Alan Petersen [00:00:00]:

You are listening to Meet the Thriller Author, the podcast that brings you face to face with the minds behind the most pulse pounding, suspense filled stories out there. I’m Alan Peterson, an author of mysteries and thrillers myself, which is why I interview writers working in the genre that I love to write Anne Read. With over 200 interviews and counting, I’ve spoken to thriller icons like Dean Koontz, Walter Mosley, Tess Gerritsen, and Lee Child, and many other amazing authors from both traditional publishing houses and the NBC. From writing tips to exclusive book previews and reviews, you won’t find a better resource for all things thriller and mystery. Dive deeper into the rabbit hole at thriller authors.com where you’ll find an archive of all my interviews, show notes, book reviews, and more. And don’t forget to sign up for my thrilling reads newsletter for exclusive deals on must read books in our genre that’s over at thriller authors .com. Stay tuned for episode 194 with best selling author and artificial intelligence expert, Andrew Mayne. Hey, everybody.

Alan Petersen [00:01:07]:

This is Alan with meet the thriller author. And on the podcast, I have, Andrew Mayne, who is a best selling author, magician, inventor, and an artificial intelligence expert who has appeared on Shark Week and A and E’s Don’t Trust Andrew Main. He was the science communicator for OpenAI and their 1st prompt engineer. At OpenAI, he worked at GPT 4 and ChatGPT, GPT 3, all that good stuff, creating many of the original prompts and examples used today. He’s now the founder of Interdimensional, an artificial intelligence service, consultancy that helps companies deploy AI solutions. And, of course, he’s a thriller author, best selling author. His next thriller, Night Owl, is a brand new series about a former cold war era counterintelligence agent who finds himself in a fast paced, and slightly insane world of Silicon Valley, which we published on December 1st and is available for preorder now. Where did you get this Silicon Valley, I wonder?

Andrew Mayne [00:02:02]:

I don’t know, man. Just this creativity is a mystery, man.

Alan Petersen [00:02:07]:

Yeah. We’ll get it to we’ll get it to all that, but thank you so much for coming on the on the podcast. I really appreciate it.

Andrew Mayne [00:02:12]:

Yeah. Absolutely. Thank you for having me, Alan. I’m glad to be here.

Alan Petersen [00:02:15]:

Yeah. You know, it was funny when I, when not, when this whole Chat GPT went to the public and I started looking into it And I started seeing your blog. Your website get popping up all the time, so I learned a lot from reading your website. And I was like, oh my god. That’s really cool that a thriller author is now, like, and, expert in this stuff and this whole artificial intelligence. And, you know, when I when when we first when you first were on the show on the podcast in twenty. I never even heard of, of JetJPT or OpenAI. Mhmm.

Alan Petersen [00:02:47]:

How crazy is that? When how did he get into AI? And then working for OpenAI, how did all How that come

Andrew Mayne [00:02:52]:

about? Well, I, you know, I’ve always been interested in artificial intelligence and robots. When I was a kid, I used to build, like I’d take coffee cans and stuff and build robots, and It didn’t work very well, but I would take apart toys and build stuff. And then when we got our 1st computer, I try to make very simple Chatbots, you know, just basic sort of thing. If you go, what’s your favorite color? Red. And then, like, oh, you know, they if red appears here, then say, cool. Like, apples are red. I’d make very simple stuff. And then by the time I got to high school, you know you know, when My senior year of high school, maybe Gopher had came out.

Andrew Mayne [00:03:31]:

That was, like, the first, you know, browser, like, the first like, the World Wide Web was Really even, you know, it’s impensating temporarily. We’re still finishing this stuff. And so I was sort of determining my career path on the idea that, like, well, Tech is exciting, but it’s not a lot of fast moving stuff. It was the role of Windows. Apple was sort of in its decline and whatnot. So I was really attracted to You know, I was a magician. And so I lived in South Florida. I had the opportunity to become a cruise ship magician and perform in resorts.

Andrew Mayne [00:03:56]:

So straight out of high school, I did that. Headline, I traveled around the world doing magic because that just was seemed more attractive. I didn’t think I would be missing anything out. And then while doing that, you know, the Internet happened then a bunch of other things happened, and I kept thinking, hey. I wanna get more involved in technology, so I followed stuff pretty closely. And I launched, you know, a blogging company In the early aughts at a podcasting company and stuff, but, you know, it was more of a thing I would follow. I’d read research papers and key you know, keep a little bit into it. But it wasn’t until About 8 years ago, 7 years ago, I decided to get kinda real serious about all of it.

Andrew Mayne [00:04:31]:

And, you know, Since then, I did a lot of fun stuff, and I was in the middle of actually working on a special for shark week. And the I pitched them the idea that I was gonna make myself visible to great white sharks. You know, very idiotic thing to do, but, you know, whatever to get on TV. And working on that, I wanted to build a system that would let me see, like, 360 degrees around me. And I’ve been playing around with different AI stuff at the time. There’s a thing called, you know, GANs, which is generative adversarial networks, which basically is a way to, like, The 1st image models you’d see where they created kind of things you could look at, use that. And I trained little ones on my computer to do that, and then, you know, I built some started playing around with, like, some text Generation models at that time were very, very simple. And then I kinda got a little more serious because I only really got picked up programming like a year before, learned Python.

Andrew Mayne [00:05:16]:

And then for the Shark Week thing, I started playing around more with image recognition and started to train models to do other stuff. And it was very fascinating. And so I just sort of got really into the field, and it was really the thing that hooked me into it was while I was learning about building image Detection systems and the way they work, they have different sort of algorithms, but one of the things they often look for is, like, you take an image, you take out all the color, and you reduce it to look for edges, and you look for stuff. If you wanna spot a face, You know, look for 2 eyes and 2 nostrils. You know, that’s one of the things you could do. And there’s a thing called the HAR like features, which is sort of an edge detection thing. While I was working on this, I was talking to a shark researcher. They made a comment to me.

Andrew Mayne [00:05:52]:

They said, yeah. You know, if you’re in a cage, the great white doesn’t See you. They just see the outline of the cage. And I’m like, That’s a very interesting thing because that involves a certain amount of post processing for what they’re seeing. And If you think about it, great white sharks have a tremendous amount of sensory input. They’ve got incredible vision. They’ve got great sense of smell. Oh, they could pick up vibrations, electrical fields, all these other bits of data, but they got small brains.

Andrew Mayne [00:06:18]:

But sharks have been around longer than Trees. You know? Sharks have been around, like, 300,000,000 years. They’re a very, very successful species, and part of that is be be by being very efficient. And I thought, at the time, the way they described it sounded very much like a very efficient image algorithm. Like, what’s the minimal amount of resources I need to compute to figure something out? And that led to me figuring out my intuition on how I was gonna fool great white sharks. You know, I pitched the idea of this sort of predator suit with LCD camouflage and stuff, which worked for 15 minutes in the ocean then totally started to short out, but I had a better method to do that. But That parallel between artificial intelligence and what was going on there in biology really got me interested. Because once I had an analogy to get hold of, I just sort of Dove right in.

Andrew Mayne [00:07:01]:

And not long after, OpenAI, who I’d been following because I thought it was a very interesting effort. This company was Start as a nonprofit, and their goal is, you know, how do you prevent AI become from becoming Skynet? How do you prevent it from, like, you know, you know, the existential threats and whatnot? And they had released they did some cool stuff like having, you know, having training systems that could beat the world’s best gamers, like in defense of the ancients and whatnot, and then They put out a thing called GPT 2. They did GPT one, which was a paper that got a little bit of attention, and GPT one basically took something that Google had discovered called the transformer. It realized it was really good for predicting, like, the next word in a sequence. And the way this did work was basically, if you took a ton of data, ton of text, whatever, The AI would learn by looking at stuff and trying to make a guess what the next word would be, but then kinda get look ahead and guess ahead. And then you can make a better effort than just trying to guess a word. An example would be if I said to you, I went down to the park and I Walked into the dugout, and I saw a bat. In a traditional model, you would just assume it’s a baseball bat because you heard park, you heard dugout, you heard this.

Andrew Mayne [00:08:09]:

But if I said a bat Hanging from the rafters and it flew away, well, then you would know it was the mammal bat. And a transformer type model can actually figure out like, oh, May not always be baseball bat. It could be this other thing. It language a lot of taking different paths and stuff. And by scaling these systems up, They did from GPD 1 and then GPD 2. They’re able to actually have GPD 2 could generate coherent text, and it could kinda follow that for a couple paragraphs. So when they released that, And so writer, you know, I’ve been writing for the last, you know, 16, 15 years now. You know, I was very interested to see what was the intersection of AI and language, And they posted all the examples on GitHub or repository for code and whatnot.

Andrew Mayne [00:08:49]:

And I read through every single example they post because at that time, you couldn’t play with GPT 2 yourself. So I read every single example they had there and trying to understand what was the process. As they did that, I could kind of see, okay, it’s paying attention to the original prompt here, Then it’s sort of going off in this tangent, and now it’s paying attention to there, and it’s going to there. And it’s kinda relates to a thing. There’s a thing called attention mechanisms, which basically has to do with how the model how it’s gonna train itself and proceed, and you could sorta see a pattern to that. So I became obsessed with this. And then Out of the blue, OpenAI reached out to me in early 2020, and somebody I had known that I was really into kind of the stuff and following a bit, but Talking about it on podcast and stuff, and they said, you’ll put you on the NBA, and then they said, do you wanna play with GPT 3? I’m like, GPT 3. Like, yes.

Andrew Mayne [00:09:37]:

I do. And I was obsessive. Like, I just would stay up late. I just put out a book that was between, you know, book releases, so I had plenty of spare time. So I just released a book and just spent I’d be up till 4 or 5 AM playing with it and trying to understand, taking what I knew about GPT 2 are taking taking a look at GPT 3 and figuring out how do you get it to do what you wanna do. To have to understand it’s between GPT 3 and chat GPT Is GPT 3 was a model that was just trained on a bunch of text? It wasn’t taught afterwards. If somebody asks you to make a list, make a list. Somebody says do this, do that.

Andrew Mayne [00:10:12]:

It literally was just trying to predict the next sequence from a bunch of text. So if you said, you know, This is, you know, Alan Peterson’s blog about, you know, the movie alien and what I like. It would go, okay. I guess this is what I would What would something on the Internet or a blog post look like? It would continue writing down. If you said write a blog post about this, maybe it might work, maybe it wouldn’t. So prompting was very different then. Prompting was really trying to figure out how to get the model to do what you wanted it to do, And that kind of real you had you had to look at the world of text to say, if I wanted to be factual, then I should probably say, You know, this is an abstract, you know, title abstract, you know, materials research study or something. Give it in the world of text where it would Find something.

Andrew Mayne [00:10:57]:

If I wanted to be sort of truthful, I would figure out a truthful story. If I wanted to be creative, I’d figure that out. So OpenAI was a very small company at that time, and, you know, here I was just showing all these examples. I’d send them videos and stuff and also be coding up examples using Making apps for the iPhone and all this and showing a lot of this potential. And they asked me, said, you know, would you like to work with us? And I’ve never had, like, A real job in my entire life. But I looked in yeah. I know. And I looked at what’s going on there, and I said, yeah.

Andrew Mayne [00:11:27]:

So They hired me to write prompts. They hired me to be I was running project called Creative Applications, and that was in 2020 and Work with different teams there, the applied team, we work with researchers. And then 2 years ago, I became the science communicator for OpenAI where, basically, I would help, you know, help figure out how do we explain these things, how do we do that, and still play with the stuff and do all that. So very long explanation, but that’s how I came in to work at OpenAI.

Alan Petersen [00:11:54]:

So we so you were there when they went to the public, when they opened it up to the

Andrew Mayne [00:11:58]:

public? Before. Yeah. Actually, yeah. Yeah. I was at they I was one of their the alpha testers, And, you know, that was and then so I wrote the the documentation for GPT 3, and I wrote the examples for GPT 3. So the stuff you go there and they see, All those things, airport goes, oh, it’s me. That was, like, late at night. Like, I wonder if it could do this.

Andrew Mayne [00:12:16]:

And then internally, they’re so busy building and doing stuff That you know, I’d get a call, you know, somebody would say like, hey. Do you think you could do this with it? I’d be like, I don’t know. And so I was the prompt whisperer internally. And and not for any amazing skill or talent, but just from sheer maniacal energy at which I threw at the trying to figure out how to get to do cool things.

Alan Petersen [00:12:39]:

Were you, what was the when they got released in November to the public, did you Were you surprised at how it exploded? Were you expecting it? How So It must have been crazy for you guys.

Andrew Mayne [00:12:50]:

Yeah. So it’s a background. So we had Back in 2020, we we released in, like, the summer GPT 3, that which was an API where you could go online to the website there and use their playground to play with it. Right? And it was very powerful, but it wasn’t as accessible to people as, you know, perhaps it could have been. So that was one of the things OpenAI work towards doing was trying to improve that. Of the things they did is they looked how people use it, and they created a category of model called the instruct models. And what instruct models were were basically taking, You know, an example of, you know, make a list of 10 science fiction books and it gives you a list of this. It’s not you are a blog and you’re supposed to be doing this and blah blah.

Andrew Mayne [00:13:27]:

And that’s why I see a lot of people doing prompting, a lot of prompt experts today that are doing wasting a lot of words and stuff too because I’m like, these new models are different. It’s not the same as it was before. But anyhow, The new instruct models became more capable, but still there was a little bit more to do. So, a team at OpenAI Basically said, let’s let’s take a ton of data on basically really good interactions between what people want and how they will succeed, also help guide the model so it’s not give you things you don’t want. You don’t wanna say, I’m having trouble at work, and it tells you you should punch your boss. That would be bad. And so, basically, it’s called reinforcement learning with the human feedback. Well, basically, ideas like, it would try a bunch of different things, and every now and then, he would say, yes.

Andrew Mayne [00:14:07]:

This is good. No. That’s not. It would keep that cycle going until it build up a lot of data. So that newer model, we the Instruqt series, they took that, and then they took a user interface. And we we were in November. We had a meeting because, Remember November last year, still not even a year ago. That’s the crazy part.

Andrew Mayne [00:14:25]:

We’re we’re less than a year at a chat GPT. Last year, we’re sitting there planning for the rollout of GPT 4. Now GPT 4 had been ready for months, but we’ve been testing it. So we’re just under the hood hat. Weren’t, you know, publicly talking about it, and we knew we were gonna have, like, a rollout in, like, March. So understand we were months ahead of from g p d four, but we knew it was gonna be very crazy with that. So one of our researchers, a friend of mine, says, hey. We wanna go test this thing, and it was called a low key research preview.

Andrew Mayne [00:14:53]:

And we’re like, well, how many users do you think are gonna wanna play with this? And we’ve had other things where people tested and whatnot. And, you know, we, You know, maybe 10,000, worst case, a 100,000 max, whatever. We had no expectation for what this was gonna do because understand Core capability of chat g p t had been in the model called GPT 3.5, which had been available for 6 months before. Everything you every cool thing you could do with it, you could get it to do with maybe a little bit more prompting and in a different interface. So it wasn’t like we thought Up capability wise, this is a new model. It was more like for ease of use wise. Well, you reduce the friction in prompting it. You reduce the friction of the interface.

Andrew Mayne [00:15:33]:

You reduce the friction of having to create an account. And pretty much, there’s no friction, and you end up with the chat GPT phenomenon. And funny as it was is The 1st week or so, it just started going crazy. But middle of December, we had an employee Christmas party, and one of our execs is like, well, you know, everybody’s talking about it now, But, you know, the Internet, a week from now, nobody will be really paying attention. Turned out not true. And so it it was Insane. And and I had my running joke in one of our Slack threads was like, okay. You I’m like, we I said it’s not when not if, it’s when we are going to be a subject of a South Park parody.

Andrew Mayne [00:16:11]:

And then, you know, we were at on Saturday Night Live, South Park parody, and then, you know, the latest, you know, Prime you know, latest primary debate, guess what? You know? If it’s a check that punch line, and it’s like, that’s how fast that adopted. So it’s a whiplash for everybody.

Alan Petersen [00:16:27]:

Yeah. I saw that episode on South Park. It was written by I think it was Trey Parker and Chat GPT.

Andrew Mayne [00:16:32]:

And I think they changed it. I think they changed it. Originally, I think they went off and change And I I looked into like, my wife’s a filmmaker, and we looked into indvpro to see if we could actually create an account for Chat GPG. Because I thought how awesome would it be if it could have all anybody who use it could just credit it.

Alan Petersen [00:16:48]:

Yeah. If I got worried about it, oh, no. No. The JettyPT is gonna have,

Andrew Mayne [00:16:53]:

Well write stuff

Alan Petersen [00:16:54]:

out of the episode. Or

Andrew Mayne [00:16:56]:

Yeah. And I and I and I get, like, there’s yeah. A lot of writers are concerned, which I’m happy to talk about that if you want. But

Alan Petersen [00:17:02]:

Yeah. Yeah. So I was gonna say because, you know, being hanging out in the writer’s forums and Facebook groups, you know, for writers, especially self publishing in the indie world. I was kinda surprised how how, toxic it it’s gotten the, the, the against, against, AI. I understand people are fearful. They don’t understand it. Do you, did you realize it was gonna be this, this is gonna be such a pushback against it. And now, like, George George r r Martin and John Grisham and others are suing it for copyright infringement.

Alan Petersen [00:17:34]:

Could you just speak to all to all that craziness that’s going on right now?

Andrew Mayne [00:17:38]:

Yeah. I mean, I can’t speak to any, you know, obviously, any litigation or like that that’s going on, but but I I could talk about and I I could say that there are some severe. I have questions about some of the technical claims made by some of these people doing that, but that’s neither here nor there. I would say in general, yeah, I am surprised In one the the the rate at which the fear is sort of set in, I you know, as a writer, as a guy who does 2 books does 2 novels a year and who does take my writing seriously. You know, I I’m a Wall Street Journal bestseller. I’m a 2 time thriller award finalist, Edgar finalist. And so I think I I, you know, Think are good, you know, for the most part, and I’ve been very I’ve had, you know, my book, Netflix, spent 7 weeks on the bestseller list on Amazon, number one for everything. And so I’ve had success.

Andrew Mayne [00:18:24]:

I’ve had experience of trying and struggling. I started off as an indie writer. I work with publishers, and I work with a great publisher. I’ve been I’ve been there have been years where the only income I had was either what I was making from Kindle Digital Publishing or from selling, optioning, film rights, and stuff. Often, I’ve had lean periods. I have great periods and whatever. And so I’ve I’ve I’ve, you know, been involved in writing as a living for a good portion of my life. I Never worried about AI taking my job, and although I assume at some point that it would.

Andrew Mayne [00:18:56]:

You know, when GPT 2 came out Could. You know? Oh, when GPT 2 came out, you know, I looked at the quality of that and said that it’s not quite there yet. And then GPT 3 came out and I said, oh, this is better. And I experiment with using it in books, use it in stuff. I think the 1st published stuff, the u g p three or g p four is probably, like, little some paragraphs and stuff in my books that I would do as an experimental pass way. Like, If I wanted to have an AI talk, I would do that, but I don’t I don’t let it write for me because I love to write. You know? Like like, you know, it’s like why do people why would you kayak if you can have a powerboat? Because you wanna You know, why would you hop mountain climb if you could have a helicopter? Because you wanna mountain climb. You know? Why would you fish if you got fish? Because I like to like, I like to write.

Andrew Mayne [00:19:32]:

And so If people don’t want to buy my books because they prefer AI books, then that is their choice. I’m still gonna write. I got into this as a writer. If you’re you know, when it comes to me as a person having a career in the space, I think we’re gonna enter a golden age. I think that we’re going to enter in a period of the the opportunity for creativity is going to And more than there are going to be creatives to fill it. And an example I give is when I present, you know, talks to creatives, I show a slide, which was the first photo that de Geur took back in, like, 18/38 or something of a person. And what it was at that time, you had to leave a camera exposure open for, like, 20 minutes, 30 minutes, whatever, because the light just the film wasn’t that sensitive. And nobody would ever sit still long enough to have a photograph, but in this case, he took a photo of a street, had a man stop to get his shoeshined.

Andrew Mayne [00:20:22]:

And so the man and the shoeshiner were captured long enough to see there. So you can look at this photo and you can see this, and that’s the first time, you know, you have A machine was able to capture people, capture life, capture trees, everything altogether, and that’s an amazing point in time. Think prior to that, no machine could really capture in detail something on par or greater than a human could be able to do. And that would have been troubling and fearful for some people. I think that some people fear what would this mean for portrait painting and what would people mean for this. But fast forward, you know, Today, there are more people employed in creatives in a on a living wage than there were at that point in time as a percentage of the population. And you have creatives who are billionaires, and there were not any billionaire creatives or anything like that time at that time. And there are people living very comfortably now as creatives.

Andrew Mayne [00:21:11]:

You take I show that photo, and I show a photo of James Cameron on the set of Avatar Way of the Water, where he’s using his digital camera, which has no lens. It’s merely a positional system that he sees in a three-dimensional space. And I say, okay, imagine going back in time to 18/38 and trying to explain to a photographer, What is TV? What is film? What is computer animation? What is, you know, try to explain them what Avatar the Way of the Water is. Like, well, There’s a thing called a movie. Like, okay. What does that mean? Well, instead of a painting, it’s this thing you just saw a photograph, and it’s going to be 24, or maybe his case, 48 frames per second, 40 days a second. Well, how do you capture it? Don’t worry. It takes me takes me an hour to capture one still.

Andrew Mayne [00:21:52]:

Like, we solved that. You know? And then people pay. They go into a theater like Play. And, like, what happened to play? Is no place are still a thing. Broadway’s still here. It’s still a thing. You know? And then you have trying to explain the world of cinema, the claim of full of video games, off to somebody, it would just melt their head. Why is this gonna be any different? Why is it gonna be any different now that we have another acceleration technology? And I think the opportunity forms or creativity are going to increase tremendously.

Andrew Mayne [00:22:18]:

And if you are a creative that knows how to use them as an amplifier, you’re going to be in a great position. And I think the Fear I hear from people is I you have to change. There is going to have to be change, and we don’t wanna change. And it sucks when something changes, but If change happens because people are making a choice, if people choose not to read my books because they can choose something else, then that is their choice. I don’t wanna in front of that and say, no. I like my way of life. I like this. Well, I do, but I don’t wanna interfere with somebody else.

Andrew Mayne [00:22:44]:

If somebody wants AR over mind, then that should be their choice to do it. But I like to think There people are gonna still want human made stuff. I’ve got Brandon Sanderson’s book collection up here. But and Brandon Sanderson, you know, he did this Kickstarter. He raised $40,000,000. $40,000,000. Okay? Publishers are looking at this going, how the hell do you do that? And I’m gonna give you as a point of context, Alan, I’ll show you something kind of a very funny stat. That’s sort more sad is he raised $40,000,000 for his Kickstarter.

Andrew Mayne [00:23:17]:

And if one of the companies I like to track just because out of perverse curiosity is Barnes and Noble. Okay? And if you look at what the value is of Barnes and Noble’s, like, market cap, It’s like well, Barnes and Noble education. Let me see if I can find Barnes and Noble. Like, it’s Barely, you know, it’s maybe 10 x that. You know? And so 1 author was able to go raise, you know, know, a significant amount of some you know, I think JK Rowling’s worth more than Barnes and Noble is now, and that shows you the power of a creative. It shows you the power of being able to have that. And the thing my point about Brandon Sanderson, I backed his campaign. Here’s a fun fact.

Andrew Mayne [00:24:01]:

I’ve never read one of his books. I’ve never read one. I hear they’re great. He’s a super cool guy. I just wanted him to win. And if your readers and audiences aren’t wishing for us to win, AI ain’t the

Alan Petersen [00:24:14]:

problem. Yeah. I think that’s a thing too with, like, even this the screen actor the writers, The writing guild right now that’s on strike in the Hollywood. They’re very concerned. One of the points is there is is, AI. They’re worried about AI use. And And, yeah, imagine now you you you you’re looking to get an advance nowadays, and he got a $40,000,000 it’s like a $40,000,000 advance.

Andrew Mayne [00:24:39]:

Yeah. And and it was it was people believing in him. You know? It it it it’s people because people believe by the way, Barnes and Noble’s operating income from 2019 was 38,000,000. So it gives you an idea of, like, how how the things are. So And and then people like, well, Andrew, that’s like your Brandon Sanderson Sanderson or Jake. Yes. But if you can achieve 1 tenth of a success, you’re gonna be great, and that’s going to take work. It’s going to be a change you’re having to adapt.

Andrew Mayne [00:25:06]:

And I’m not saying it’s gonna be fun or it doesn’t Suck. It does. And I’ve gone through this myself. I had a DVD publishing business where I made magic DVDs and instructional DVDs, and it was a great business. I could put out 3 or 4 of these a year. I had really good income. I mean, good income for a single guy living in South Florida. And then the Internet came along and file sharing, and I watched I watched, you know, people who I in the business who I knew, you know, start pirating stuff.

Andrew Mayne [00:25:33]:

You know? And I have friends Friends would just oh, no. It’s you’re gonna sell more. I’m like, no. But I said, that’s a lot. I said, that is a wish. That is a wish that people keep telling me about piracy that I’m going to sell more because of this, and it did not. It killed my business. That being said, I’m not a fan of copyright piracy.

Andrew Mayne [00:25:50]:

And, you know, when somebody’s you know, People but the thing I got out of that, the Internet was great, and my next business was even bigger and was a bigger platform. And the next Platform at that was even bigger. Every time I’ve had to adapt, do you the next platform I moved to has been an even bigger, better one.

Alan Petersen [00:26:06]:

And that’s one of the concerns that I hear from, people in the writing community that are worried about it. They’re this they’re saying that if you use Something like chat gpt, you’re, you’re basically participating in copyright infringement. But can you explain that a little bit? Because that’s It’s so hard to explain it, but it’s, like, it’s it it’s anticipating things. It’s not like actually just copying a book

Andrew Mayne [00:26:29]:

and No. It These things, they’re not retrievable. Like you said, these things are building predictive models. I my favorite authors are Michael Crichton, When I share an agent with, Isaac Asimov, Stephen King, and like Tom Clancy. If you read my and then for a point of view, Susan Collins’ Hunger Games first person POV. Like, I write first person POV because I read Hunger Games. I’ve read Susan Collins’ style, and I love That felt like a fresh way to write, and it felt very easy for me. I ripped them all off.

Andrew Mayne [00:27:02]:

I ripped them all off. I’m here admitting it right now that if you look at that, my 1st book I wrote, Public Enemy 0, I was trying hard to do, like and Michael Crichton and Stephen King are very different writers, But I was trying to sorta do a hybrid of that. And and this was the way I would learn phrasing, the ways I’d learn that. I read James Patterson. I’d learn how to do this. So I learned. That’s how I became A writer. And the idea that we can say, you know, that you can’t learn from culture In books and separate part of culture scenes, it just seems crazy to me.

Andrew Mayne [00:27:33]:

And then because you sort of say, okay. What can you and can you not do? You know, if you say, you know, what Can I learn? What can’t I learn? If I is a can I can I learn this? Can I save book? Can I save examples and copy things into my notebook to learn from whatever? And people might argue, well, it’s the scale. You it’s a scale. It’s like, okay. What’s the limiting factor of this? You know, where is the point where it’s a problem? I don’t you know, to me, it’s just it’s it’s arguing it’s Partially, problem number 1 is I think that we culture is built upon this. I’ve learned everything from people around me. It continues on to do this. If I build a Tool, whatever that does the same thing, that’s the same thing.

Andrew Mayne [00:28:07]:

Why should that be kept separate? And if we just say, well, it’s scale or whatever like that, that means electricity. You know? It’s it’s unfair to steam. It’s unfair to this. Well, no. It’s just we can say, well, it cannot it can’t be too good, you know, and that’s then we lose out on the things that we gained from it. So I just just signed a 4 book deal. K. And I’m a very busy person as we all are, and I have time to sit down and write.

Andrew Mayne [00:28:32]:

But the time it takes me to sit down and to organize all my notes and put that stuff into that format to go to the editor to say this is what I wanna do, all that. That is a pain. That is a pain for me. It’s not a fun part of the creative process. So, you know, I use chat CPT to take My notes, it didn’t write any of my pitches. I wrote all those, but I said, hey. Take this take all these notes and all this messy space and consolidate this down for me, And it’s made my life better. You know? I use it to do edits.

Andrew Mayne [00:28:58]:

I use it to do research. It’s an amplifier for me. I’m able to do 2 books a year and work at an AI company and do all my side projects in part because I use this as an amplifier. And I think the sooner other people do this, The better. And I get it. You know? People in writers’ rooms and stuff are afraid of what’s going to happen. Well, like, listen. Like, you need to be more worried about TikTok.

Andrew Mayne [00:29:18]:

You need to be worried about, you know, Americans liking to watch soccer more. There’s a lot of these other factors are gonna be way more displacing than I think, you know, AI is and I think long run AI is gonna create a huge way a whole world of which we can create that we haven’t thought about.

Alan Petersen [00:29:32]:

Yeah. I think that’s really interesting too. It’s like having a a a whole staff now. Like, you know, You and Chajeepiti is kinda like a a a staff to help you get organized and and do your research and and all that stuff. Okay. Curious how else you’re using it. I mean, do you do you like, were you are you an outliner? Is that or do you, like, write from the chin of your pants?

Andrew Mayne [00:29:53]:

No. I’m definitely an outliner. That’s the only way I can write as efficiently as I do. I’ve got actually outline for book over here. So, Very much I used to be a seat of the pantser, but then, you know, when you’ve got 10 days to turn around a novel, seat of the pantsing does not work for me. And so I I I build build up outlines, you know, and, traditionally, what I did is I just did book outlines, but now I start with world outlines. I start with the outline of what’s Going on in the world. And then I because a lot of times now mystery thrillers is sometimes I have to uncover stuff that happened before the story started.

Andrew Mayne [00:30:26]:

And usually, I would just put that in the middle, but now I’m starting with the timeline of the world, then I go write the book outline based on that. But That’s made it you know, that makes it extremely efficient for me because I get to just spend my time doing the thing I like, which is sit down and, like, Throw myself into a scene and go, what happens? How do they solve this? How do they get out? You know, what’s the change? What’s this?

Alan Petersen [00:30:48]:

Yeah. I think it’s been it’s really good too because, especially writing thrillers and mysteries, you know, the readers there expect certain tropes and certain things to happen. So, you know, it’s not like we’re now it’s not like as we’re reinventing the wheel here.

Andrew Mayne [00:31:02]:

Yeah. But I don’t yeah. But I don’t use AI for that, though. I don’t use AI for my lot. Right? Yeah. I mean, I I don’t and and, like, AI Can’t write a novel right now. It I mean, it could it could put a bunch of words in a row. It doesn’t get structured.

Andrew Mayne [00:31:13]:

The I can go on at length, and I’ve heard, you know, some of the things I’ve heard will claim like, no. It’s not. I can, I can tell you when that’ll probably happen because you have to have certain technical things happen, but it’s not there? It can write short sketches and sort stuff like that. And I think that I I had I was just on a podcast right before this, and somebody said, well, I think, you know, I think mediocre writers are gonna be the ones most scared. I’m like, well, the is is we all are afraid of remedial. We’re all afraid of hacks. We’re all afraid of that. And so, like, it’s scary, but it’s also The future is going to happen.

Andrew Mayne [00:31:43]:

The future is going to happen. And and even if, you know, OpenAI shut down right now and Google shut down right now, There’s a huge open source movement right now to train models. There’s a huge group with there, and those things are getting better and more sophisticated. And You can wish the world was another way. And, I mean, one thing if it was a clear and out and out horrible thing, if if if we were talking about, you know, it was 1939, And it’s not. You know? It it is the age of electricity. It’s the age of the atom or whatever. It’s some other prospect, which could be good, could be bad.

Andrew Mayne [00:32:15]:

But if you focus on, like, well, Progress progress is happening. How do you adjust to

Alan Petersen [00:32:20]:

progress? Yeah. That’s the thing. We hear all the the clickbait articles about, you know, AI books flooding Amazon. But that’s scammers, and scammers are always gonna scam. Even before AI, they

Andrew Mayne [00:32:31]:

were Yeah. They get, like, Three reviews, and I’ve seen people post, like, code to write it. And my challenge to anybody doing that right now is tell me you read one of those books. Yeah. You know, that was like because, like, Amazon just put out a thing today where they’re limiting a person can’t upload more than 3 books here. And so, like, And there’s ways to sort of, like, if the danger is crap, that’s fine. You just go look at what’s got 2 reviews. Chances are nobody else wanted this book.

Andrew Mayne [00:32:56]:

But eventually, they will get good. They will get good.

Alan Petersen [00:32:58]:

Do you think that in the future, like I mean, you’re saying they will get good. So are those fears correct sometimes? Will we be replaced, or will there always be a room for a writer like us, human?

Andrew Mayne [00:33:07]:

It will always pull up from artist. I mean, do do you do you buy stuff purely because, like, Taylor Swift is doing blowout con and I’m gonna say it again. I’m not making a concert argument because I always hated that. But, like, Taylor Swift is, you know, doing blowout attendants at concert arenas around the world. You can just listen to our album. It’s cheaper. You just listen. Why? Why you have a better experience in home headphones? Why? There’s something more there.

Andrew Mayne [00:33:32]:

You know, I buy books from authors that I like as people, and maybe I will like some AI generated fantasy stuff. But we love We love that there is a George r r Martin there creating stuff. We love that there was a Robert Heinlein. We love that there was a person that we identified with we connected to. We love that there’s this Crazy weird guy in Maine named Stephen King who’s writing this stuff. We’re just as fascinated sometimes about the people who make the stuff. But if you’re anonymous, I don’t mean if you’re anonymous And you’re interchangeable, then, yeah, AI is gonna be a threat. And and and we may find that some of the office we really like were actually AI, and I don’t think that’s gonna go over with all.

Andrew Mayne [00:34:07]:

You know?

Alan Petersen [00:34:08]:

Yeah. Yeah. No. It’s good. It’s gonna be interesting future. That’s for sure. Because you are traditionally published and you have agents and everything. Is is that something that are they asking about AI? How are you using AI? Because now with Amazon, KDP, Kindle Direct Publishing, you have to check a box whether or not you’re using AI or

Andrew Mayne [00:34:26]:

not. Yeah. So The the the publishing industry is certainly talking about it. The the questions about that. But I remember the publishing industry is still getting over, you know, the introduction of the Selectric typewriter. So their bill yeah. Their ability to like, yeah, even ebook stuff, that’s I could go on about how They they’re don’t even realize how much the paradigm of what a book can be and what the opportunities are right now. Like, there’s just you know, we So many books would be better at a 120 pages, but published like, oh, the economics doesn’t work.

Andrew Mayne [00:34:58]:

When you were shipping paper, it didn’t work. But in the world of, you know, bits, it does, and there’s people doing really well by that. So I’d say there’s some concern and whatnot about that. I mean, a lot of concern, a lot of lot of talk about that in that world, and more of like, well, what does this mean? What does this mean? You know? I don’t know. I mean, maybe it won’t all be rosy, but I think that the opportunity is huge there. And I think that, like, I I, you know, I have some foreign editions of my books, and I get covers back. I’m like, Midjourney. You know? Not even DALL E.

Andrew Mayne [00:35:27]:

I’m like, I know I know how you did this. You know?

Alan Petersen [00:35:31]:

So tell us about your new new book that you have coming out. It’s a new series.

Andrew Mayne [00:35:34]:

Yeah. It’s called night owl. It’s about a a former counterintelligence operative named Brad Trasker. And I wanted to tell a story of somebody who was sort of kind of a bit of Cold wear or or, you know, post cold war sort of person trying to adapt to what it is like today. And if you think about the last 10 years, what’s happened is From cryptocurrencies to blockchains to AI, things have changed rapidly. And you have entire you hear about gangs that what they do is they might a game could be, like, 4 dudes in Guatemala who are really good, you know, Python programmers who figure out an exploit and are able to take your hospital data And lock it up and charge you, you know, make you pay a ransom. And so these things are changing a lot, but also that role to where we are right now, but it still interfaces with the real physical world. And, you know, you look at the Ukraine, which is launching attacks against Russia using drones, using DJ Mavics and stuff.

Andrew Mayne [00:36:26]:

And so it’s a guy trying to adapt to this changing world. And so he goes off his Dissolution, obviously, goes to work for a tech company because he comes across a young woman who’s very smart and, you know, is in a situation where she could use somebody with his skill set And basically finds himself in a very different world where the stakes are even higher, so to speak, and the the the level of sort of –ceptions, and he has to adapt and figure out how does he apply what he knows in this world.

Alan Petersen [00:36:52]:

And that’s coming out on December 1st?

Andrew Mayne [00:36:54]:

I believe so.

Alan Petersen [00:36:55]:

Where can the, listeners find you? Because, you have some great information on your books and AI and everything. What’s the best place to find you?

Andrew Mayne [00:37:02]:

On Twitter at Andrew Mayne. That’s at Andrew, m a y n e, and then my website’s just Andrew Main, andrewmayne.com.

Alan Petersen [00:37:10]:

Alright. Well, Andrew, thank you so much. It’s just fascinating talking to you, but thrillers and AI. It’s just a crazy time. It’s guys it’s exciting, I think. But

Andrew Mayne [00:37:19]:

Yeah. I I think that I I know. I’ve I’ve deal with some very charged emotions online with people, and I think that it’s one of these things when you start to play with the tools and you see the potential behind the scenes. I’ve Talk with some incredibly talented creative people, you know, very, very successful people who are very excited about it because they look at the biggest limitation on them is not their creativity. It is tools to amplify, and that’s where we are right now. I think we are heading towards a golden age if you want it.

Alan Petersen [00:37:46]:

Alright. Thanks, Andrew. Appreciate it.

Andrew Mayne [00:37:49]:

My pleasure.

Alan Petersen [00:37:50]:

Thanks for tuning in to today’s episode of meet the thriller author. I hope you had as much fun diving into these mysterious and thrilling worlds as I did bringing them to you. Your support keeps the show going, and I’d really appreciate it if you could take a moment to rate and review the podcast on your favorite app. It helps me make the show better, and it also helps other thriller enthusiasts find us. For more details about this episode, past interviews, and other great resources, head over to thriller authors.com. While you’re there, make sure to sign up my thrilling recent newsletter. You won’t wanna miss the exclusive deals and book recommendations.

Alan Petersen [00:38:26]:

If you’re interested in my own writing, you can find more about my mystery and thriller novels at alanpeterson.com. You’ll find all the details about my books and what I’m working on next. And remember, that’s Peterson all e’s, not Peterson. So that’s elenpeterson.com. Until we meet again, keep turning those pages and enjoying great reads. Thanks again, and stay thrilling.