In this week’s episode, I take another look at generative AI tools and consider what they might mean for writers. I also take a look at March’s ad results and discuss the very excellent board game HEROQUEST.
TRANSCRIPT
00:00:00 Introduction and Reader Updates
Hello everyone. Welcome to Episode 153 of The Pulp Writer Show. My name is Jonathan Moeller. Today is April the 6th, 2023, and today we’re going to talk a bit more about generative AI and writers. We’re also going to see how my ads did for the month of March and talk about a board game I’ve been enjoying lately. I’m recording this a few days earlier than usual because I want to take a couple days off for Easter, though the show stood still should still come out at the usual time.
First up, writing progress. As of this recording, I am 72,000 words into Cloak of Dragonfire. I think it’s going to end up around 100 to 110,000 words long, so not too much left to do in the rough draft. If all goes well, I’m hoping to have that out in the first or possibly second week of May. I passed the 10,000 word mark of Silent Order: Thunder Hand, and I also finished writing the first chapter of Sevenfold Sword Online: Leveling, which will be the second book in that series and Silent Order: Thunder Hand will be the 13th book in the Silent Order series. Both those books will probably come out sometime this summer once Dragonskull is finished, so I only have two more books to write in Dragonskull and I want to do that and finish that series before I really do anything else.
In audiobook news, Dragonskull: Sword of the Squire is out and you can get it at Audible, Amazon, Apple Books, Google Play, Kobo, Chirp books, Scribd, and probably a few others that I am forgetting off the top of my head. Oh, my Payhip store as well. Initial reviews are good and I’m glad everyone is enjoying that. In other audiobook news, I uploaded Ghost Exile Omnibus Three which will be a combination of the audiobooks of Ghost in the Throne, Ghost in the Pact, and Ghost in the Winds. And if you’ve read those books or listened to the audiobooks, you know they’re long and the combined omnibus version of those three books will be over 45 hours long. So that is good value for your Audible credit. That should hopefully be out sometime before the…excuse me…sometime before the end of April.
00:02:12 Reader Questions/Comments
Before we get into our main topics for the week, let’s have a few questions and comments from readers. Our first comment is from Scott, who says: I’ve just started reading your Dragonskull series. I must say they are brilliant. I’ve just read the latest, Wrath of the Warlock, and finished it in three nights. Love your characters in these books and can’t wait until summer when you bring the next episode of this amazing story. Love your work and really looking forward to your next book. Thanks, Scott. I am glad you’re enjoying the Dragonskull series and if all goes well, hopefully the next Dragonskull book, Dragon Skull: Doom of the Sorceress will be out in June, if all goes well.
Our next question is from James, who asks: How many words in all the books you’ve published? I had to look at some spreadsheets and figure it out, but I believe for all my 139 novels combined (Cloak of Dragonfire will be the 140th), I have published 12.1 million words of fiction in novel form. I don’t know how many in short stories, but assuming roughly 5,000 words is a short story for an average, it would probably be another half million. So we know for sure, it’s at least 12.1 million words of fiction I have published in novels in the last almost 12 years.
Our final comment of the week is from Wayne, who says: Concerning the fact that I’m going to be publishing Cloak of Dragonfire soon and I am hoping to write another Caina novel before the end of the year, Wayne says: This is great news, really looking forward to more of Nadia Moran’s adventures. I have read all the Cloak books and I’m on Ghost in the Ashes now. Nadia is my favorite, but gotta love Caina, really want to read the Cloak and Ghost novels, but not sure what point, so just wait until I finish The Ghost series. Thanks, Wayne. I’m glad you have enjoyed both Nadia and Caina’s adventures. About the Cloak and Ghost books, you can probably read those…if you read all the Cloak books, you can read those whenever you want. They’re not canon for either the Ghost series or the Cloak series. It’s just a little experiment I wrote, where I thought it’d be interesting if Caina met Nadia and they went off and had adventures together.
00:04:21 Facebook and Amazon ad results for March 2023
So let’s take a look back and see how my Facebook ads have performed for March 2023…actually, my Facebook ads and my Amazon ads. First up, let’s start off with Facebook ads. And here is what I got back for every dollar spent on Facebook ads. For the Frostborn series, for every dollar spent, I got back $4.74, which includes the audiobooks. For The Ghosts, for every dollar spent, I got back $3.05, and that also includes the audiobooks, especially the omnibus edition audiobooks. For Cloak Games and Cloak Mage, for every dollar I spent, I got back $2.06, and for Sevenfold Sword, for every dollar I spent, I got back $2.21, so that is a pretty good showing for Amazon ads this month. I need to take some time to adjust them, but I probably won’t have time to do that ‘til I finish the rough draft of Cloak of Dragonfire, which should hopefully be soon.
For Amazon ads (and these are for Amazon US ads), remember that for every six to eight clicks, you need a sale for the ad to work. If you’re not getting a sale for every six to eight clicks, then it is time to adjust the ads. But here is how we did on Amazon US. For Dragonskull: Sword of the Squire, I got one sale for 0.68 clicks and $8.43 back for every dollar I spent, which is just exceptional. Thank you for reading that, everyone. For Cloak of Dragons for every 3.38 clicks, I got a sale and I got back $1.73 for every dollar I spent, which is not as good as like, but much better than last month. And for the Demonsouled series, I sold a copy of Demonsouled Omnibus One for every 3.59 clicks and for the entire Demonsouled series as a whole (which I measured that way because it’s in Kindle Unlimited at the moment), I got back $3.64, which is the best that Demonsouled has ever done while I’ve been running it on Amazon ads. I was pleased enough with how the Amazon ads performed that I’m finally going to break down and try them properly for Amazon UK, so I will report back on how that goes next month.
00:06:24 Thoughts on the game Heroquest
Now, before we get into our main topic of generative AI, let’s talk about something fun: Heroquest the board game. Back at Thanksgiving 2022, my brother recommended I read Game Wizards by Jon Peterson, a business history of the early days of Dungeons and Dragons. It was an absolutely fascinating book, and I’ve talked about it many times on both my blog and my podcast. It’s honestly one of those books that I think every small businessperson should read, even small business people who wouldn’t touch Dungeons and Dragons with a 10,000 foot pole. There’s a lot of valuable business lessons to be learned from the various self-inflicted misfortunes of TSR. There is also quite a lot about related history in the book. One of them was that for a while in the 1980s while TSR was at its peak, a lot of people tried to jump on the fantasy gaming bandwagon with varying degrees of success. One of those companies was Milton Bradley, which teamed up with Games Workshop to release a board game called Heroquest, which distilled down Dungeons and Dragons to its dungeon crawling essentials. It also tied in Games Workshop’s Warhammer Universe. The game did quite well and numerous expansions were released through the late ‘80s and early to mid ‘90s. But eventually Milton Bradley stopped producing the game and it vanished into relative obscurity. Milton Bradley let the trademark lapse and then someone else bought it and it passed through a couple of other hands.
So we fast forward to 2020 when Hasbro brought the trademark and then in 2021 they brought the game back via crowdfunding campaign. Then in 2022 I heard about it for the first time in Game Wizards and then in January 2023, I was sufficiently intrigued to pick up a copy of the core game via Amazon since it happened to be on sale at the time. It’s quite an enjoyable game. I would describe it as a Dungeons and Dragons-lite. It’s boiled down to the essence of dungeon crawling, one my favorite genres of both games and literature without a lot of excess baggage or the tedium of listening to someone describing their tiefling bard’s 15,000 word tragic back story.
You have four characters with different abilities: the barbarian, the dwarf, the elf, and the wizard. You lead them on various quests against the forces of the evil wizard Zargon. The game’s board is modular and using the pieces of furniture included with the game, the board gets reconfigured into different dungeons for each mission. A good selection of monsters comes with the games: goblins, orcs, evil fish people people, a variety of undead and dread warriors which are renamed from Chaos Warriors in the original so Hasbro doesn’t get sued by the famously litigious Games Workshop. All the miniatures that come with the game are excellent and detailed though they are unpainted, more on that to come.
The handiest part of modern Heroquest is the companion app. Given the complexity of modern board games, quite a few of them now come with companion apps. Heroquest’s companion apps essentially runs the monsters for you, which is quite useful, though you still need to keep track of quite a bit of information, which you’ll want to do since gold coins and various treasures are supposed to carry over between quests. The barbarian looks formidable charging into battle bare chested, but he’s actually more formidable once you get the poor guy some armor. So, I thoroughly enjoy the game. I came across it as a middle-aged adult, but I found it as a kid that would go nuts over it. On the rare occasions I have a Saturday afternoon free, I’ll relax and run a dungeon. Each one of the dungeons only takes about an hour or so. Now, for an amusing aside: whenever I discuss Heroquest, people invariably suggest that I should get into miniature painting.
Once I started reading about Heroquest, I discovered that there is a large community dedicated to painting gaming miniatures, complete with YouTube channels devoted to the topic. It is very unlikely I will take this up as a hobby because I hate, hate painting. It is one of my least favorite homeowner chores, and I rather doubt that I would enjoy it even more with tiny brushes. Like, after I paint the porch for the summer, that’s totally what I want to do to relax: come inside and paint something else, but for fun. No, I’m just going to enjoy Heroquest with my unpainted miniatures, and whenever I talk about Heroquest, quite a few people mentioned that they too enjoy it. Reader JD wrote in to say: I grew up with the original Heroquest game. That was great. I had advanced Heroquest with its jigsaw like board pieces. I moved on to playing Dungeons and Dragons, but ended up buying both Heroquest games on eBay for a bit of nostalgia. You can get the older version on eBay for varying amounts of money depending on the condition, but the new game is available at…you can get it off Amazon pretty cheap. Well, not pretty cheap. It’s about $100, but sometimes it does go on sale and get it for a little cheaper than that. Another comment comes from Genelin (I think I pronounced it right), who says: I owned the original Heroquest board game as a kid, but I never actually got to play it because I didn’t know anyone who’s interested in stuff like that. As much as I’d love to try out the new one, the situation hasn’t changed. I’ve got nobody to play with. Well, the good news is that you don’t need to find people to play with, you can use the app. Because the app will run the game for you and then you control all four of the characters, so you no longer need to recruit people willing to put up with one’s interests in order to play Heroquest.
00:11:20 Main Topic: Recent Developments in Generative AI for Writers
Now on to our main topic this week: recent developments in generative AI for writers. If you’ve been listening to the show or reading my blog for a while, you know that I am generally very critical of generative AI for a variety of reasons that I’m going to recap shortly. But here is a sentence I’d never thought I would say, but it’s possible Adobe of all people might have done something to address my concerns about AI generated art. This is because before I started using Photoshop during COVID in 2020, my previous history with Adobe Products was long, troubled, and unhappy, including but not limited to: my computer crashing in 1994 the first time I tried to install Adobe Reader, many, many, many tech support difficulties related to Flash Player in the 2000s, many more tech support difficulties related to maintaining Adobe Creative Suite programs in a computer lab, and all the many times that Adobe Updater crashed while trying to install updates to update Adobe Updater. If you’ve lived through that, you know what I’m talking about.
If you listen to my previous podcast on the topic of AI Art, my chief concern is the ethics of it. I remain unconvinced since doing anything other than salami slicing plagiarism. The counter argument is that the AI isn’t copying, it’s learning patterns. My counter-argument to that is that a machine is incapable of learning, just applying a more refined formula that generates better results and it does that by copying a million different images and generating the average of them. That’s why if you go to the most of the AI image generators and type in a prompt like “Magic the Gathering plains cards,” it will generate an image as an average of all the Magic the Gathering playing cards in the data set, including weird symbols where the card would have text because a MGT playing card is statistically likely to have text in that portion of the card. It’s not creating a new image, it’s not learning anything, it is just averaging suitable images from its training data.
However, Adobe recently announced plans for a beta of an AI image generation tool they called Firefly. According to Adobe, their AI has only been trained on public domain images and stuff within the Adobe Stock Collection that they have the rights to use. If you’re thinking that public domain means the photos will be low quality, remember that every photograph taken by a U.S. Federal employee in the course of their duties is in the public domain. So there is a vast amount of high quality public domain images to use for training. Adobe also claims that it is working on the payment scheme for people whose images were used in the training. I mean, obviously this is not a perfect solution. Pirated stuff does sometimes turn up on Adobe Stock and well, everywhere else on the Internet. And it’s entirely possible Adobe is lying through their teeth in a “technically legally true” way about some of this.
That said, and this is a clever approach, most of Adobe’s customer base is image and video editing professionals, and they were the angriest about the potential abuses of AI art. Adobe doesn’t really have enough goodwill among its customer base to afford yet another viral uproar, but since all the big brains claim that AI is the next big thing, the company needs to compete in the space to survive. So it was a smart move to let others rush into that space first and then announce their own solutions that addresses some of those concerns. Anyway, I signed up for the beta of Firefly, and I actually just got in right before I started recording this. I will test it out and share my thoughts next week. If it does work and the promises are true, Firefly is the sort of general generative AI product I feel I could probably use in good conscience.
What I’d most likely use it for would be to generate assets I would include in my book covers and Facebook ad images. I’ve already used various stock photo sites in DAS 3D for that, so it would be another tool in the toolkit. Doubtless fiddling with the generator to find the exact correct prompt would be no different than scrolling endlessly through stock photo results or trying to get another rendering in DAS to make the output look right. On a related note, a reader asked if I was considering using generative text AI like ChatGPT to help with my writing output. Short answer? Absolutely not! Longer answer: Reason number one, at this point it would feel like committing fraud and cheating readers. Like, if you buy a book that says Jonathan Moeller on the cover, the implicit promise is that Jonathan Moeller wrote the book, and if you’re buying that book, you want it to be written by Jonathan Moeller. That means you’ve decided you like my writing, warts and all (and I appreciate that) and have decided to pay actual money for it.
So, buying a Jonathan Moeller book that was actually written by an AI is a bit like a restaurant offering French fries that are allegedly 100% potato but actually turned out to be tofu, soybeans, and sawdust. Granted, if I did write a book with AI and the cover said by Jonathan Moeller and Chat GPT, that would be different, but I don’t want to do that anymore than I want to eat a French fry made out of sawdust. Reason two: generative AI still kind of sucks. I have to admit that when I listen to people who are very impressed with and enthusiastic about AI, I kind of wonder about their credulity. It’s like a wizard cast a spell on horse manure to make people think it was the finest pepperoni pizza and people are wolfing down this horse manure and praising the flavor while the horse looks on in bemusement.
A lot of excitement has been generated from the fact that you can tell ChatGPT to write text in the style of Sir Arthur Conan Doyle or whatever other author you want. But what comes out when you use a prompt like that tends not to be very good, especially when compared to the original. Granted, in certain applications, horse manure is actually highly valuable. It’s just not pepperoni pizza. In past blog posts, I’ve joked that generative AI is the infinite crap generator. Maybe it’s more accurate to say that generative AI is the most likely random crap generator, but occasionally not, and you have to iterate a whole lot with your prompt and results to find something that isn’t crap. You can use it to write things, but then you have to wade through a lot of crap to get there first and then edit it and then put it all together, at which point you might as well just write it yourself for less hassle.
And that brings us to the third reason: copyright. As in this writing, the copyright situation around generative AI is highly unsettled, but trending towards negative. The official position of the US Copyright Office is that AI generated images can’t be copyrighted, though legislation might end up changing that at some point. Someone tried to copyright an AI-generated comic book and the copyright for that was revoked and then revisited to say to that while the image themselves could be copyrighted, the arrangement in human written text could. And this case is ongoing, with the creator of the comic book and her lawyers going back and forth with the Copyright Office in the US about what is copyrighted in her AI generated comic book.
There’s also the ongoing lawsuit Getty Images has against several of the AI models, and that no one can predict at this point how that’s going to all settle out. So overall, the opinion of the current authorities is that machine generated art cannot be copyrighted, and presumably that is true with machine generated text. So for those three reasons, I won’t be attempting an AI written book anytime soon. Perhaps in another ten years, writing a novel will involve typing like 1000 different scene prompts into an AI generator and then editing the various outputs together. That said, I really doubt it, though.
So that is it for this week. Thank you for listening to The Pulp Writer Show. I hope you found the show useful. If you enjoyed the podcast, please leave your review on your podcasting platform of choice. It really does help. Until then, stay safe and stay healthy and see you all next week.
