Project Title: LLM-driven Turing Game Bot

Objective: In this project, the student will develop an intelligent bot capable of playing the Turing Game (turinggame.ai), where users try to distinguish between human participants and AI-based bots. The bot will leverage a local Large Language Model (LLM) and aim to improve its performance over time.

Key Deliverables:

1. Bot Development: Implement an AI-driven bot that can participate in the Turing Game. The bot should engage in natural conversations while attempting to mimic human-like responses.
2. Integration with Online System: Communicate with the administrators of turinggame.ai to integrate the bot into their platform for real-world testing and feedback.
3. Performance Optimization: Continuously monitor and improve the bot’s performance based on feedback from real sessions, making the bot more difficult to detect over time.
4. Conference Paper: Write and submit a research paper summarizing the design, challenges, and outcomes of the project for a suitable AI or machine learning workshop.

Required Skills:

• Proficiency in Python.
• Familiarity with large language models (LLMs) and machine learning frameworks.
• Ability to communicate with external platforms to facilitate integration.

Tools & Technologies:

• Python
• Local LLM models (e.g., GPT-based models)
• API communication (to interact with the Turing Game system)

Project type: Cmpe492