The Future of Game Writing: Will AI Take Over Your D&D Adventures?
Exploring the rise of AI in the world of role-playing games and what it means for small-time game creators
I’ve written several articles about Dungeons and Dragons (D&D) and some of its alternatives. After our six-month Tomb of Horrors campaign ended as expected (80% of us died), the group decided to take a break over the holidays and regroup in 2024.
A few of us discussed trying out a different kind of role-playing game. We tossed around ideas like a Zombie Apocalypse or Cyberpunk setting. This got me into research mode over the holidays, diving into different games and generic role-playing systems to find something fun for us to play.
I went down the metaphorical rabbit hole and got deep into the theory of generic role-playing systems, trying to understand how they work. I discovered FUDGE and GURPS — both well-documented, tested systems that can be used for almost any game style or setting.
This was exciting. I could dream up a Wild West-style game or a post-Climate Change apocalypse adventure. These generic systems were flexible and made sense, especially with their z-based scoring methods for abilities and skills.
But despite all the possibilities, I kept coming back to a simple truth: no matter what game system you use — generic or not — the most important thing is gathering with friends and agreeing on rules that everyone can live with. It could be as simple as knowing which poker hands beat others or as complex as DnD 5th Edition.
Writing D&D Games
At the start of 2023, I challenged myself to write a few short D&D adventure modules. My goal was to complete 12 modules, but I managed to finish eight. Some were D&D modules, while others were world settings, like the World of Amum. I put them all on my Ko-fi shop page, making them free to download because I believe in open-source gaming.
I wasn’t bothered by not hitting my goal of 12 modules. What I learned along the way was more valuable. I realized that while I was passionate about writing these games, the demand for detailed game mechanics and system understanding was a challenge.
Writing D&D games is a completely different experience from playing or running them. I made it even harder by trying to make each module game system neutral. My focus was on the adventure and the story, aiming for games that could be completed in one sitting.
Of course, this unraveled into a big mess. During playtesting, the holes in the story and game mechanics became glaringly obvious.
For example, in my adventure Survive the Silverwall Pass, I created a backstory about “snow elves” and a special arrow that could incapacitate Frost Giants. But during gameplay, my friends were more concerned with making the Frost Giants scarier and more imposing since they were the game’s focal point.
Playtesting is humbling, and I wouldn’t skip it. Ironically, writing these modules reminded me of product launches in the startup world — just when you think you’re ready to go, you realize there are loose ends to tie up.
You Need a Hook
One key lesson I learned from playtesting is the importance of having at least two, but preferably three, adventure hooks. These hooks get your players interested — whether it’s finding lost treasure or saving a princess from a high tower.
Hooks keep your experienced (and sometimes jaded) players excited. They build the story, and storytelling is crucial to making any game a success, especially in role-playing games.
For example, my latest module (still a work in progress) has three hooks:
- The characters are hired by a local historian obsessed with solving the mystery of the Montmort family and recovering valuable artifacts from the mansion.
- The characters stumble upon a treasure map mentioned by treasure hunters in a tavern or marketplace, intrigued by the prospect of finding hidden riches.
- A local lord or lady tasks the characters with investigating the undead presence at the mansion, fearing it might spread and threaten their lands.
If my players don’t feel hooked by these, I’ll know it’s time to go back to the drawing board. Writing games is different from writing essays or poetry, where you write for yourself. Writing games is about your players. You have to create something that resonates with your audience and meets their expectations.
The Curious Problem of GPT
Large language models like ChatGPT are making an impact in the gaming world. For game creators, this is both exciting and terrifying.
It’s exciting because AI can act as an assistant. It can outline a game for you or analyze your creation for loopholes. But it’s terrifying because it can write an entire D&D module in seconds, cutting creators out of the process. Will small-time game creators become obsolete in a post-GPT world? The answer is yes, it’s a real possibility as GPT technology improves.
I experimented with some of OpenAI’s writing assistants and found them far superior to the basic ChatGPT-4. Why? Because they seem fine-tuned on specific data sets, giving them a unique voice. OpenAI’s marketplace alone should send a chill down the spine of any game creator.
Likely, large corporations like Hasbro are already looking at customized GPT systems to automate game writing. These AI systems don’t need salaries or sick days. All they need is hardware and electricity — both tax-deductible. Human writers? Messy and expensive by comparison.
I suspect the trend of layoffs in the gaming industry will only accelerate in 2024 as corporations “fine-tune” their staffing and technology.
What About Small-Time Creators?
I admit, I tested GPT for my latest module but abandoned it quickly. It was generating names and places that felt eerily familiar. It made me uneasy because I couldn’t shake the feeling that the training data included copyrighted game worlds. With lawsuits like the NY Times vs. OpenAI looming, you have to wonder where all this data came from.
I support open-source gaming, but I don’t support stealing. I don’t support taking someone’s hard work, repackaging it, and selling it without giving credit. That’s always been a weakness of the open-source model, and it only works if everyone — including big corporations — pays into the system.
Writing Games in 2024 and Beyond
Writing games is tough, no question about it. I admire small-time game creators who are both writers and artists. I’ve seen some incredible teams create fun games with amazing artwork. But the sad reality is that many of these small teams won’t survive. It’s hard to make a living, and without enough revenue, many of them will close up shop.
Some will make it, while others won’t. The trick for success in a GPT world will be for game creators to stay rooted in their values. As the market evolves, I think we’ll see a divide between mass-produced games and authentic ones — those written without the use of AI.
I believe we’ll start seeing disclaimers like “Not written with GPT” on new games, especially those offering free D&D modules with non-commercial licenses. Small creators will need to protect themselves more than ever if they want to carve out a space in this changing market.
Yes, the industry is going to change, but one thing will always remain: everyone loves a good game.
So, start writing one.
Follow Me
