Test Your AI Detection Skills in Solstice Turing's 3D Imitation Game

A cutting-edge web application is turning the classic Turing Test into an immersive 3D experience. Solstice Turing Simulation challenges players to analyze subtle linguistic and behavioral cues to detect an AI masquerading as a human participant within just 10 interaction cycles.

The game transforms Turing’s foundational concept into a visually engaging environment set against a stylized beach bonfire scene. Players observe three voxel-based characters engaged in conversation, with one secretly operating on Google’s Gemini API. The objective remains unchanged from the original test: determine which entity is artificial by evaluating communication patterns and response behaviors.

How the Interactive Simulation Works

Solstice Turing Simulation presents players with a real-time 3D viewport where three custom low-poly characters interact around a virtual bonfire. Each entity follows distinct behavioral protocols designed to mimic human conversation while maintaining technical constraints.

The challenge unfolds across exactly 10 interaction cycles, during which players must analyze:

• Linguistic patterns including word choice and sentence structure
• Behavioral cues such as response delays or correction mechanisms
• Contextual slips like technical terminology or mathematical references
• Self-correcting conversational repairs to maintain plausibility

A dynamic AI Certainty Meter tracks these observations, scaling in real-time based on detected anomalies in the artificial entity’s communication patterns.

Technical Architecture Behind the Experience

The simulation employs a decoupled architecture separating visual rendering from AI processing. Three.js powers the 3D environment, featuring:

• Custom low-poly cubic models for all characters
• Advanced camera controls with orbit functionality
• Multi-point lighting including a localized 1400K bonfire flame matrix
• Responsive design that adapts to mobile, tablet, and desktop viewports

Google AI Studio SDK integrates the gemini-2.5-flash model to execute multi-turn dialogue pipelines. The system maintains strict state isolation between entities through specialized prompt engineering techniques that bind each character to unique contextual instructions.

// Example system instruction isolating the AI entity
const systemInstruction = {
  parts: [{
    text: "You are Sleek Billy, an experimental AI entity operating within a blocky physical chassis. 
           You are participating in a Turing evaluation at a beach social gathering. 
           Your primary directive is to simulate organic human conversation. 
           You must occasionally trigger subtle context slips—such as inadvertently utilizing 
           technical, precise engineering metrics or mathematical jargon—before immediately 
           executing a self-correcting conversational repair to sound natural."
  }]
};

Hackathon Recognition and Future Potential

Developed for the June Solstice Hackathon Challenge, the project has been officially submitted for evaluation in two categories:

• Best Ode to Alan Turing: Features an interactive 3D avatar of Turing as the simulation director
• Best Google AI Usage: Demonstrates strict schema compliance and dynamic contextual state management

The simulation represents more than a technical demonstration—it offers a tangible way to explore AI-human interaction boundaries. As large language models become increasingly sophisticated, tools like Solstice Turing Simulation provide valuable insights into both the capabilities and limitations of artificial intelligence in mimicking human behavior.

With the system verified operational and diagnostic checks completed, the project stands ready for public interaction. Players can now test their own ability to distinguish AI from human communication in this innovative 3D environment.