The use of generative AI in game development has sparked growing controversy within the gaming industry and player communities. On one hand, AI tools can accelerate production by generating dialogue, character models, landscapes, and sometimes even entire quests. This can save developers time and resources, allowing for faster prototyping and expanded creative possibilities. Indie developers, in particular, may benefit from these tools to build ambitious games with smaller teams.
However, critics argue that AI-generated content risks diluting creativity and originality. Writers, artists, and voice actors have raised concerns that their roles could be replaced or undervalued. Some also worry about the ethical use of training data, AI models are often built on vast datasets that may include copyrighted or fan-created content without proper attribution or consent.
Players have voiced skepticism too, questioning whether AI-generated narratives or art can truly match the depth and soul of human-crafted stories. And there’s also concern about overreliance on automation, potentially leading to bland or formulaic game design.
While generative AI holds promise as a tool for innovation, its integration into game development requires careful consideration. Transparency, fair labor practices, and creative integrity must remain priorities as the industry navigates this new frontier.
We also experiment with generative AI. And we found out one thing AI definitely can’t do (thankfully?) is generate convincing pixel art. It completely misses the mark there. But when it comes to early-stage concept art, it can actually be quite handy. Feeding a character description or a scene into an AI model often results in a rough visual that helps our artists get a feel for the mood or setting. None of that AI-generated imagery ends up in the game, it’s just a jumping-off point, a creative prompt for our actual human talent.
For instance, every quest-giving NPC in our world has a written character description. Feeding that into an AI to generate a basic portrait gives our artists something to build on, accelerating the concepting stage. It’s not about replacing anyone, it’s more like giving our creative team a sketchpad that draws back.
We’ve also played around with AI for NPC dialogue. Not the critical, plot-relevant conversations, but just those little ambient exchanges that make the world feel alive. The results are… interesting. Technically, the language is fine, often grammatically correct and even contextually appropriate, but something always feels off. There’s a certain spark missing, a human touch that’s hard to describe. Even in blind tests, the AI-generated lines stand out as subtly hollow. It just doesn’t feel the same.
Then there’s coding. Not the main focus of this post, but it’s worth mentioning because here, AI really shines. Using AI as a co-pilot during programming has been a game-changer. It speeds up development, boosts productivity, and often provides spot-on suggestions. It’s clear that most models have seen a massive amount of code during training.
That said, it’s far from perfect. Sometimes the suggestions are hilariously off, like a GPS confidently telling you to drive straight into a lake. But the upside is undeniable. As a programmer, I’m not even slightly worried about being replaced, because AI still needs close supervision and constant double-checking. But let’s not get sidetracked, maybe this topic even deserves its own post. Back to storytelling with AI.
Should AI Help With How You Tell a Story?
The idea that AI could assist with how you tell a story, but not what you tell, is at the center of an ongoing creative debate. Supporters argue this distinction preserves human originality while using AI as a supportive tool. For example, a writer may know the core message or plot they want to convey but use AI to refine dialogue, suggest structure, or improve pacing. In this way, AI acts like an editor or co-writer enhancing the storytelling craft without replacing the storyteller’s voice.
On the other hand, some argue the line between how and what is blurrier than it seems. When AI proposes character arcs, alternate endings, or even tone shifts, it inevitably influences the story’s substance. This raises the question: is it still the writer’s story, or a collaboration with the machine?
Critics also warn that overreliance on AI in the how can lead to homogenized storytelling, shaped by patterns found in training data rather than genuine inspiration. Still, if used intentionally, AI can empower creators without overshadowing them.
Ultimately, the balance lies in control. When the creator decides the what, and uses AI to hint thoughtfully for the how, the result can be both efficient and authentic.
Conclusion
While generative AI has made impressive strides, it’s far from being able to create entire stories, games, or artworks on its own at a truly human level. Right now, it lacks deep understanding, emotional nuance, and true originality. The outputs often need heavy editing, fact-checking, or creative refinement to meet professional standards. It may one day evolve into a more autonomous creative force, but for now, it works best as a collaborator assisting, not replacing, human creators. The technology is a tool, not a storyteller. Relying on it entirely just doesn’t produce results that feel meaningful, consistent, or truly inspired.
Note: This blog post is an example of how AI can be used as a co-writer and editor. Since English isn’t my native language, I used AI to help structure the text and improve clarity. However, the message and ideas are still entirely my own. The voice is still mine, just expressed more clearly with a bit of assistance.