⛵ Our Mission: Gamified HITL

AI is shifting from generic to specialized; worlds.fun is pioneering the LLM narrative by enabling zero-cost optimization powered by gamified human feedback.

Our journey begins with an MMOLLM(Mass Multiplayer Online LLM) game, where players interact with AI agents in a dynamic worldmap game. But it’s more than a game, it's a live agents optimization ground where:

1. Players immerse in interactive gameplay, providing real-time inputs.
2. Developers leverage specialized gameplay data that enriches retrieval pools with high-value, context-specific interactions.
3. These specialized data turn optimized intelligence into valuable, tradable assets.

Potential applications may include:

• Content — Baby Shark develops a worlds.fun mission where players persuade virtual animals to join their zoos → later using the interaction dataset to enhance its AI-driven children’s content recommendation.
• Gaming — GTA6 introduces a worlds.fun mission where players threaten agents to surrender their lands and power → refining context-aware NPC behavior by integrating real player persuasion tactics into AI decision trees.

🃏 Gameplay

🦑 How to play

Up to 100 players start a 15-minutes limited game. Players can select any of the 200 countries in the world map and begin a chat with respective state leaders or presidents. The player that convinces the most leaders within the timeframe wins.

Topics always change, ranging from convincing leaders to turn their country communist to building up a nuclear arsenal. Let the battle of words begin 🏁

🐳 How worlds.fun stands out among LLM games:

• Multiplayer (up to 100 players)