Tired of the monotonous content generated by AI? Mirra AI uses InfoFi to train AI agents.

The InfoFi sector has entered a warring states period, with Kaito, Cookie, and BubbleMap competing fiercely. Mirra AI, on the other hand, has relatively low attention, but perhaps this InfoFi project that combines AI model training has some lying in ambush potential. Mirra AI uses community platforms like ( and X) to train its AI. Quality content creators and scouts looking for high-quality content can earn points from this. The scoring rules of Mirra AI emphasize authentic content creation rather than rephrasing the White Paper or content generated by AI.

Mirra AI trains AI agents using community platforms.

Mirra AI is a project that uses social platforms ( such as X) to train AI, which can be simply divided into Creator ( Content Creator ) and Scout ( Scout ) Two roles. If you find good creative content on Twitter, you can mark @MirraTerminal in the comments section. At this point, Mirra will perform content validation, which is equivalent to submitting AI training data, contributing data, and making AI smarter. Both creators and scouts will earn points after verification, please note that you need to link your X account to claim points.

As a content creator, the content you create becomes part of training the AI. You'll also earn points when your insights are validated. When Mirra points are widespread enough, the leaderboard also represents who is a recognized thought leader. For the scout's role, you decide what the AI learns, and you also get points from it. Mirra AI is building decentralized AI in this way, and the first training case is the Mirra Agent, and there seems to be a possibility of training other agents.

Mirra likes content that provides space for thought and is written by real people.

The founder of Mirra, Lester, shared the process of earning points on Twitter. Overall, the content detection is divided into two phases. The first phase occurs when scouts tag Mirra in their posts, which involves scanning the scout's content and account. If it detects that ( is suspected to be a bot, ChatGPT content, spam, or contains blacklisted keywords ), the scout's request will be ignored. When the first detection passes, it will display: "Content has been detected. Under evaluation."

In the second detection, Mirra will score the content, and both parties can earn points in the system, but the scores for both sides may not be the same. Mirra may also provide the message "Your content did not pass the detection."

Lester also explains the scoring criteria, which are generally required to achieve points:

Related topics.

Useful information.

Operability.

Based on facts.

Share personal views.

Critical thinking.

Interesting, sharp humor ( Mirra likes black humor and playful expressions ).

Original content.

At the same time, try to avoid:

Writing similar to ChatGPT.

No opinion.

Lack of critical thinking.

Low quality mouthpieces ( most kaito yappers just rewrote the project's White Paper ).

General opinion.

Overuse of emoticons.

Hashtag.

Mentioned multiple accounts in the post.

Overly technical content.

In short, Mirra likes content that allows for good thinking while reading and is written by real people.

However, on the on-chain reputation platform ethos, some users have expressed negative views about Mirra AI.

Tired of the same old content generated by AI? Mirra AI utilizes InfoFi to train AI agents, first appearing in Chain News ABMedia.