
Building ChatRPG an interactive choose your own adventure game with PartyRock
Learn how we can use LLMs to build interactive experiences with no code.
Super Mario RPG
and yes, I'm as excited as you are about the remake! Series like Final Fantasy
and Elder Scrolls
further fueled my interest. While these games provided expansive, predetermined worlds, it was the agency and narrative control of those early interactive books that stayed with me. This blend of experiences made me wonder, could I build something that merges the depth of modern RPGs with the open-ended narratives of text-based adventures?

- Chat Box with the DM Role: Analogous to the initial trigger in a step function, this is the player's primary interaction point. As they input responses or choices, the AI—serving in the Dungeon Master (DM) role—processes this input, crafting narrative responses. Each interaction determines the subsequent flow, much like a lambda function reacting to specific input parameters.
- The Summarize Function: Given the extensive narrative arcs, it's crucial to periodically distill the essence. This function operates similarly to an intermediary lambda step, taking the accumulated chat history, processing it, and generating a concise summary representing the current narrative landscape.
- Image Generation Prompt from Summary: This step can be likened to a function converting data into actionable insights. Harnessing the generated summary, it crafts a precise prompt tailored for an image generation model, converting textual narrative into a request for visual representation.
- Displaying the Image: The final step in our narrative step function sequence. Post-processing and visualization are crucial in data workflows, and in ChatRPG, this translates to presenting players with a visual snapshot of their current in-game scenario.

Here are some examples builders are already using AI to streamline game development:
Any opinions in this post are those of the individual author and may not reflect the opinions of AWS.