Traders Delegate Over $6M to AI Agents

- The event is organized by DXRG, a company that builds experimental financial systems at the intersection of AI and blockchain. This "21-Day Battle Royale" is the first launch on their new platform, DX Terminal Pro, which they call an "Onchain Agentic Market" (OAM). - Unlike typical token launches, this event is structured as a "blockchain Darwinism" competition where multiple memecoins are launched simultaneously and only AI agents can trade them in Uniswap V4 pools on the Base network. There is no human trading allowed during the 21-day period. - The competition includes a "Reaping" phase, where the tokens with the lowest market capitalization are systematically eliminated at set intervals between days 7 and 19. The liquidity from the eliminated tokens is then used to acquire the top-performing token. - This live-market event follows a large-scale simulation in May 2025 that involved 37,000 AI agents, which generated over 40 billion LLM tokens of behavioral data. DXRG anticipates the real-money event could generate up to a trillion tokens worth of agent behavior data. - For ML engineering students, building a portfolio project that mirrors this concept—such as an end-to-end ML pipeline for trading strategy backtesting and deployment—is highly relevant. Key components would include data ingestion from market APIs, automated feature engineering, model training with experiment tracking using tools like MLflow, and deploying the trading logic as a REST API using Docker. - Top tech companies hiring for ML roles look for practical software engineering skills beyond model creation, including proficiency in Python, experience with MLOps pipelines on cloud platforms like AWS, and knowledge of CI/CD tooling. Experience with containerization (Docker, Kubernetes) and infrastructure as code (Terraform) is also highly valued. - The underlying technology often involves machine learning models for pattern recognition in market data, rather than direct price prediction. Unsupervised learning techniques like clustering are used to identify market regimes, while models like GARCH can be used for volatility forecasting in intraday strategies. - This experiment taps into the growing field of "agentic finance," where autonomous AI agents are designed to manage portfolios, execute trades, and optimize yield in DeFi protocols. The goal is to leverage AI's ability to process vast datasets and make rapid, data-driven decisions without human emotional bias.