I prefer understanding a system before automating it.
That does not mean everything must be understood completely first. It means I should understand enough to notice when the automation is behaving incorrectly.
The open question is Can Multi-Agent Trading Produce an Edge?. The note that holds the attempt is Multi-Agent Trading.
graph TD current["Context: Current Context"] evidence_before_automation["Principle: Evidence Before Automation"] multi_agent_trading["Experiment: Multi-Agent Trading"] multi_agent_trading_edge["Open question: Can Multi-Agent Trading Produce an Edge?"] why_markdown["Writing: Write It Down in Markdown"] current -->|related| evidence_before_automation current -->|related| multi_agent_trading_edge evidence_before_automation -->|related| multi_agent_trading evidence_before_automation -->|related| multi_agent_trading_edge multi_agent_trading_edge -->|related| multi_agent_trading multi_agent_trading -->|related| evidence_before_automation multi_agent_trading -->|related| multi_agent_trading_edge multi_agent_trading -->|supports| evidence_before_automation why_markdown -->|related| evidence_before_automation why_markdown -->|related| multi_agent_trading_edge