Writing a Book with AI
Welcome to wake up with AI, the podcast where human powered meets AI assisted. Join your hosts, Chris Carillon, Niko Lofakas, and George b Thomas as we dive deep into the world of artificial intelligence. From the latest AI news to cutting edge tools and skill sets, we are here to help business owners, marketers, and everyday individuals unlock their full potential with the power of AI. Let's get started.
Chris Carolan:Good morning. Good afternoon. Happy Monday, December 16, 2020 4, halfway through the last month of 2024. It is time to build some specific skills related to AI today. We're here with Mike Tedasko, writing books and and publishing books in record speed.
Chris Carolan:How are you doing, Mike?
Mike T:Good, Chris. I'm looking forward to teaching somebody, teaching folks about all all all the good and the plenty of bad that comes with AI writing.
Chris Carolan:Excellent. We're we're excited to have you. At the very least, you'll be teaching us, about this publishing part. We talk about our own books all the time, and we hope that there's a takeaway today. I know there will be for those watching that, you know, half the time when we talk about AI tools, it's the difference between, you know, creative works or even professional works, existing or not existing.
Chris Carolan:A lot of times, we we try to redirect the conversation away from replacement and more about just enablement.
Mike T:Look, and I I concur. I I mean I mean, these are tools to help us be more effective. You know, for those of us who don't have access to a cowriter or an editor or a proofreader or whatever it might be, this is a way to get to that. And it's do you want? It ain't good as good as a human.
Mike T:It could be a lot faster, but it's a lot cheaper. And and again, not everyone has the luxury. I'm working on a book project where I actually do have the luxury of having an editor. And you know what? It's awesome.
Mike T:But I work on a lot of other things, like my blog posts. I don't have somebody edit that. I have AI help me with that. We could talk about all that stuff today.
Chris Carolan:Yeah. So we'll we'll spend a little bit of time just talking about writing process in general, and then we'll we'll get into some live demonstrations, which we definitely love doing that on here. But, George, as we, as we started talking about having Mike on, what was the, you know, the first question about using AI for writing that that kinda popped into your head?
George B. Thomas:Yeah. Well, the first one and the one that I now have are 2 totally different things, but I will go with the one that I had, first, and that is how do we get it to write in a more human way? And when I say that understanding that I'm literally asking it to write as Mike t or George b or Chris c. Right? Like, how do we how do we hone that into existence?
Mike T:So here's what I would say. Like, Claude are are you all familiar with Claude? For me as a writer, that's probably my favorite tool that I use by far. There is a new feature that they actually have in Claude where it will actually learn your writing style and be able to more closely mimic your own writing. And this is similar, like I'm sure everybody's done the thing of, like, give me a Shakespearean sonnet about Darth Vader when he has a cold or whatever.
Mike T:And it'll do something amazing like that. Or then you could say, like, okay. Now write it like Ernest Hemingway. Now write it like this. Like like, there is some level of mimicry that it can do.
Mike T:I will say in playing with Claude, which is probably the best model for at least short form writing and capturing your voice, even then, it doesn't do it exactly. And and here's why and here's what's going on. I am no expert in Shakespeare. So if I ask for a Darth Vader catching a cold Shakespearean sonnets, I read this thing, and you know what? I'm like, well, that seems pretty good.
Mike T:Because you know what? I don't have expertise in that area. And while it might get, like, 70% of it, in my naive mind, it's getting a 100%. I'm sure if you get somebody who works at The Globe up in, London or whatever the hell his, theater place was, who's, you know, studied this their whole life, they will look at that and say, like, what is this? This this isn't right.
Mike T:This isn't close. And therefore, it's the same kind of thing. You've been writing as you your whole life. And so for it to actually get your tone and all that right, it's really damn hard for you. Like, you could give it to a stranger, but like, oh, yeah.
Mike T:That sounds like you maybe or whatnot. But, like, to really get yourself right, it's not going to quite get there. It can do corporate speak generally well, but, like, these are all generic, middle of the road type stuff. If you're really trying to refine it, it's just super hard, and it's it is absolutely not there yet in doing that.
Chris Carolan:Something interesting, I think, is, like, you've been writing as you for a long time. That is definitely different depending on who you're talking to. And, like, one of the the magical things about AI, in my opinion, is is maybe people who have not been writers or or even done that much writing outside of maybe some business emails or things like that. Right? I'd love for you to dig into the process there.
Chris Carolan:Like, how do you like, if you've always had something, like, it'd be awesome to write a write a book or something. Like, what is even the process that you'd recommend to to get started there?
Mike T:And we're talking about, like, non AI world. Just just like how how do you do this in general?
Chris Carolan:No. I'd say, like With AI? I I try to live in the AI world now because I think that's our in the AI world. Yeah.
Mike T:We're in the we're in the AI world already. Why not live in it? It's in this question too. Look. Here's the first thing I would say to you is the AI is not gonna do the work for you.
Mike T:Like, that that is it. Like, writing a book you know, I think Tim Ferris has a great quote about, like, when you should write a book. You you should only write a book, and I'm gonna paraphrase here, when it is more painful for you not to write a book than it is for you to write it. Like, you need a like, writing a book is hard. It takes a long time.
Mike T:And again, like, the hardest part is the rewriting. So I'm gonna quote Hemingway now. So Hemingway's famous quote about about writing, one of his famous quotes is, all first drafts are shit. And this came from Hemingway. And, like, there's a lot of truth in that.
Mike T:Like, whenever he writes, go, like, find the first draft of Star Wars online. Like, you know, and, like, you will read it, like, what is this guy? Like like, it's like, woah. It's all over the place. Well, like, Lucas rewrote that thing 20, 30, 40 times.
Mike T:Got other opinions, other insights, all this. Like, that is what actually happens in writing. So the first thing is, like, make sure you wanna do this. If you do and you're like, yeah. Yeah, Mike.
Mike T:I'm up for it. It's like, I I'm I'm willing to do the work and all that. Look. What I would say is you just take this thing one step at a time. There are people who will outline everything upfront, and they'll have, you know, let's say it's a fiction piece, they will they will know the twist and all that.
Mike T:They'll just work on an outline for weeks or months or whatnot, and then do that. There are other people who are just kind of like, okay. I'm just gonna start the story, and I'm gonna see where the story takes me. And then there's nonfiction, all of these other things in between. The one thing I would say, that doesn't change.
Mike T:That's how people have been writing books for 100 of years now. The way that AI makes things different is it could actually help do many different tasks for you. It can help you project manage this thing to keep you on task. So, you know, you say, hey. I wanna craft a 200 word book by the end of 2025.
Mike T:Realistically, what's the schedule for that? Like, what's that going to actually look like? Take me back step by step, month by month, and eventually, you can even even ask what AI to do week by week. It's gonna make a project plan for you. It could you can then just start following that project plan.
Mike T:And, like, one of the things that AI actually does do pretty well is giving you feedback on this. You have to prompt it in the right way. But all the time, I write a lot of blog posts about AI stuff. I will write the whole first thing myself, but then I will feed it into Claude. I will feed it into chat gpt, and I will say, give me feedback on this.
Mike T:I will have it fact check me. I will have it do all of those things. I would rather have a human editor do that, but I don't have the money or the time to have do this on my my free newsletter. So, like, AI is wonderful for that. So, like, those are a couple ways, and I could get into way more detail.
Mike T:But, like, throughout the process that you know, like, again, once you commit to doing that, the AI is going to help you throughout. You're still gonna have to do the hard work of writing yourself.
George B. Thomas:So what I hear is feedback on what you've written, and I hear maybe research to make sure that the stuff that you're saying, potentially lines up somewhere. I do actually wanna step out of AI for a second because I'd like to know what the building blocks are when we say, okay. I've got this pain point. I need to give birth to this book. What are the steps to get to the point to where you even feel like as a human, now I've got this idea, I've got this structure, I'll start to write.
George B. Thomas:Like, give us those, like, steps or that flow so I can kind of think about that in an AI assisted world.
Mike T:Yeah. I mean, look. I so all these things start with ideas. I've I have an Evernote file that I've been keeping for, I don't know, 15 years, and I probably have a 100, 200 book ideas in there that I've just been keeping. And I'll sometimes look back over them and frankly just see what inspires me.
Mike T:Because again, this is a time commitment that you're doing. What I would say is, like, everybody out there, no matter what, you should have idealists. Like, ideas are such a precious commodity, especially when they're your own ideas. You should be just be keeping them. It could be in a notepad file, wherever.
Mike T:Keep your ideas for businesses, for books, for whatever it is because you never know when you're gonna be inspired for whatever it might be. Say you have that. Say you start on the process. Again, I I there's no right or wrong way to do it. But what I would say is you just gotta know you're gonna have to just do it one part at a time.
Mike T:And you are going to hit a wall probably in the 1st week, where you're like, what the heck am I doing? Like like, you know, I think for most people, it's better to set a schedule. So say you already work full time. You're gonna be writing this book. You got kids, you got all this other stuff.
Mike T:You know what? You might the schedule might actually have to be, you get up at 5 instead of 6 every morning, and you you get 45 minutes of writing in in there. Just, you know, to do it ahead of time, to just do it, and you just try and churn out a thousand words. The one thing I would say is do not worry about quality. There there's a concept Safi Bahcall, a great writer.
Mike T:He's only written, I believe, one book. It was called my god. It's totally escaping me. I actually interviewed him back when I was at PayPal for his book. It will come to me in a in a second because this is something I do recommend.
Mike T:He in the book, he actually talks about the process that he went through a lot. And, you know, there is the kind of the writer's brain and the editor's brain. Don't let those two things like, when you're writing, just write. It doesn't matter how bad it is. It doesn't matter how crappy it is.
Mike T:You will edit this stuff later. You just gotta just barf all those ideas that you have in your brain down on a piece of paper, on a computer, and you just start churning through it again and again and again and just go through the reps. And again, after doing this for about a week, you're gonna hit a point at some point where you're like, oh, god. This is painful. This is awful.
Mike T:I don't know if I wanna continue doing it. And frankly, that is the point where you do then need to decide, okay. Maybe I just take a couple days off. Maybe I can push through it. Maybe whatever.
Mike T:There are so many 5000, 10000 out of a 100000 word manuscripts that exist out there where people started, they were ambitious, and it didn't work. And you know what? Sometimes that's fine. It's okay to just say you're gonna abandon things sometime. And, like, you need to make that call because maybe you know what?
Mike T:As you're getting into it, maybe it wasn't as good of an idea. Writing forces you to think. That's why I do it. It forces you to get no matter how much sense it makes in your brain, when you actually start to get it down on paper, you're like, oh, maybe this isn't as good. What I would say is just always with this stuff, never just abandon it right away, but just say, okay.
Mike T:I'm gonna pause for a week. I'm gonna come back in a week. I'm gonna reread this. And then if I still have the passion I did before, then I'm gonna jump right back into it. And that's the one piece of recommendation I give folks.
Nico Lafakis:I was gonna say you had mentioned that you get to a point where you sort of, like, hit a wall after you begin the the process. Is there do you notice that being a general thing? Is it, like, okay. You know, generally, after I get my outline done, that was so painful. Or is it, like, you know, you've gotten a few chapters done and and now you're just trying to, like, push your way power your way through the rest of it.
Mike T:And so I don't know how you got super soft there, Nico.
Nico Lafakis:Oh, I'm so sorry.
Mike T:My my apologies for that. I think you were saying, like, just how do you push forward, like, with all the other chapters and things like that? Am I getting it right, or did I miss something there?
Nico Lafakis:So I think you were saying that you get to a point at which you sort of hit a wall when you like, right out of the starting gate. And I was wondering, is that a definable moment? And if so, what what is one of your ways or what are a couple of your ways of getting around that? Because my my thoughts towards this and and hitting walls is, okay, but the tool is also sort of your salvation and that you can use it for ideation to sort of help you bridge that gap or get through that.
Mike T:Okay. I I totally got that. Thank you. And and I think the levels are better now. It just wasn't coming up to me.
Mike T:So, Nico, your question is just about, like, hitting a wall. And and this was something I was going through very recently. I'd I I'm so actually, long story short, I'm actually getting an MFA. I'm a weird tech guy who's actually getting an MFA in writing, if I didn't explain that in the beginning already. And one of the last classes I'm taking in my degree is actually a humor writing class.
Mike T:Super fun, exciting, like, really, really enjoyable to do that. And so this is all short form stuff, like, 5 to 700 word pieces. I I've been writing over the last few months and, you know, and getting published now. One of the things that I found out with, like, hitting that wall, and this is in a much more condensed time frame than the books that we're talking about, but I think it's still applicable. What I'll generally do is, like, I have, you know, hundreds of headlines for what might be funny articles, and I'll find something that, like, piques my interest, and then I will just start writing.
Mike T:And I I call this the collective soul problem because I'm I because whenever I can, I reference 19 nineties grunge metal bands and and stuff like that whenever I can? So some folks who are younger than me, there's this band in the nineties that was popular for a couple years called Collective Soul. It was a very poppy rock band. One of the things that they basically did as a band, they would start playing a song. They would all bring in ideas.
Mike T:They would literally start playing it. And after 15 minutes, it wasn't clicking. They were done with it. Like, the like, they they were just done. They were just like, okay.
Mike T:We're never playing that. Like, that is not going to be one of our actual songs. And with at least short form writing, I thought like, with the humor stuff, that works really damn well. So I basically just give myself, like, 15 minutes, and I just write for those 15 minutes. At the end of the 15 minutes, if I'm like, oh, god.
Mike T:This is not funny at all. I don't know where this is going or whatever. Then I'm like, okay. This is just going in the trash bin. I I was just moving that to the side.
Mike T:If there's still something to them, like, oh, I've actually kind of cracked myself up here or whatever it might be. Like, it's the same kind of thing. So that's the short form version of your question, Niko. I think the longer form version of that is probably about a week or so seems right. Like, you know, 5 days spending an hour a day, so 5 hours or so.
Mike T:It's probably about the right time where you could at least give yourself a decent sense of, like, okay. I'm gonna I think because if you only spend an hour or something like that, you didn't really push through it enough. So so at least in my somewhat uninformed opinion, I would say about 5 hours is that equivalent amount of time for a full fledged book that I would then take and just use that to kind of reevaluate.
Nico Lafakis:Like, where do you see things, you know, like printing press obviously changed the way in which everybody is able to consume knowledge. Right? And so now we have this tool where, rather than consumption, it's more production. Right? So, like, everybody is able to produce now.
Nico Lafakis:If they had a a like you say, like a book idea, it's been rattling around in their brain for years, you know, all of a sudden, at least from my view, I feel like next year we're just just going to see, like, millions and millions of books just, you know, audiobooks and all sorts of stuff just thrown out there. Do you feel that that's going to be the case, or do you think that it's more of a helpful tool for people who are already into writing?
Mike T:Look, I would bring the analogy to, like, photography, to be completely honest. You know, go let's go back a 140 some odd years. You got, like, Matthew Grady in, like, New York, basically, taking photographs of, like, the president and, like, you know, there's, like, 10 people in the country who know how to use this new and emerging technology. Like, there's almost nothing there. And things evolve and becomes cheaper and cheaper and and more effective.
Mike T:And and now we're at a point where, yes, everybody in the world has a camera on them at all times. And and actually, I don't know the numbers. Are there more artists doing photography than than before? Like, you know, has that world diminished and so forth? I don't know.
Mike T:I mean, I don't know if it's taken away from that. I'm there are plenty of gallery pieces focused on photography and different art styles and and all that. And it doesn't diminish it because you have 4,000,000,000 people around the globe who have a camera in their pockets and on their person at all times. And I think that's what we're going to see with, like, writing. Look.
Mike T:Again, taken far enough back, you know, most people were not literate to be able to write. You know, most people did not have books then, you know, as you talked about the printing press and all that. And then, like, okay, maybe the town just has a Bible. Now it has a few other books and so forth, and then you're getting into pamphlets and all this stuff to to today, where I'm sure the amount of content that is published on the Internet in a given day would probably be more than you could possibly read in your lifetime. And, you know, on Amazon, you I mean, there are there are millions and millions of books.
Mike T:You would not be it's not possible for any human to read all of that. So I think curation does come into play here, and I think that is more of what you're going to be able to see. The the tools are there for anybody to do this. Just like, you know, again, to use the same device, like, you can create movies. There there are directors, Hollywood directors, that use iPhones to make movies.
Mike T:I don't see as many people really running out there and doing that because you want it's still work. It's still vision. It's still all of these things to do that. AI still sucks at making something from scratch. We can get into that.
Mike T:It is just not good. For folks like watching this since 2022, I have been producing AI written books under the pen name Alex Irons and publishing them on Amazon. It's actually super easy to publish stuff on Amazon. And this is what Nico is getting to with this. So I have 14 books on Amazon today under this pen name, and I will tell you honestly that none of them are good.
Mike T:They're not really worth my time in reading them. And And I'm doing my best. I'm not trying to make a bad book here, but I'm also just trying to let the AI go. The AI can help you. It doesn't go from 0 to 1 and create something of art that you're like, wow.
Mike T:This is this is worth 5 hours of my time or 10 hours of my time to sit down and actually read and spend time in this world that this machine created. We are not near that yet.
Chris Carolan:Well, I'll let you bring in the the screen share whenever you feel is best. But I think it's interesting what you say there. If you've been doing it since 2022, you know, without trying to control too much, like, it would be super interesting. And maybe as you talk us through this, like, what have you seen, you know, get better, like, in that time? Right?
Mike T:Absolutely. Absolutely. And and and I will say here. I'm I'm gonna go ahead and start to share right now, share a few different things with y'all. First, actually, I'll I'll just show this.
Mike T:And and, actually, I I do wanna make really clear. So this up here, this is Alex Irons. Kinda looks like me. Not not quite look like me, but, you know, close enough and all that. It's an AI version of me.
Mike T:I'm very clear that this is AI work, and I'm very clear it sucks. And I take 0 money from this. So, actually, all the money I take goes to charity. So we're gonna actually create a book here, and these are all different things. The mediocre Gatsby instead of the great Gatsby was one that I had on there.
Mike T:So often, I will have, like, topics that I'll bring in there, but then let AI run with it, and we could see this. This was the only one, the autobiography of George Santos. That was one where I kinda co wrote with AI. It's fair, but, you know, still still not something that I would really, you know, spend any any time with. But that's that world.
Mike T:So what I'm gonna do is I'm going to get into chat gbt now, and we're gonna share this with you all. So anybody who's spending the $20 a month for chat gbt is gonna be able to see this. And George, Chris, Niko, I'm gonna be just be prompting you all here for ideas and concepts here. So for anybody watching, we're just gonna simulcast. This is literally how I would do it.
Mike T:So I'm using the Chat gpt01 model, which I've never actually used before. Well, never used for writing a book. I've used it for many other things. I haven't actually done a full, quote unquote, book as part of this. And and here's what my prompt is going to be.
Mike T:So apologies as I'm typing and writing out loud. And then in a second, I'm gonna ask you gentlemen for thoughts and feedback. So
Chris Carolan:And as he puts this prompt together, if people are looking to explore these tools more, like, don't discount the fact that we're using o one for the first time as opposed to 4 o or, you know, Mike mentioned Claude earlier. Half of the interesting battle slash challenge is understanding which models do what and how and when to use them. And, you know, this is where we we talk about on the on the show all the time. Like, George is able to create lots of great content because he's done the work to use different tools as, like, you know, a composer of sorts. Right?
Chris Carolan:Understanding all these different instruments to put this cool stuff together. It's not just trying to ram everything you want through one model because that's the only one you're you're used to using.
Mike T:That's right. Yeah. In in all the books I've that Alex Irons has written, I've used different models. Like like, I've used BARD. I've used like like, every time there like, it was almost always an excuse for me to use a new model and understand it by by doing this.
Mike T:So you will see, like, what are the different you know, the good and the bad of it. Let's get this thing running, though. So so this is a super simple prompt. Super simple. And and, again, intentionally, I'm not just saying write a book, step 1, because actually that ain't gonna work.
Mike T:Like, even with the most advanced model that OpenAI has, it's not gonna be able to just one shot write a book for you. So I'm actually starting with an outline as many authors would actually do. But, like, let's let's let's give it a topic here. What what do you guys want this book to be about? It could be about truly anything.
George B. Thomas:Let's make it about street racing.
Mike T:Street racing.
George B. Thomas:Yeah. We're gonna do a Dominic Toretto spin off.
Chris Carolan:Oh, wow. Here we go.
Mike T:So instead of the so instead
George B. Thomas:of the Fast and Furious, we're gonna do the slow and something, but we'll have AI help us with that. So inner city street racing.
Mike T:Street racing in nursing homes or something? Or
George B. Thomas:I mean, it could be, like, when when Dominic Treddo is 90 years old. So, like Yeah.
Nico Lafakis:Yeah. Let's do it.
Chris Carolan:Retired street racers. AARP Street Racing.
George B. Thomas:I mean, we're gonna have fun if we're gonna do this.
Mike T:Absolutely. Yeah. And and so the one thing I do wanna point and so I'm just gonna hit I'm gonna hit return here, and we're just gonna see what it starts to think and come up with. So it's gonna it's gonna think for a while, and we'll see this whole pros progress as it's doing that. The the one thing I do wanna pull out is, like, already though, we are using some human creativity, on top of this.
Mike T:So we're already cheating the model a little bit by saying, let's give it actually something that, hey. I would watch that movie, a bunch of, like, 8 80 year old street like, that would be pretty damn awesome. So, I think there's something here. So the Rust and Rubber, the the revolution of senior street racers by Alex Irons. And then it's giving me a chapter outline of all of this.
Mike T:And so you could see it is given 13 chapters, blah blah blah, and an end note and an outline. So it has done a pretty decent job of this. Let's just
George B. Thomas:Now, Mike, when you're when you're doing this, do you usually circle back around and ask it, like, hey. Can you give me 5 more title ideas so that you can pick one that really, like, resonates with you?
Mike T:Look. If I was doing this for myself, I would say give me 50. Yes. Yes. So to answer your question, yes.
Mike T:Typically, with this, I just let the AI choose whatever it deems to be best with that. I do give it some help for this, but but yeah. Like, if I were actually using this for my own book, absolutely. All the time, I will when I'm writing in general, I will often put up little brackets around things that either I don't like in my writing or, I'm not sure if that's factually correct, or I have no idea what to say here. And then what one of the things I will often do is run it through Claude, run it through ChatGPT, and say, hey, give me some feedback.
Mike T:Give me 5 options for the 4 places in this piece where you see brackets. And then, you know, I will do that all the time, but I'm constantly asking for multiple options and then letting you choose the best one. That is the best way to do it. But anybody who's seeing this, you could see you got Walter Wheels Mendoza, street racer in the seventies, blah blah blah. Great.
Mike T:Great. I'm gonna do one more thing here, and then we're actually gonna have it start writing. So I'm gonna, like, I'm gonna first say because this is something that can really skew. I'm gonna say define the writing style and method to create the piece and give me a list of the main characters with story arcs for each. And real simple prompt there, but what this is doing and, again, this is all stuff a writer would typically do if they're writing a fiction piece.
Mike T:The writing style, maybe this is a little bit they know their writing style potentially, but this is forcing the AI to kinda define that and then actually giving some real character arcs because, otherwise, it tends to just wanna go from point a to point b and so forth.
George B. Thomas:And, Mike, for those of us in the room that have no idea, you just typed in writing style and method like it was like, boom. Of course, you would do this.
Mike T:Sorry. Yes.
George B. Thomas:What method? Like, what is method? And why is writing style and method, and how can we think of those as a rubrics that we might pick and choose for if we can envision the outcome we want, this is a very foundational beginning piece that, like, helps direct in that way. Explain writing style and method a little bit deeper, please.
Mike T:Yeah. And, honestly, method wasn't a special word. I I just kinda threw it out there, to be honest. But but but, like, what we're trying to do is it doesn't know should it write, like, should it write, like, the the news page of the Associated Press? Is that the style that we're looking for?
Mike T:Is it looking for, like, Aaron Sorkin style writing, like West Wing and Social Network, that kind of stuff? Or what is it? And so it's basically just trying to define how we want the AI to actually write this out. So, yeah, so Unity actually need to use the method, but writing style is what I'm actually trying to define here. And so you could see, so it's going to and it said we're going to do close third person.
Mike T:It's going to primarily follow one character. It will have flashbacks. It's focusing on setting, all this other stuff. It just decided all this on its own. And again, in the meth in the in the process we're going through here, it's mostly pushing it off on the AI to do it itself.
Mike T:And then the character arts is just so it has some logical way of doing that. So then what I'm going to then say is given everything, I could press spell everything you have laid out already. Start writing this book. And so I'm gonna actually say specifically, take your time, quality is greater than quantity. Focus on making it a memorable experience for the readers.
Mike T:Do the first act now and adjust the outline if necessary. So all of this that I had written here was basically to tell it, write this book, take it slow, make it interesting, do what you said, but also don't fear from straying from the outline if it's better to do it later. You know, a better way to do this would be to actually say, okay. Given what you just said, now redo the outline. Okay.
Mike T:Now redo the characters. It is the whole process of writing and rewriting that humans go through. The more you do of that, the better it's actually going to be. But we're just gonna kinda do this in sort of one shots.
George B. Thomas:When you hit that arrow, I do have a question because words matter. And I would have thought you would use the word chapter, but you used the word act. Is it matter? Talk to me about act versus chapter when you're doing this prompting.
Mike T:Yeah. So typical story is gonna be 3 acts. But it's just like how we if if anybody has ever done Joseph Conrad and, you know, a man with a 1000 faces and all this. Like, the typical western fiction story is going to be something that is done in 3 acts. So even though they did 12 chapters, a better way would have and actually to say, hey.
Mike T:Let's do it chapter by chapter. I'm just moving along a little bit faster here, instead of doing the prompt 12 times because it I think it gave us 12 chapters initially. I'm actually saying, okay. Take the first 3 or 4 chapters, whatever you define as the first act, and go with that. So, yes, the more times you do this, the more the more outputs you give it, the the better the quality is going to be.
Mike T:So a long time since the finish line, it started to write chapter 1. It gets you into chapter 2 here, and I think oh, yeah. It's still writing, and it's into chapter 4 now. It's still going, and it's going.
Chris Carolan:If you wouldn't have maybe done some of those foundational things like writing guidelines and voice and tone, what are we accustomed to seeing, or what are you accustomed? When you say AI just wants to go from a to to b, what do you see happening when you don't set some of those foundational aspects?
Mike T:AI is lazy. Like, that's the honest truth. AI is lazy just like they they learn from us. And so often, what AI will do is it will just try to go to the the path of least resistance. It won't always look back and say like, oh, let's plant a seed here in chapter 1 that we're going to revisit in chapter 11.
Mike T:These are the things that, like and, you know, and better pieces of written work we will often do, and we will often expect that as readers. The AI itself, though, it's just going to move next piece one piece to the next very matter of fact. And frankly, it's gonna get less creative as it goes through. Like, I have found, like, it can add like, it could start off actually, oh, this seems like an interesting story. And but by the 3rd chapter, it's like, oh, god.
Mike T:This is just it is just like rubber stamping. It's just fine finding a formula and, like, the most boring average formula you could find. There's no surprise, no nuance.
Chris Carolan:Check-in the box.
George B. Thomas:So, Mike, have you found ways that you battle against that?
Mike T:Yeah. So so let's let's do one more prompt here. Is it is it done right? Is it done? Yes.
Mike T:So how many chapters did it did it do here? Chapter it went up to chapter 5. I'm kinda surprised by that. Good. But this is getting to your point, George.
Mike T:I want even more excitement, intrigue, and elderly humor.
George B. Thomas:So, obviously, this is because we're writing a fictional book. But, Mike, do you think this would work too if if this was, like, words like education for the reader, like, go from 101 to 102, like, things like that? You know what I'm asking there?
Mike T:Yeah. I I mean, I I think the more colorful you are in your language, the more colorful the AI is going to be in turn. I think you are going to be able to see that there is again, I I do not like to use author's actual names. Like but but I I've just been reading the latest or listening to the latest Malcolm Gladwell book. But, like, think about his style.
Mike T:You know? I mean, it's very it's, you know, very vivid descriptions, a lot of detail in it. You know? I mean, he's a he was working for the New Yorker. He has, like, just these vivid stories that he tells.
Mike T:And so, like, to say, hey. I want you to explain how deodorant works in this chapter of this book. And I want you to give it just a vivid story about the origins of the creation of deodorant and what is the most amusing factual tale. Now the more that you push it in nonfiction, I will say, the more likely it is going to just completely make stuff up. So, like, like, that is the one so I have not done any nonfiction books this way, in part because I don't wanna, like, just put junk in the world that is completely made up that I made up that I don't wanna fact check myself.
Mike T:Fiction, you don't have to worry about that. But in nonfiction, yeah, you gotta watch out for that much more. But it's not to say you can't do it there. So for folks watching this, again, what I'm gonna do is I will do this. I'm gonna copy and paste, and I think within 24 hours, this will be live on Amazon.
Mike T:We will do that. But I actually wanna show you all how to make the cover. Is that worthwhile to do that? Like, we could do that real quick.
George B. Thomas:Yeah. Absolutely. While you're doing the cover, I do wanna hopefully, we have time to swing back around too. Like, I I've heard, and I could be totally wrong, that Amazon is getting to the point where they're, like, blocking AI content. Is there a way that you get around it being flagged for AI content versus human content?
Mike T:So it's funny. When I started doing this, they didn't care at all. But, like, you know, like, since then, for anything that you actually upload into Amazon, you do have to say there's AI content or not. And so I do and I say it's a 100% AI written. I guarantee that downgrades it in the algorithm.
Mike T:So Grammarly has a free AI detection software that seems decent out there that people can use. I've used it on content I know is AI written and I know is hay is human written, and that seems to be directionally correct. I'm sure Amazon is using something like that on their side to actually detect it. So if you are doing an AI book and it's not actually and you're saying it's not AI written, I think they will detect it and they'll probably punish you more is the probably the honest answer to that. So here, let's let's do a quick book cover for you all.
Mike T:So let let me just share it one more time, and I'm gonna use 2 tools here. I'm going to use Grok's image generation tool, and I will actually use the image generation tool called Midjourney, which is located in Discord. So let me just share again. Can I share more than one tab of my browser here? Let's see.
Chris Carolan:And what what I love doing in cases like this, when you already have, like, a bank of content to work from, even if that place can't make the image for you, like Claude can't make images, I'm usually asking it to create the prompts about the thing that we've been talking about so that I can go make an image with it, you know, very quickly in the other in the other tools.
Mike T:Absolutely. Yes. And so I could ask chat gbt right now, and I could actually say, hey. Give me something that would blah blah blah blah. I could absolutely do that and probably give me a better prompt.
Mike T:But I'm just gonna use this right now just so folks could see it. So in the bottom there, you could see what I'm writing, hopefully. So I'm saying book cover for rust and rubber. And I shortened it because it's not gonna be able to get all the words right if I give it too long of a title, and we'll get to that in just a second. Rust and rubber, a book about elderly street racing.
Mike T:Okay. I'm just gonna make it real simple. Oh, and actually, I have one thing I'm gonna do. You could do this with mid journey. I'm gonna change the aspect ratio.
Mike T:So 9 to 16, I think is the standard aspect ratio for for books usually, and let's just let that run. And I'm gonna put the same prompt into Grok. Oh, no. No. No.
Mike T:I didn't copy it. Hold on a second. I can't do the aspect ratio in there, but there's other match that we can do. Let's see what it does. Okay.
Mike T:So we got both of these things running. We'll see which one comes through first. Okay. So when we blow this up, we can see this. No.
Mike T:No. No. Wait. We lost it. We lost it.
Mike T:Hold on. Open in browser. Oh, why is it not doing it? Okay. I don't know why it's not doing it, but can can folks see it?
Mike T:It's a little bit small. Apologies for that. But you can see the thumbnails here. So I don't know. Fair book covers.
Mike T:It did not actually put the title on it. So I'm just going to change this real quickly. And you can also use Canva or something like that. I have done that as well, especially in the days before it was able to generate words.
Nico Lafakis:If you blow up that second one, I think it put it in the signage.
Mike T:Oh, yeah. Yeah. So yeah. Actually, what does that does that actually say on the signage? Why will it not let this thing get bigger for me?
Nico Lafakis:I kinda figured it hadn't fully rendered yet.
Mike T:Oh, yeah. Yeah. Yeah. There it is. Yeah.
Mike T:It you're right. Actually, you're you're absolutely right. If I zoom in here, you it does say rust and rubber. It says rust and and rubber here on this one. Yeah.
Mike T:You can. So what yeah. So, again, you could use Canva. You can use whatever tool. Cover for a book titled Rust and Rubber featuring elderly racers.
Mike T:So I just added the word title in there. I'm not even gonna worry about the aspect ratio. Let's see. And let's see what Grok did. Grok?
Mike T:I don't know. Like okay. That, like so it's giving people an old cars. This does not look anything like a like a book cover, I will say. And, like, you just gotta play around with it.
Mike T:Like like, this this guy looks pretty awesome. Like, maybe there is something you could work with, for this one here. But, again, this is what I would do. And, again, if, here we go. Here we go.
Mike T:We at least got one. We at least got one. Looks more like an album cover to me than a, sort of book cover, but we'll we'll go with it. Yeah. So, yeah, upper upper left there, you you know, that's something and you could play with styles and things like that as part of it.
Mike T:But yeah. And that's how I would then do that book cover. And look. That's it. You just kinda put these things together, put it in a file, and upload it into pretty much in no time can upload something like that into Amazon.
Mike T:So Rust and Robert will be live in, like, in 24 hours. I and and if anybody wants again, I'm not going to I take no money. I will try and make it I think I can make it free for, like, the 1st week. So if anybody wants to download and read it, like, I think I'm pretty sure I can make it just totally free for anybody. I think you just have to have an Amazon account.
Mike T:It ain't taking John Grisham's job anytime soon, but down the road, who knows? Who knows?
Chris Carolan:Well, I know if I had a teacher drop that assignment on me at this moment, I'd definitely be using AI to create that book for it. So
Mike T:Yeah. And and look, the shorter the form, the better it is, I will say. Like like, longer form stuff like this, this where it gets tough. Like, in over 10,000 words, it starts to fall apart at certain pieces, and I'm sure everyone will start to see that if they actually just skim through it. But we'll we'll make it free for all your listeners.
Mike T:So at least for a week here, I think I'm able to do that.
Chris Carolan:Awesome. Well, we appreciate you showing us that. What's
Mike T:your fun. It was fun.
Chris Carolan:What's your one takeaway for those out there that have had a book rattling around forever in their heads?
Mike T:Just play with the tools. Experiment. That that's it. Just honestly, all this stuff, you hear people talk about it, you do whatever, just figure it out yourself. Like, you don't need to be an engineer to use these tools.
Mike T:You just need to have access to a computer. Some of them, you don't even need $20 a month for. You just, like, literally can use the free version, figure it out. Just spend a couple hours, figure it out yourself. You know, if you want to publish a book that no one's ever going to read just to figure out, well, what's it like to publish a book?
Mike T:And just don't do it under your own name and have fun. Figure it out. So that's what I would say. Just experiment, really embrace that inner child. Well said.
Mike T:You don't need to be an engineer. You don't need to be a writer.
Chris Carolan:Need to be anything special. I'll start diving into these tools and and start learning. Aside from from checking out Amazon tomorrow to find this book, if people wanna follow along with what you're doing and what you're up to, where where should they go?
Mike T:Absolutely. Probably LinkedIn or Medium are the best places. I write on both platforms, same stuff. So follow whatever is more convenient. Just look for the last name, Tedasco, and I have a newsletter.
Mike T:It's totally free. It's just me musing about AI stuff. So, you know, follow me there, and hopefully, there might be some insights you'll be able to take into your work or personal life from the stuff I write about.
Chris Carolan:Awesome. Well, we appreciate you coming on. I look forward to catching the the book later this week, and we hope everybody has an awesome week out there. Have a good one.
Intro:That's a wrap for this episode of wake up with AI. We hope that you feel a little more inspired, a little more informed, and a whole lot more excited about how AI can augment your life and business. Always remember that this journey is just the beginning and that we are right here with you every step of the way. If you love today's episode, don't forget to subscribe, share, and leave a review. You can also connect with us on social media to stay updated with all things AI.
Intro:Until next time. Stay curious, stay empowered, and wake up with AI.
Creators and Guests
