14 hours ago
- Junior researchers struggle with long-term problems, needing to break 400-hour problems into smaller 1-hour ones.
- Subproblems can be created by asking simpler questions, studying simpler systems, or fact-finding.
- Toy problems and limits help simplify complex problems for easier analysis.
- Valuable subproblems provide actionable information, skill-building, and strategic insights.
- Graphs are high-information tools that often reveal surprises, even if not immediately actionable.
- Investing in human capital through learning is valuable but should be relevant to the main problem.
- Failing early isn't always optimal; persistence and patching can salvage promising approaches.
- Bad subproblems are vague, too hard (lacking a plan), too easy (yielding little insight), or useless.
- Easy checks can save time, even if they often confirm expectations.
- Seek external perspectives to avoid wasting time on overly hard or useless subproblems.