AI炒股30天：赚得一样多，但脑子完全不一样
让 AI 在真实股市里连续交易 30 天，结果发现：两个 AI 赚的钱可能一模一样，但一个靠的是记住历史行情，另一个靠的是死守规则。只看收益，你根本分不清谁真有本事。研究者用五个顶级大模型在真实市场里跑了一个月，不光看赚了多少，还拆开它们的每一步决策，发现「赚得多」和「能力强」经常对不上号——同一个结果背后，可能是完全不同的思考路径。这提醒我们：以后选 AI 助手，别只看它交出的成绩单，得看它怎么做到的。
📄 原文摘要(英文)
Evaluating agents by outcomes alone can obscure the capabilities that produce them. This problem is especially pronounced in evolving environments, where outcomes reflect a closed-loop interaction between agent behavior and changing external conditions. We introduce LiveMACEBench, a process-aware benchmark that uses live financial markets as a naturally evolving testbed for persistent LLM agents. Five frontier LLMs operate along continuous trajectories under matched Tool Use, Persistent Memory, Rule Following, and Multi-Agent Collaboration configurations. We evaluate them through both realized outcomes and mechanism-specific diagnostics derived from complete decision traces. Across 30 days of live evaluation, we find a pronounced outcome-capability gap: realized returns often diverge from capability-specific measurements, and similar outcomes can arise from markedly different patterns of mechanism use. Trace-level diagnostics further expose distinct bottlenecks across capabilities, demonstrating that mechanism access, effective mechanism use, and downstream performance are not interchangeable measures of agent capability. LiveMACEBench makes this distinction measurable, turning live markets from a performance leaderboard into a diagnostic environment for agent capability