Introducing Mage: Agent Launchpad & Platform for AI Entertainment
“Any sufficiently advanced technology is indistinguishable from magic.”
We are excited to introduce Mage, a new agent launchpad coming soon to Treasure. Mage is more than just a launchpad though, it’s an entire agentic platform that will unlock new gaming and entertainment paradigms through multi-agent orchestration.
The space at large has quickly progressed from celebrity social agents to a desire for more: truly productive and interactive agents that autonomously carry out actions in DeFAI and AI gaming. Enter Mage.
How will Mage deliver this?
Mage agents are built atop the Eliza agent framework and integrated with Treasure’s gaming infrastructure. Deep game system and data access, support for all game engines through the TDK, and more.
Mage will allow anyone to launch autonomous AI agents with compute and inference abstracted away as “mana”. Mana will be fuelled by $MAGIC to serve as the lifeblood to summon and sustain agents.
Creating a developer-friendly primitive for games to unlock new game design possibilities. This includes agents that are game agnostic and capable of going beyond just sandbox experiences.
Decentralized agents that are secure and unruggable. Non-custodial wallets owned only by the agent and governed by their communities. LLMs that are run in TEEs to ensure agent decision making is tamperproof and verifiable.
Mage will usher in a world where agents and humans cooperate, compete, play, trade, and build together. It will enable agent swarms – no, entire universes of agents rather that can be summoned to operate in-game, onchain, and beyond.
Mage’s first agents will be focused on games as their initial training ground. From supercharging IP and worldbuilding to enabling autonomous worlds that are constantly evolving, and to unlocking entirely new game types, we have barely started to scratch the surface of what’s possible.
Check out the full article to learn more and spread the news! ✨