What is deep work optimization?

Deep work optimization is a concept in personal epistemology: Dedicate focused time blocks to optimizing specific agents rather than trying to optimize everything continuously.

Example: You have a decision-making agent that handles career evaluations, a writing agent that processes raw ideas into structured arguments, and an energy management agent that regulates work-rest cycles. All three could use improvement. Instead of vaguely 'working on yourself' across all three simultaneously, you declare: 'This week, Tuesday and Thursday mornings are optimization sprints for my decision-making agent.' You pull the performance baseline from L-0575, isolate the specific bottleneck (the agent stalls when options exceed three), and spend two 90-minute blocks redesigning the evaluation criteria. By Thursday afternoon, the agent handles five-option decisions reliably. The other two agents didn't get better this week — but one agent got measurably better, and you have a documented protocol for repeating the process.

This concept is part of Phase 29 (Agent Optimization) in the How to Think curriculum, which builds the epistemic infrastructure for agent optimization.