Like many video game hobbyists, I’ve had dreams of making a game of my own. Before I was even in high school, I had a binder full of designs for a game I had dreamt up, but I never really got the means to even attempt to make it. Programming has always bored me, and I tend to get frustrated with art. Even attempts at small projects to just motivate myself with some visible progress fizzled.
The closest I got was a brief stint writing for a game company. I co-wrote a game that is still unannounced and may, at this point, be canceled. I’m not sure. However, I got the gig at the lowest point of my life, and it gave me something to focus on. I’ve always wanted to return to doing it, but again, there’s a wall I have to climb over to make any progress.
Jordan Weisman is among those looking to make game development more accessible. Weisman was one of the founders of Virtual Worlds Entertainment and FASA Interactive, responsible for the creation of the Battletech and Shadowrun series. He later returned to Shadowrun when he created Harebrained Schemes to help with the creation of Shadowrun Returns and the sequels that followed. After that was acquired by Paradox, he moved on. Now, his latest endeavor, Endless Adventures Inc., is set to release Adventure Forge later this year. Promising “No-code, no Limits, only imagination required,” it’s a platform to allow anyone to get their feet wet with game design.
Mr. Weisman and his business partner, David Reid, sat with me to discuss the upcoming toolset, as well as address my concerns about its use of generative AI.
The goal of Adventure Forge is “to create a toolset that allows storytellers, writers, artists, and designers to create narratively-focused video games without any coding or scripting necessary,” Weisman explained. They’ve spent “many years of tinkering trying to find authoring paradigms that enabled very rich, sophisticated logic for narrative games.” That led them to “a system of highly contextual, auto-populated drop-down menus so that you can create very sophisticated logic but never get syntax wrong.”
Weismann explained that the driving force behind it was years of creating interactive experiences, including both digital and tabletop games. “The part that brings the biggest smiles to my face are the stories that players told me, rather than the ones I told them,” he explained. “In video games, I never really felt that we were able to offer that level of creativity to players. To people who wanted to tell stories and make games.”
He further elaborated and said that he hoped to see more diverse voices making games and telling stories. “For the most part, we’re still in an industry where it’s old white guys like me who are telling stories. The entire universe plays games, and the entire universe is not old and white.”
“My other hope is that we also see a lot more innovation in narrative game design. Even those of us who have the privilege of having development teams available to us to create games, we don’t get that much time to really play with new ideas because it’s expensive. I write a design doc, send it to engineering, engineering sends it back to me, it sucks (because everything I design sucks the first ten times). That loop is slow and expensive. Your time for real innovation is relatively narrow. So, selfishly I also wanted a tool that allowed me to just screw around and try different things. Try to shorten that loop.”
The plan for Adventure Forge is that it won’t just be a toolset but also a platform. In order to share and play content, you’ll need the application itself. Its target platforms are Windows, Mac, iOS, Android, and Chrome.
The reason for the limitation is, as Weisman puts it, “Sometimes art, specifically narrative art, isn’t meant to be shared with the whole world. It might be a very personal story.” So, the platform will allow you to control how it’s shared, whether it’s with the whole community or just a select group of friends.
Currently, the plan isn’t to allow you to publish your creations externally from the app. However, Weisman says, “The premise is to see what games start getting traction with people and then reach out to those creators and work with them to move it into a standalone application that will be available in app stores to expose their great work to an even larger audience.
All this empowerment for solo developers comes at the price of Adventure Forge’s integrated generative AI. Behind those contextual menus is an LLM, and it goes beyond that. One of the touted features of the platform is that if you can’t find a premade asset for something you need within the extensive list of items provided, you can have the AI create one for you, replicating the art style and perspective. It can also generate written text based on the context given, so it might handle the environment descriptions for you.
AI – or rather, what industry backers are pushing as AI – is an extremely nuanced topic of debate. Procedural creation has been part of the industry since the beginning before video games even migrated off of huge mainframe computers. Game engines have long integrated tools to make certain processes of game design easier or even automated. Generative AI might be seen as the next step in this, or even a step too far. There are facets of it leaking into every form of media, and creative people are rightfully afraid of being devalued further in a world where people already seem to believe art should be free.
“The way we view generative content creation is as an author-facing tool, not player-facing content,” Weisman explains. “For us, the most important thing is that the designer/artist/writer/creator of the experience is in control of the experience. That they can guarantee that their creative intent is coming across to the player.” Further, he explains that attempts to use chatbots to supplement the characters and story just don’t provide the desired result. A carrot that is currently being chased by some other big publishers.
Instead, you can get writing prompts from the LLM based on what you’ve written previously. Weisman says, “It’s just text for you to manipulate. For myself, maybe 5% of the actual words that the language model made may end up in something I’m working on. But often, I find that it’s blathering of some tropes out will inspire me quicker than if I were just staring at the blank prompt.” He describes it as “the super-next generation of predictive text that we’ve had in our word generators and our emails for quite some time.”
Using generative AI is a bit of a tightrope walk, to put it lightly. The crew at Endless Adventures are being careful with how they approach it, but I still have conflicted feelings on the matter. It’s important to keep in mind that learning to be a traditional solo developer is not impossible. There are a lot of people from different backgrounds who have taken the dive and met with success. The truly hard part of the ordeal is getting your work noticed. This would be even more difficult to navigate if marketplaces were packed full of quick, low-effort AI-generated projects.
“I think there are many axises to this challenge. To me, even in a big team, players want to play something created by a human that has the thought process of a human,” said Weismann. “The human condition that they’re bringing to that story that makes it interesting and relatable. To me, it’s a matter of if they can use those tools in a controlled enough environment to get that creative intent across in a real and cohesive way and develop a style that is going to make each game look unique. Because games need to look unique to stand out in the marketplace. You can’t have your game look like everyone else’s.”
David Reid added, “AI is really good at looking backward and building things that humans have thought of doing already. But if you make a game solely on AI, the only things that will play it is AI. Humans will find them very dull and uninteresting. Really, what we’re trying to do with Adventure Forge is unleashing the creativity of people who have been unable to turn the stories and ideas they have into full-fledged interactive games. AI is a tool for that.”
“A lot of entertainment is pattern-driven,” Weisman explains. “Patterns will always have a place in entertainment. One of the things when I lecture that I talk about with designers is that you have to establish the familiar in order for your audience to appreciate the exotic. And the familiar is patterns. Things that we’re used to in our world and our own life experience. And then the human element is how to find the interesting twist. That’s the part that I don’t see a computer coming up with soon.”
That’s pretty optimistic. In my experience, humans have an insatiable appetite for the familiar that seems to get more and more overwhelming as the years go by. There’s often a pop where we get tired, and suddenly, super-hero movies aren’t getting the same traction, but I don’t think that results in seeking something unique. There’s usually a holding pattern while we seek the next trend that the masses cling to.
Generative AI and LLM are inevitably going to find a place. If we’re lucky, it will be as Endless Adventures envisions it, and creative people will be empowered to create more. The general fear is that there will be fewer places where creativity has value because AI has filled them all faster, cheaper, and without complaint. It’s hard to have your voice heard over the din of so many robots.
In the Shadowrun universe, one of the first sentient AIs developed took over a large arcology, locked the doors, and killed most of the inhabitants in what’s known as the Renraku Arcology Shutdown. It’s a good read, but with the way AI has been applied to non-sentient programs that browse data to generate content based on patterns, it’s more likely that our world’s AI is just going to suck all the flavor off the bones of the world than destroy us all. Such a shame.
There’s a lot more to discuss on generative AI, but I find the whole debate so tiresome. From this conversation, I’m at least convinced that Jordan Weismann is doing what he’s always done; attempting to empower people to tell stories. There doesn’t seem to be anything malicious about Adventure Forge’s approach, and he’s spent a good amount of time considering the potential issues of AI and weighing them. He sees value in pursuing it.
I’m generally on the mixed-to-negative side. As I’ve said, games have used procedurally generated content as far back as 1980’s Rogue. The first two games in the Elder Scrolls series used it to craft their worlds (which were so terrible that they never did it again after that). So, maybe it’s, as Weisman says, and that as long as there’s a human behind the wheel, generative AI can be more of a tool than a crappier substitute for creativity.
At the time of this interview, Adventure Forge was expecting to enter its beta period this Fall. It could still happen, but I haven’t heard anything since, so we’ll have to wait and see.
Published: Nov 21, 2023 04:00 pm