Clevis Logo
ExamplesUse CasesPricing

Use Case

Automated video game level design

Generative AI models, like Language Models (LLMs) and image generators, provide opportunities for automated video game level design. Firstly, they could auto-generate textual descriptions using ChatGPT to create game lore, NPC dialogs, and quest narratives facilitating immersive gameplay. Secondly, image generators trained on pixel-style video game images could generate new game levels or characters, introducing novelty in each session. Further, combining these models to output both graphical elements and relevant textual content provides a unified, dynamic game universe. Finally, algorithm-customized level generation tools powered by these models can adjust difficulty levels or modify environmental aesthetics based on player's progress, preferences or feedback. Hence, the opportunities are profound, with vast potential to revolutionize video game design as well as player experience.

How to build with Clevis

Text Input

Input desired level size and difficulty

HTTP Request

Fetch data about existing video game levels for inspiration

Prompt ChatGPT

Prompt ChatGPT to generate a list of game elements and obstacles

Prompt ChatGPT

Prompt ChatGPT with the output from the previous step to generate level layout ideas

Display Output

Display the final generated level layout to the user

This is an illustrative app example you can build using Clevis, an efficient tool to develop advanced AI applications. This specific application is called 'Automated Video Game Level Design', and it leverages AI to generate innovative video game levels.

The process starts when the user inputs the desired level size and difficulty. This action is described in the Text Input step. This input gives the application a basis on what the user needs.

Next is the Http Request step, where the application fetches data about existing video game levels that can serve as inspiration for the new level it will generate.

After acquiring the necessary data, the application then taps into the power of ChatGPT, OpenAI's language model. In the first ChatGPT step, it prompts this AI model to generate a list of game elements and obstacles that are suitable for the specified level size and difficulty.

The second ChatGPT step takes this further, using the output from the previous step to stimulate ChatGPT to come up with creative level layout ideas.

Finally, the application displays the final generated level layout to the user in the Display Output step. This entire process signifies how you can build a similar app for game design or other applications using the Clevis tool.


© 2024 Clevis. All Rights reserved