AI helps you play games, and NFT holders can earn money without doing anything? Interpreting NIM’s AI agent economy

As the first highly anticipated RollApp on Dymension, NIM Network, which combines AI and GameFi narrative, is building an AI agent ecosystem to bring higher economic value and a new gaming experience to users.

Compiled by Alex Liu, Foresight News

Introduction

AI gaming chain NIM Network has launched an AI agent ownership economic framework driven by new AI agent standards that can play games, trade and generate income, as well as an AI gaming experience called TITANS.

As the first highly anticipated RollApp on Dymension, NIM Network, which combines AI and GameFi narrative, is building an AI agent ecosystem to bring higher economic value and a new gaming experience to users.

This article will briefly introduce NIM’s blueprint concept, functional design and economic model for AI game agents, and explain the implementation details and future plans.

Blueprint

In short, NIM will create a world where “AI participates in games”. Here, AI agents can interact with different games. They are driven by the latest AI models and do not need to rest. In the game, these agents become AI players, they can imitate the behavior of human players, and even surpass them in some aspects of game performance. Players, developers and holders can train, develop, collect or trade these AI agents, and at the same time, predictors can participate in the prediction of e-sports between AI agents.

In such a world, NIM positions itself as an AI hub that holds the economic value, ownership, and logic of AI gaming agents.

This is a beautiful blueprint. The potential economic value and innovation in entertainment methods behind it provide plenty of room for imagination.

Functional design and economic model

By integrating a series of new features into the NIM chain, NIM aims to open up the blue ocean of the game AI agent economy. These features mainly include:

• The ability to verifiably use AI models on-chain to drive AI agents.
• Innovative Genesis Proxy. It is scalable and can adapt to different games and applications and make corresponding adjustments for them.
• Introducing incentives for the creation of AI agents. Creating chatbots, avatars, TITANS, or other innovative agent types can receive incentives.
• Comprehensive AI agent ownership model. A carefully designed model allows for ownership and monetization of agents at different levels and degrees. Players, developers, holders, and predictors will each have a specific economic model.
• AI agents are able to leverage the latest technologies and innovations in open source AI to achieve continuous self-improvement cycles.
• A set of primitives tailored for on-chain automation and better AI agent user experience.

Implementation details

The first step to realize the AI agent game world is to create the first batch of genesis agents.

NIM is launching a new AI NFT primitive to facilitate the creation of its Genesis Agent. The Genesis Agent is based on the ERC6551 standard. This new standard allows NFTs to have smart accounts, called AA Agents (Agent Primitives Driven by Account Abstraction). AA Agents are game-specific agents that use AI models to control the execution of agents in the game.

AI NFTs like TITANS NFTs are equivalent to using ERC6551 to control the keys of these smart agent accounts. Each AI NFT can derive multiple accounts, representing different AI players that can access the AI model. AA agents can receive rewards from the game, and AI NFTs are a collection of different AI players across multiple games.

In addition, using the new AI NFT standard, the following will become possible:

• Staking mechanism: used to upgrade agents (enabling them to access newer AI models or make more calls to models).
• Revenue sharing: AI NFT will serve as a composite target, enabling users to become diversified AI investors and users. Holding an AI NFT, you can store different AI players collected and created in it.

ERC6551 can drive various types of users to participate in the AI economy:

• Players: Players can activate and monitor AI agents in the game, train them based on feedback, and improve the performance of AI agents. This will be achieved through a marketplace where agent data and agents can be acquired and rented from owners.
• Developers: Developers can build authoring platforms, AI models, and AI tools while sharing fees/revenue with active users, or simply let builders use them.
• Holders: Holders can simply hold AI assets and enjoy the passive income generated by players, creators, and developers activating them.
• Predictors: In NIM’s vision, such an AI gaming world will eventually usher in e-sports. People can build an entire entertainment layer around AI agents. High-performance AI players will get topics and attention, and predictors who are good at predicting game results will flock to try to profit from correct judgments.

future plan

NIM has planned how to iterate on the versions of such AI agent world in the future. Here are the key points of different versions:

V1 - Bringing creators and players closer

Starting with TITANS, NIM is trying to bridge the gap between creators and owners, such as generating intelligent agents with natural language interfaces in games. In this way, users will be able to directly experience the new experiences brought by AI agents.

V2 - Hold an agent without activating it

NIM believes that the economics of AI agents should take into account different categories of users and communities. Developers are good at creating AI agents, but they are not necessarily happy to participate in the game. Some communities may want to invest in the ownership of AI agents and provide liquidity, but are reluctant to activate and use them. Players, developers, and holders are the backbone of the AI economy, and NIM will provide a unique economic model for each role.

V3 - Introducing the Entertainment Layer

NIM believes that AI-powered games will naturally evolve into an entire entertainment layer, just like traditional games. The ability of AI agents to continuously improve, run non-stop, and challenge each other with new AI capabilities unlocks the possibility of an automated AI entertainment layer. This includes e-sports, which can attract millions of gaming viewers around the world to participate in real-time prediction markets and other forms of monetization.

Reference Links:

_network/status/1775133480633995619