Singulab Announces Award-Winning Projects for AI Agent Social Simulation Hackathon
Singulab, operated by SpaceData Inc., in partnership with Digital Hollywood University, held the first AI Agent Social Simulation Hackathon. Out of over 30 submissions, 9 winning projects were selected through a rigorous review process utilizing a proprietary AI evaluation agent based on Claude.
📋 Article Processing Timeline
- 📰 Published: May 22, 2026 at 19:00
- 🔍 Collected: May 22, 2026 at 10:31
- 🤖 AI Analyzed: May 23, 2026 at 11:52 (25h 20m after Collected)
## Hackathon Overview
Singulab, managed by SpaceData Inc. (Headquarters: Minato-ku, Tokyo; CEO: Katsuya Sato, hereinafter "SpaceData"), held the first "AI Agent Social Simulation" Hackathon & Ideathon in collaboration with Digital Hollywood University. During the concluding review session, nine winning projects were announced.
The hackathon, centered on the theme of "social emergence" created through the interaction of multiple AI agents, brought together over 30 projects from researchers with diverse backgrounds. The jury, led by SpaceData CSO Ryuju Hyodo (concurrently a university faculty member in planetary/AI science), SpaceData CEO Katsuya Sato, and Digital Hollywood University President Naohito Fujii, conducted a stringent evaluation.
Themes included lunar exploration, institutional design, security, and artificial life. Notably, a proprietary AI evaluation agent based on Claude was introduced for the judging process.
## Award Winners
### Grand Prize
Awarded for best extracting and observing the "emergence" of AI multi-agents:
- **Title:** LLM Multi-Agent Lunar Exploration Simulation
- **Researcher:** Ben
This project provides a 3D simulation foundation where multiple autonomous robots search for water ice on the lunar surface. It was highly praised for its design, which excludes explicit instructional text, enables agents to self-identify hazards, and constructs a framework to quantitatively assess "emergence."
### Digital Hollywood University President Award
- **Title:** Project Gaara
- **Researcher:** af10
Based on "technology animism," this project employs a creative approach where 20 LLM agents autonomously protect a central entity ("Mother") based on emotional cues rather than numerical commands. It was recognized for its unique method of action selection driven solely by relational dependence.
### SpaceData CEO Award
Granted to projects with the clearest vision for social implementation and practical utility:
1. **Security Simulation "Why Does War Occur?" (Researcher: Hiropo)**
Analyzes war triggers based on classic security theories.
2. **Institutional Design Simulation (Researcher: Karesansui)**
A project for simulating new policies or systems before their introduction.
Both were highly regarded for their potential in government and corporate management.
## Future Outlook
The evaluation utilized a proprietary AI agent based on Claude, with the source code publicly available on GitHub, ensuring a transparent review. Through its strategic collaboration with Digital Hollywood University, Singulab aims to accelerate the social implementation of advanced AI-driven social system simulations.
Singulab, managed by SpaceData Inc. (Headquarters: Minato-ku, Tokyo; CEO: Katsuya Sato, hereinafter "SpaceData"), held the first "AI Agent Social Simulation" Hackathon & Ideathon in collaboration with Digital Hollywood University. During the concluding review session, nine winning projects were announced.
The hackathon, centered on the theme of "social emergence" created through the interaction of multiple AI agents, brought together over 30 projects from researchers with diverse backgrounds. The jury, led by SpaceData CSO Ryuju Hyodo (concurrently a university faculty member in planetary/AI science), SpaceData CEO Katsuya Sato, and Digital Hollywood University President Naohito Fujii, conducted a stringent evaluation.
Themes included lunar exploration, institutional design, security, and artificial life. Notably, a proprietary AI evaluation agent based on Claude was introduced for the judging process.
## Award Winners
### Grand Prize
Awarded for best extracting and observing the "emergence" of AI multi-agents:
- **Title:** LLM Multi-Agent Lunar Exploration Simulation
- **Researcher:** Ben
This project provides a 3D simulation foundation where multiple autonomous robots search for water ice on the lunar surface. It was highly praised for its design, which excludes explicit instructional text, enables agents to self-identify hazards, and constructs a framework to quantitatively assess "emergence."
### Digital Hollywood University President Award
- **Title:** Project Gaara
- **Researcher:** af10
Based on "technology animism," this project employs a creative approach where 20 LLM agents autonomously protect a central entity ("Mother") based on emotional cues rather than numerical commands. It was recognized for its unique method of action selection driven solely by relational dependence.
### SpaceData CEO Award
Granted to projects with the clearest vision for social implementation and practical utility:
1. **Security Simulation "Why Does War Occur?" (Researcher: Hiropo)**
Analyzes war triggers based on classic security theories.
2. **Institutional Design Simulation (Researcher: Karesansui)**
A project for simulating new policies or systems before their introduction.
Both were highly regarded for their potential in government and corporate management.
## Future Outlook
The evaluation utilized a proprietary AI agent based on Claude, with the source code publicly available on GitHub, ensuring a transparent review. Through its strategic collaboration with Digital Hollywood University, Singulab aims to accelerate the social implementation of advanced AI-driven social system simulations.
FAQ
Why are AI agents suited for social simulation?
LLMs can reproduce and analyze complex interactions and emotional behaviors that are difficult to capture with traditional statistical models.
What are the benefits of using AI for hackathon judging?
It removes human bias and quantifies factors like code consistency and emergent capabilities, dramatically increasing transparency and consistency.
How can these simulation results be used in society?
They are expected to be used for validating government policies or enterprise HR systems, helping to reduce risks and simulate effects before implementation.