AI in TTRPG Creators: To Use or Not to Use?
AI is here if we like it or not, but what it mean for us TTRPG Creators, should we use or not? Should we be afraid of backlash? Will customers vote with their wallets? Let's analyse this topic!
Ah, yes, an AI art-related topic, something I've wanted to discuss for a while. It's a subject that might make me lose followers, but as they say, "You only live once!" Let's go!
I'm not sure if I can answer the first question that comes to mind: "Is AI moral?" Simply put, I don't know. On one hand, I understand artists who are upset about their art being used to train AI without their consent. On the other hand, there's the argument that this is how artists learn, by copying others' work. It has some merit, but it's not entirely correct. Artists do look at others' work and try to replicate it, but they do so in their own style. It's also incorrect to claim that AI art is stealing. Copyright law gives artists the right to stop others from copying their work. The issue with AI is that it doesn't copy an artist's work; it tries to reproduce their style. When we compare AI-generated content with an original artist's work, it's challenging to find similarities. The styles may be very similar, but art style cannot be copyrighted.
I will take the artists' side on this and say that AI should not be trained on artists' work without their consent or compensation. The morality of training AI is in a grey area and needs improvement.
I've been using AI for my private campaigns, something I don't charge people for, something I do for myself. However, for some reason, it seems more acceptable to use art found on Google without the artist's consent than to use AI for your own purposes. When I mentioned this on Twitter, I was called a "thief." When I tried to discuss using AI in an adventure module, I received direct messages filled with hate and inappropriate language. Not a single person who "attacked" me was an artist. In fact, one of those Twitter accounts had NSFW content filled with AI art. As soon as I pointed that out, I was blocked. Even by mentioning AI art, people are being attacked by "knights in shining armor." However, not a single artist attacked me. Artists usually tend to explain why they feel it's immoral, and I appreciate that.
I've been researching and gathering knowledge about using AI in TTRPG products recently. As some of you may know, I'm working on my first adventure module with a non-existent budget, and AI is one of the options on the table. Therefore, I wanted to do research about it and share my findings with all of you.
My first steps while gathering data were small polls on my Twitter account, which yielded some surprising results. The very first question I asked was:
As you can see, over 55% of 616 votes were fine with me using AI art in the adventure module. However, 80% of the comments were against it. I did receive DMs from two people saying they are fine with AI art, but they'd rather not say that publicly. At this point, I was not sure what to think about the results. Yes, the results were clear, but the difference between comments and votes was significant. There were 114 comments, with the majority suggesting either no art at all, royalty-free stock art, or hiring an artist. A few comments echoed sentiments from my DMs, saying things like, "Not going to argue the AI art thing again. I am exhausted from being called nasty things by artists." And "A friend of mine got criticized by the masses for using AI in a free 5e supplement," indicating that there is significant backlash against people who use AI in their work. Again, I wish to underline that not a single artist in the entire conversation was rude or aggressive, and people who were had no mention of being artists in their Twitter profiles.
But I needed more data. I stated in the first question that I was not planning to sell the module, but I was wondering how the poll results would change if I asked about monetizing material with AI art. In my second poll, I asked:
This was a huge surprise. Over 12% of all votes were in favor of big companies like Wizards of the Coast and Paizo using AI-generated art. Over 8% were in favor of small/medium companies that release their books on Kickstarter, like Kobold Press. 38% of votes were okay with first-time creators using AI art, and over 41% of votes were against using AI in paid materials at all. I was not expecting over 12% of votes to be in favor of Wizards of the Coast using AI art. It was a revelation.
How do these two graphs compare to each other? In the first one, 55.2% of votes were in favor of AI, and 44.8% were against using AI. In the second graph, 58.5% of votes were in favor of AI, and 41.5% were against using AI. I think it's safe to say that AI is splitting customers down the middle, with those against AI being more vocal.
I didn't stop my research there, but before proceeding, I would like to address some comments from those polls. The most common suggestion and solution was to use royalty-free or low-priced stock art. I had never considered that option, so I checked some bigger sites like Adobe Stock, Shutterstock, and iStockPhoto. I was shocked to find that these sites, which offer paid content, are filled with AI art. Adobe Stock at least discloses that the art was AI-generated, but Shutterstock doesn't even do that. After some more investigation, I found that Adobe Stock allows you to turn off the search for AI-generated art, but there is still AI-generated art that is not marked as such. There is no guarantee that you can acquire stock art of fantasy art without being accused of using AI art down the line. In that case, why take the risk? Wouldn't it be better to generate the AI art you want, exactly as you want it, and then disclose that the art on the cover was AI-generated? At least you would be honest about it. As an author of such a module, you can be 100% honest and pay for an artist, but the person who provided you with art can just say, "I made this," when in fact it was made by AI. Guess who will get all the blame if someone identifies the art you paid an artist for as AI-generated? Exactly, the author of the book.
Another interesting comment that came up under those polls was "Don't use art at all," and to be honest, that is quite good advice. You can have a very nice book with no art in it, as many have done in the past with success. The problem now is that the market is oversaturated, and you need to compete with other creators. If I know anything about the market, it's that "crap covered in a gold wrapper" will sell better than a "gold nugget covered in crap." Customers are more likely to pick a book with a nice-looking cover, and you can't judge what's inside the book until you buy it. Having no art in a book might affect the sales of a module, especially if you are debuting your work and your name is not associated with quality content.
Let's talk about the last type of comments: hiring an artist. First, I'd like to mention artists who specialize in TTRPG art. I reached out to four of them and received a quote of $500-$600 for a cover and $300-$400 per character portrait (without a background). My module requires a cover and nine character portraits, which would cost between $3,200 and $4,200. For an adventure module that I was planning to give away for free, we can argue that an artist may or may not lower the price if all the art is ordered from them. However, that is something that new creators making their very first book may not be able to afford without taking a loan, which may or may not pay itself back. But those are professional artists specializing in TTRPG art. There are cheaper options, right? As soon as I mentioned working on my book, I was flooded with offers from artists, each being polite and showcasing their art. When I asked about the price, I received a price range of $100-$150 per piece of art (without disclosing whether it's for a cover or a portrait). Quick math puts the price range at $1,000-$1,500, which is still a significant amount of money but might be more affordable than a professional artist. Some of them even offer monthly payments. However, be cautious, as some of them seem sketchy, offering art that is inconsistent, created by different people, with accounts that have a small number of followers and were created a few weeks prior. The chances of being scammed are quite high. On the other hand, there are artists with impressive skills charging low fees, and I made a friend or two while talking with them.
There are artists with amazing skills who are just starting out and don't charge much. Here, I would like to give a shout-out to Sonya (@Sonya100years), an artist who provides high-quality art at an affordable price. Please follow her, as she deserves all the love!
To conclude, new creators often do not have access to the funds that would justify the cost of hiring an artist, especially if they are creating their first book. It's essential to understand that game masters/dungeon masters often take on most of the financial burden of running the game by buying books and providing tools for players. These same game masters/dungeon masters are more likely to create TTRPG content, often working full-time and spending late hours to finish their first work, just to be able to sell it for $5, split with a shop. These are the realities of first-time TTRPG creators, and to pay for art, they would need to sell at least 500 copies, or 1,000-1,500 copies if they wanted to pay for professional artists. It seems highly unlikely that they could cover these expenses, and I found some data on the matter.
DriveThruRPG sorts its products by "metals," with various thresholds for book sales, from copper to platinum.
So, what does this data show us? It shows that nearly 60% of all creations never sold more than 50 times. Take into consideration that this data represents the entire sales cycle, so not many sold at the premiere. Less than 5% of products sell more than 500 copies, and less than 2% sell 1,000 copies. What are the chances that a product made and published by a new creator will cover the expenses of hiring an artist? We can safely say it's less than 1%.
Why am I explaining all of this? Is it a justification for shady practices when AI is being used? No, it's context that TTRPG creators need to consider when creating their first book or module. It's essential to understand why TTRPG creators even consider the option of using AI. Here, I would like to offer advice to new, aspiring TTRPG creators: do not, under any circumstances, take out a loan to pay for art in your book. There are other options, but do not burden yourself and your family with debt that probably will not be paid off by your work.
Before we move to conclusions, I would like to bring one more thing to the table. Many comments in the two initial polls mentioned that a product with AI art will not sell. The polls show that it will sell, but probably in lower amounts because there is a group of people who will not buy AI art. However, that does not mean the project will fail. The best example is a story from a Kickstarter campaign that happened a few months ago.
Terraforming Mars was designed by Jacob Fryxelius and published by FryxGames in 2016. On September 6, 2023, a Kickstarter campaign for the next expansion, "Prelude," started. Kickstarter demands transparency from projects, so in the "Prelude" Kickstarter campaign notes, we can read the following statement:
"We have and will continue to leverage AI-generated content in the development and delivery of this project. We have used MidJourney, Fotor, and the Adobe Suite of products as tools in conjunction with our internal and external illustrators, graphic designers, and marketers to generate ideas, concepts, illustrations, graphic design elements, and marketing materials across all the elements of this game. AI and other automation tools are integrated into our company, and while all the components of this game have a mix of human and AI-generated content, nothing is solely generated by AI. We also work with a number of partners to produce and deliver the rewards for this project. Those partners may also use AI-generated content in their production and delivery process, as well as in their messaging, marketing, financial management, human resources, systems development, and other internal and external business processes."
Big news sites like Polygon, Dicebreaker, BoardGameGeek, and many others picked up on this, and soon, it was widely discussed. Reddit was filled with angry posts, and Twitter was flooded with angry messages. The Kickstarter target was $10,000, but the campaign ended with $2,222,703, over 2 million dollars. Did people vote with their wallets? Or is it possible that AI art is not as significant a factor in sales as some may think?
It's worth mentioning that FryxGames has promised that no AI will be used in their next expansion. If they were successful, why change their approach? What could have persuaded the company to change its ways? The answer, as always, is money. Let's use our poll votes as an example. They show that 50% of people are against AI. How far-fetched would it be to think that FryxGames could have made an extra $4 million by not using AI instead of real artists? Probably not that high. Let's lower the amount to 10%, which means they could make $200,000 more by not using AI. Even if it's just 5%, that's an extra $100,000. How much would artists cost them? A fraction of that price.
The example of Terraforming Mars shows us that a project with AI art can be successful, but not using AI can generate even bigger revenue. Another thing about AI-generated art is that there's no specific law for it. A US court has already ruled that AI art cannot be copyrighted, meaning anyone can take AI art from your book and use it as they wish, even copying the cover, which can create confusion for customers. The future of AI is hard to predict, with big companies like Microsoft, Google, OpenAI, and others supporting it, and we know that politicians can be influenced by these companies. But AI laws may be passed, and you might have to pay royalties to artists for using AI. You might even be forced to take your book down. The fact is, we don't know how AI art and AI, in general, will be regulated. Creating a book with AI art is a big risk.
In the entirety of the conversation under the polls, one specific reply struck me the most. Gregory Wunderlin (@gwritespretty), an amazing writer and author of "The Soul of Chaos," said, and I quote:
"Honestly, as a creator, the problem for me comes from the lack of supporting other creators."
That one sentence struck me more than anything. My entire journey has been about being supported by others, and I've said that I want to support other creators, writers, and artists. How could I be myself if I would betray my beliefs?
This blog post is already quite long, but I needed to say all of that. Now I need to conclude with this statement. At this point in time, AI art in your book can potentially improve its sales, but your work will sell better without AI art. Additionally, AI art exists in a legal gray area, a kind of "wild west" that will eventually be regulated by law, but until those regulations are in place, you're taking a risk by using AI art in your work.
My conclusion is this: I don't mind using AI art in my work, as I think it can look nice and fit well. However, I will NOT use AI art in my books because it goes against the values of Elven Maid Inn. I will continue to promote artists, writers, and creators, even those who use AI in their work, but I will not use it myself until it's regulated by law.