will link a couple of the articles that I read that changed my perspective on this in the description of the episode. One of the ones well, okay, let me tell you what the gist of the old episode was -- me essentially saying, I've been sticking my head in the sand for a while with this whole AI thing because it makes me terrified. It makes me terrified for the future of jobs. It makes me terrified for the future of art, not just books, but other types of art. And I just felt like it was anxiety inducing for me to the point where I didn't want to look at it. I wanted to pretend that it wasn't happening, which was not a realistic thing. It took me awhile to even be able to think about it. And in the old episode, I sort of came down on the side of like, okay, well, I don't see a problem with people using chat GBT or some sort of AI as a jumping off point, in terms of like, coming up with some ideas, helping them clarify, brainstorm that sort of thing. But I do have serious concerns about someone generating an entire book written by AI.

And what's what was concerning to me is okay, is it going to be disclosed to the public that this thing, whatever it is, is written was written by AI. That was sort of my my feeling on it, then. The first article that I read, well, I read a few articles. I'm just linking to one of them in this episode, but it was about how authors are suing open AI, which is the I think the company behind chat GPT. So two authors are suing because they're saying that open AI violated copyright law by using their published books to train chat GPT without their consent. So I'll just read a line from the article.

They argue that chat GPT ability to produce detailed summaries of their works indicates their books were included in datasets used to train the technology. And then it goes on to say, this is the latest example of tension between creatives and generative AI tools, which I think is a nice way of wording it, I think it's maybe more -- I think it's stronger than tension, but so there are multiple lawsuits coming out. Another article that I'll link to is about Sarah Silverman, the comedian who is suing as well, because chat GPT can produce a summary of a copyrighted published book, which is a big deal in copyright law. That's not it okay, because chat GPT has to access that copyrighted work in order to create a summary of it. I, so all of this, learning more about how the tool works, how chat GPT works, I am much more now on the side of holy shit. This does not seem ethical at all.

And it raises serious concerns for people who are using it. Chat GBT open AI scrapes, published works, I saw a really interesting thing on Instagram. And I wish I remember who shared it or exactly how they worded it, I'm sorry, if I remember, I'll update this and give them credit. But they were saying that using the term intelligence, like artificial intelligence is somewhat of a mistake, because it's not actually intelligent. It only is as good as the data that you feed it, it has to have something to work off of, in order to generate anything new. So essentially, there's, it's just scraping all of these published works all of these published web pages. I mean, there was even a news story about how it was scraping fanfiction to make it intelligent to make it be able to, you know, produce something. And so anything if you're using chat GPT, my understanding of this, again, not an expert, but my understanding is that anything you input into chat GPT further helps train it and refine it, to give more accurate answers when it spits something out.

That combined with the fact that it is just scraping published works, again raises serious freaking ethical concerns, for me, about copyright, about ownership, about just anything related to intellectual property, which is a huge, huge deal. For authors needing to have rights over their intellectual property. I just I can't state how important that is. So that's, that's what led my led me to change my opinion on this, I still wrestle with the idea of using chat GPT at all, honestly, like myself, I have used it, I did use it. Maybe two months ago, I got on there. And I wanted to see exactly what it was capable of. So I asked it for a marketing plan for something.

I was like, hey, this is my business. Here's what I'm launching, give me a marketing plan for this. And it spit out some things that were okay. Like, it was fine. I didn't end up doing anything with it. It was more just an exercise for me to see what it was capable of. I have not used chat GPT for writing. So I don't know what the capabilities are there. I'm not planning to, again, it's like, it's such a complicated thing that the issue of like, Okay, where is it okay to use chat GPT? And where is it not? Okay. And I don't know where that line is. Again, I think just using it brings up so many ethical problems that I don't know that I can in good conscience, use something like chat GPT at all, I just don't know, it's like now that I know how the sausage is made I don't know that I feel comfortable using it.

There's also I saw some tech talks about how Google Docs is now well, Google has updated their terms of service for something called Google Labs, I don't know I'm not even going to share specifics of it, because I can't remember them. But something along the lines of like Google changing their terms, so that now Google Docs, anything you put in Google Docs is fair game to be scraped by artificial intelligence to help train these different models so that they can, you know, become better, which, again, I'm just I'm so not comfortable with that. And, yeah, it's like, the more that I, the more that I look into this, the more problematic it becomes. And I think I'm going to go back to something that I mentioned before, which is that I really feel like there needs to be an immediate regulation, something in place, so that it is very clear when a work has been produced by AI. And when it has been produced by a human -- a book that's being published, a piece written online, a video segment, just anything.

I want to know when I think people have a right to know what's created artificially, like art through AI and what's created by human even if it's just something small. I think that needs to be so obvious. And my fear is that the regulation I mean, I live in the US -- I can just talk about the US -- our government. Our government is a shitshow and can't accomplish anything, and they're so behind the times when it comes to technology and regulations already, that I mean, I can't even think of how many months or years it's going to take for them to like address AI. Maybe that's just me being pessimistic, maybe this will get to be a big enough thing that there will be some regulation around it, who knows. But I am not optimistic. I don't have high hopes about that. Which means that, you know, companies and news outlets and all of these places, publishers, they're going to have to take the responsibility of doing that. And I don't know that they're going to be willing to do that.

One of the articles that I linked to is a New Yorker piece that talks about an AI writing tool. And it's more of a writing tool for businesses like a copywriting tool, but the article talks about a number of the startups that have been formed that use AI and there's one in particular that says, hey, give us your like, for companies and it says give us your, your brand guide your voice something, some examples of stuff that you've written, and we our promise is that we will turn around and spit out copy or pieces that are better than what you could do. And there are a hell of a lot faster.

And it mentioned some companies who are already talking to these startups or implementing them already big companies like Victoria's Secret is one in the article. And I just, you know, I don't trust corporations to do the right thing ever, like corporations, companies are motivated by profit. So if they can find a way to have something written in 30 seconds, and that's going to be decent, why would they not do that? Of course, they're going to get to that point where it's going to replace humans, which is just a terrifying thing to think of, I think I, it's going to be interesting to see the impact that this has on writing, creative writing, copywriting, freelance writing, just ghost writing all sorts of writing over the next couple of years. And I'm really weary of the direction that we're heading.

The other thing that I want to mention is just as someone whose job could be very much impacted by AI in the next 10 years, and maybe even five years, 10 years may be too long. There are not only editing tools for a like AI editing tools, there are also now AI, writing tools, and publishing companies. I talked about this on Instagram a while ago, where I saw an ad for an AI book writing and publishing company. And I clicked through because I was like, what, how does this work? Exactly. And the website, there were like two ways that it worked. You could use their AI tool, which was not chat GPT, I don't think my understanding was it was something proprietary, where you could use this tool to put a couple of ideas, and then this tool would spit out a book for you essentially, and then it would use AI to publish it somehow. With which, okay, I'm not even gonna get into that.

But what was so concerning about this particular company, and I wish I had screenshotted it I'm sure there's so many of these popping up, but I can't remember which one company it was or what website or anything. Y'all it looked so scammy there were typos in the copy. They did that weird initial caps thing. It -- just the testimonials seem to fake like, it looks so scammy they could be legit. I don't know, I just got a really bad vibe. There were so many red flags on their website. And I don't want anyone to get scammed. I don't want some writer who's new, who's like, you know, this is my ticket to success as an author. Or this is my ticket to financial success because I can be churning out books and publishing them like crazy because they're written by AI. I don't want someone to get taken advantage of.

There are people who are just in it for the money and want to take advantage of writers looking to make you know, a living as an author or make some money as an author. And that scares me. The fact that these these kinds of companies are going to be popping up and you know, are they legit? I don't know. Again, not to be super pessimistic. Maybe these companies are fine, but this particular one had so many red flags. So if you're listening to this and this is something that you are seriously considering. Please, please vet any service that you use, please. Also with self editing tools, I will say I've looked into I've talked about this on another podcast episode where I looked into Grammarly and I used it for some thing I put a manuscript into it and it caught some things. I think Grammarly is better than nothing. But it's nowhere near ready to replace a human. It miss things, it flagged things as errors that were not errors.

So, again, I think it's a great, it can be a great option, something like Grammarly for someone who can't afford an editor, because let's be honest, editors are expensive. And it's definitely again, it's definitely better than nothing. But it's not to the point where it can replace a human editor like not even close. So those are kind of my complicated feelings about AI. And I can't put my head in the sand anymore, like this is happening. It's here. This is just how I'm feeling today. Obviously, this could change. Hopefully, we won't have to rerecord this episode again, because more stuff comes out in the next week or so the next couple of weeks. But yeah, this is very much an ongoing thing. I'm still sorting out my feelings about it. But I at least wanted to put this out there. And now I want to read some comments from people on Instagram. I'm not going to read every message or DM but just some of them.

So I did get a DM from someone that I thought was interesting and wanted to read. Okay, so I posted about this on Instagram said, hey, I'm gonna do an episode send me your thoughts about AI. And someone said, AI isn't nearly able to do what people are selling it to do. I'm not worried you shouldn't be either. My day job is that I am an IT consultant, I have gotten a good look at AI under the hood. And it's nothing. It's a search engine that returns results in sentence format instead of list format. I think writing is the literal definition of the human touch. And that is something AI is nowhere near being able to fake. I think a lot of people would feel better about it if they knew how basic it really is.

Chat GPT can form sentences because someone programmed it to do that. And it isn't horrible in the way that it is if it is a sophisticated engine of programming powering it and actually works. But what people are afraid of is when computers start to learn beyond what we programmed them to know. And this point is called the singularity and we're not close to that. Okay, so that was that person's reaction or feelings. And then I have a few other shorter ones. I put like a little comment box in my stories where people could fill in their thoughts. Okay, someone said it trivializes the hard work people are putting into writing and honing their manuscripts.

Someone else said doesn't AI essentially learn and learn is put in quotes. By studying media that already exists, then copies it? That's not okay. What else do we have? I feel like it could go one of two ways. One way would be it puts many writers out of business in writing is changed. Or the other ways people call it a hoax and it becomes all jumbled. Someone else said we as consumers need to make sure we are prioritizing books written by individuals. AI should exist only in science, art and literature. And a AI should be separated completely. A few people talked about self editing tools. I got a comment and said AI tools to help with self editing was great but AI writing stories I'm not okay with. While it does a great editing tools, it does not replace a human eye. Someone said no people write to try to explain their human experience AI takes that away. And then a few people talked a little bit more about plagiarism saying it's a form of plagiarism AI learns from other writing, but no one has paid or no one has allowed it. Do not submit your original work to open source AI you are handing over your copyright. So that's just a sample of of thoughts and feelings from people.

I again, I don't I think this is one of those things that I'm just going to have to be vigilant about and kind of monitor and I think that we as a writing community are going to have to do the same thing. Because I mean, no one was talking about chat GBT or this kind of a like AI on this scale months ago. Like it's only like a matter of months that all this stuff has come out. And of course, I'm sure it was happening behind the scenes a lot longer than that, obviously, but in terms of like the collective being focused on this issue, and all of the articles and all the attention and discussion and debate and lawsuits now. This is all happening so quickly. It's all recently that this has come out.

So I have no doubt that this is something that's going to continue to kind of evolve and grow as we navigate this and I think we just have to be number one supporting authors who are you know, who have these lawsuits against Chad GBT against open AI, we have to just, you know, make sure that we're not falling prey as a community to any sort of scammy AI writing or publishing tools that are promising, like, get rich quick. And this is your ticket to success kind of things, and then don't have this stuff to back the technology or ability to really back it up. You know, my concern with an AI written book, at least right now, where we are is sure it could turn out a AI written book, but is the quality there? I don't know.

And then I think we also have to be I don't know if other people would agree, but at least for me, like I really want to be pushing for people to disclose for this to be required that if a AI wrote something that that be disclosed, I think that's going to be important. So that's the -- I think I'll just end it there again, at some days, I get really panicked and doom and gloom about AI and the effect that it will have on writing and on jobs and industries. I mean, let's be real. And then I have this counter reaction where I think, okay, am I just being like, kind of a? I don't know. Am I resisting the flow of technology, and resisting the flow of the inevitable and being grumpy and grouchy about it? But I don't think that's the case here. I don't think that's what's going on.

Yeah, maybe I'll do another episode and a few months or something, because I'm sure more is going to come out in the next few months. I'm sure that that there'll be updates and everything. But for now, again, today, this is how I feel. And this is how y'all feel. So yeah, I don't know, man. It's wild. Sometimes I'm like, are we really having these conversations? I feel like I'm living in the, in the future. Like I never thought that I would be having these conversations about AI this point in my life. Like maybe I thought maybe artificial intelligence would be more of a thing when I was like an old woman, but I didn't expect to be having them at this point. But here we are. Alright, thank you for listening to this kind of complicated mess of of feelings and reactions and everything. And I'm always happy to hear what people think about this so you can always reach out to me and share your thoughts.

Katie Wolf