Autonomous Agent Behavioral Model Coaching via Contextual Feedback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video games lack an engaging experience for players who want to train virtual characters without directly controlling their actions, as interactions are limited to either player-controlled characters or non-player characters with no strategic influence on the virtual character's behavior.
Innovation Solution
A method for operating a video game that includes maintaining a gaming environment with an autonomous agent and a behavioral model, allowing players to coach the agent through instructions and contextual feedback, adapting the model to influence the agent's actions in the game environment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If the player directly controls the player character through inputs, then the player can perform actions in real-time, but the player cannot strategically train the virtual character's behavior
Solution Approach 1:
The system dynamically switches between two operational modes: gameplay mode where the player directly controls the player character in real-time, and coaching mode where the player strategically trains the behavioral model. This dynamic transition allows the system to satisfy both real-time control needs and strategic training needs at different times, resolving the contradiction between ease of operation and adaptability.
Solution Approach 2:
The game session is segmented into distinct phases: gameplay sessions for real-time action and coaching sessions for strategic training. During gameplay sessions, the player character acts autonomously based on the behavioral model, while during coaching sessions, the player provides instructions to adapt the behavioral model. This segmentation allows each mode to optimize for its specific purpose without interference from the other.
2Speed
If the virtual character actions are controlled by player inputs, then the gameplay is responsive, but the degree of customization is limited to physical attributes
Solution Approach 1:
The system performs preliminary action by adapting the behavioral model during coaching sessions before the actual gameplay occurs. The player provides instructions and contextual feedback that shape the behavioral model's decision-making capabilities in advance, allowing the player character to autonomously make strategic decisions during gameplay without real-time player input, thus maintaining responsiveness while enabling deep behavioral customization.
3Adaptability or versatility
If the player influences the virtual character's behavior through training, then the customization extends beyond physical attributes, but the system complexity increases
Solution Approach 1:
The behavioral model serves as an intermediary layer between the player's strategic instructions and the player character's actions. During coaching sessions, the player provides high-level instructions and contextual feedback, and the behavioral model translates these into actionable decision-making rules that the player character executes autonomously during gameplay. This intermediary simplifies the interface between player and character, managing system complexity while enabling deep behavioral customization.
4Measurement precision
If the player provides contextual feedback during coaching, then the behavioral model adapts more accurately, but the coaching session duration increases
Solution Approach 1:
The system maintains continuity of useful action by allowing the behavioral model to continuously learn and adapt from player feedback throughout the coaching session. The model processes each instruction and contextual feedback incrementally, building upon previous learning rather than requiring complete retraining. This continuous learning approach improves training accuracy while minimizing the time required compared to traditional retraining methods.
Data Source
AI summary
A method of operating a computer for carrying out a video game is provided. The method includes maintaining a gaming environment in a non-transitory memory, with the gaming environment defining an autonomous agent and a behavioral model associated with the agent. The method also includes implementing a gaming session of the video game during which the agent carries out actions in accordance with the behavioral model, and implementing a coaching session of the video game by receiving instructions from a player. The instructions indicate a gameplay action to be performed by the agent in the gaming environment. The method also includes implementing the coaching session by receiving contextual feedback from the player regarding said gameplay action. The method also includes implementing the coaching session by adapting the behavioral model based on the instructions and the contextual feedback.