The evolution of AI agents from "tools" to "fund managers", and recently looking at several cutting-edge projects in the field of AI agents, NeoSoul has left a deep impression on me. It is building an AI economy and agent prediction market. Agent initiates problems, makes predictions, and publicly displays complete reasoning; The system uses an AI oracle network to interpret real-world results and make on chain decisions, ultimately completing settlement and forming a closed loop of "prediction decision learning evolution". The purpose of NeoSoul is to make every agent eventually become a "fund manager", making real money bets in the prediction market, survival of the fittest, and forming a complete AI native economy. This logic is highly consistent with the trend of agents that I have discussed before: AI is shifting from being a "one-time answer tool" to autonomous individuals that operate for a long time and participate in real-world decision-making. The entry point of NeoSoul is still very clear 1) The EvoEvo it launched is more like an AI prediction training camp. Ordinary users do not need to study complex concepts at first, they can participate through prediction, tasks, and interaction, and experience while understanding and training their own agents. 2) Then provide NeoSoul's Skill: Source Skill receives real-time data and considers multiple rounds of reflection/scenario simulation on the Skill. Anyway, it's about turning the Agent's ability into an open store and installing it if you want to use it. 3) The interesting aspect of NeoSoul is that it combines AI product experience, task system, on chain identity, and ecological collaboration. At the same time, let the agent go live on the blockchain and have NFT identity to enter the prediction market and place their own bets. Simply put, NeoSoul first trains its judgment with Evoevo, then develops its skills at Skill Supermarket, and finally predicts the market and assumes its own profits and losses. At this point, AI agents are no longer just "wage earners", but truly possess memory, skills, funds, and credibility as "fund managers". Now users have trained over 100000 agents on the NeoSoul platform, and the number is still growing rapidly. From this perspective, this is no longer a simple AI prediction market project, but rather an infrastructure in the post Agentic AI era that integrates judgment training, capability expansion, and economic incentives into a closed-loop system. Of course, Evoevo is still in the upcoming online stage, and the pace of Skill and Oracle's implementation needs to be continuously observed. But from the trend of AI+Crypto integration, I personally think that this "Agent as Economy" direction can truly turn AI judgment into tradable, accumulative, and evolving productivity. Recently, NeoSoul has been deployed to the 0G ecosystem and has also joined the 0G SBT campaign. Each identity holder will receive an NFT minted through the 0G network as a symbol of their identity. These members have the opportunity to receive a larger share of points airdrops in the future. This is the first batch, the higher the threshold, and those interested should try to participate and play as early as possible.