Summary of Game Development As Human-llm Interaction, by Jiale Hong et al.
Game Development as Human-LLM Interaction
by Jiale Hong, Hongqiu Wu, Hai Zhao
First submitted to arxiv on: 18 Aug 2024
Categories
- Main: Artificial Intelligence (cs.AI)
- Secondary: Computation and Language (cs.CL); Human-Computer Interaction (cs.HC)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The proposed Chat Game Engine (ChatGE) uses large language models (LLMs) to enable non-technical users to develop custom games using natural language. The ChatGE is powered by an LLM that can perform three processes in each turn: configuring the game script, generating code snippets, and interacting with the user through guidance and feedback. A data synthesis pipeline is used to generate game script-code pairs and interactions from a few seed examples. A progressive training strategy is proposed to transfer the dialogue-based LLM to ChatGE smoothly. The effectiveness of ChatGE is evaluated for a poker game case study, considering both interaction quality and code correctness. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The Chat Game Engine uses big language models to help people who aren’t experts make their own games by talking to the model. It’s like having a partner that can help you create a custom game using simple language. The engine has three main tasks: creating the game script, writing the code, and guiding the user through the process. To train the model, the researchers used a special pipeline to generate examples from just a few starting points. They also developed a training method to adapt the model for this new task. As an example, they tested the Chat Game Engine by making a poker game and evaluating how well it worked. |
