The author is initiating a 30-day experiment where an AI agent, built on the DuDuClaw platform, will autonomously manage a real stock trading account with an initial capital of NT$2,000. The primary goal is to test whether a Large Language Model (LLM) can evolve into a Large World Model (LWM) capable of learning and adapting in a high-stakes environment, rather than just predicting the next token. The experiment will track the AI's strategy development, trading decisions, and its ability to learn from both successes and failures, with a target of doubling the initial investment. AI
IMPACT Tests the practical application of LLMs in complex, high-stakes decision-making environments like stock trading.
RANK_REASON The article describes an experiment using an AI agent platform for a real-world task, rather than a new model release or core research.
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
