Weekly Summary - 2026-08-03 to 2026-08-10

Overview

This week connected long-horizon education, career, and decision-making with practical health and systems thinking. The notes emphasize choosing paths by durable fit and real conditions rather than prestige, popularity, hindsight, or short-term outcomes; several of these ideas were consolidated into human-review workflows and reusable reflection skills.

Themes and insights

Education and career are fit-and-ecosystem problems

A “good major” is framed as the intersection of sustained interest, relevant ability, strong program-level training, and credible market demand—not a national hot/cold label. University value includes skills, peers/networks, and field-specific influence; for doctoral study, the mentor and research group can matter more than institutional brand. The academic-career note also stresses that exceptional achievement rests on a broad research ecosystem, sustained commitment, community participation, and circumstances beyond individual ability.

Long-term strategy should remain adaptive

Long-range planning is presented as maintaining a map of important unresolved problems, building relevant capabilities, running small experiments, and updating when evidence or readiness changes. This rejects both chasing short-lived trends and treating a calendar-based plan as proof that conditions are ready.

Evaluate decisions by process, not a flattering narrative

The counterfactual and finance notes distinguish facts from imagined alternate futures, and decision quality from eventual outcomes. A fair retrospective restores the costs, risks, information, and personal constraints present at the time; its value is to improve a future rule, rather than continually punish a past self. Likewise, good results can follow poor reasoning through luck, so principles and understanding should precede price, excitement, or apparent bargains.

Turn reflection into bounded review loops

The extracted skills capture related reusable processes: audit first reactions and evidence; begin learning with the smallest safe attempt and feedback; journal uncertain decisions before and after commitment; protect resources with explicit boundaries; and test side work or AI-workflow changes with limits, metrics, and human judgment where stakes require it. The family workflow keeps finance, parenting, education, and career material as a human-review checklist rather than autonomous advice.

Health exposure can inform ordinary daily choices

The PM2.5 note treats fine-particle exposure as a sometimes-invisible health risk and proposes checking PM2.5 separately from composite AQI before deciding on ventilation and outdoor activity. It notes that protective measures depend on context—such as mask fit, a relatively enclosed room, appropriate purifier capacity, and local interpretation of the displayed measure.

Notes covered

• 梯队筛选：为何说学术道路不好走？ — Argues that elite scientific achievement depends on a large, layered, long-running research ecosystem rather than isolated genius. It highlights intrinsic research interest, long-term commitment, methods, mentors, networks, and luck or institutional context.
• 大学：好专业，怎么选？ — Defines university returns through knowledge/skills, networks, and industry influence, with knowledge conversion depending on student fit. It recommends assessing a specific school-major combination against training quality, graduate supply, local demand, and adaptability rather than popularity alone.
• 我会如何定义“好专业” — Condenses a good major into four intersecting conditions: interest, ability fit, adequate program quality, and real, durable market demand. It cautions that popularity alone or preference alone is insufficient.
• 006 - Chỉ số PM 2.5 — Explains PM2.5 as fine particulate exposure with potential respiratory and cardiovascular implications, often without immediate symptoms. It proposes using daily air-quality information to guide ventilation, outdoor time, masking, and purifier use while noting measurement and fit limitations.
• 反事实的复利 — Examines how imagined, more fortunate alternative life paths can compound regret by omitting their costs and uncertainty. It offers a reflection sequence: separate fact from imagined continuation, restore omitted constraints, and retain only lessons that can improve future decisions.
• 真正的长期规划 — Presents long-term planning as sustained attention to important problems, capability accumulation, small experiments, and evidence-based updates. It favors readiness triggers over rigid multi-year schedules.
• 家庭财商、学习与天赋培养工作流 — A consolidated, explicitly non-autonomous review checklist for family money modeling, child-owned funds, housing and business review, learning, talent discovery, university use, and major selection. It centers evidence, safeguards, and human judgment for high-stakes family, financial, and educational decisions.
• critical-reaction-audit — Documents a workflow for treating fast emotional or socially conditioned reactions as provisional, separating evidence from familiarity, and choosing proportionate reversible responses. It includes claim-strength, alternative-cause, and analogy checks.
• minimal-actionable-learning — Describes MAKE: reduce a vague learning goal to the minimum safe knowledge needed for a short, observable first attempt. Feedback, one-variable iterations, and capacity-aware practice turn information into behavior.
• decision-journaling — Records options, constraints, facts, assumptions, uncertainty, contextual distortions, and review triggers before committing to consequential decisions. Its retrospective method evaluates the original process separately from the result and includes counterfactual-regret review.
• adaptive-long-horizon-planning — Formalizes a living strategy for uncertain, consequential ambitions: problem map, root-problem test, capability bets, bounded experiment, evidence update, and next move. It keeps direction stable while timing and tactics remain revisable.
• side-hustle-evaluation — Provides a guarded review of freelancing, monetized hobbies, and small businesses against main-career compounding, hidden costs, legal constraints, recovery, and protected assets. Promising options are framed as reversible, time-boxed tests with stop criteria rather than income promises.
• ai-workflow-bottleneck-review — Addresses the case where AI increases output but not delivery quality or speed. It recommends mapping end-to-end flow, automating deterministic checks, repairing data ownership and handoffs, and reserving human review for meaningful exceptions.
• resource-boundary-review — Offers a process for evaluating requests on time, money, energy, attention, reputation, or family support. It identifies displaced priorities, establishes reusable limits, and favors concise, clear boundaries with proportionate responses to repeated harmful behavior.
• Second-Brain Skill Mining Log — Records the history and rationale for second-brain scans, workflow consolidation, and skill extraction. The latest entries document the long-horizon planning skill, a historical recovery backfill, and mirrors for the extracted skills.
• 怎样不让欲望、运气和短期结果，扭曲自己的判断标准。 — Uses an investment example to argue that outcomes do not alone validate decisions: luck can reward a process that violated one’s principles. It prioritizes understanding and value over price or apparent bargains, and integration into stable judgment over passive information consumption.

Questions to carry forward

1. For an education or career choice under consideration, what evidence is available for all four conditions—interest, ability fit, program quality, and local or target-market demand?
2. Which important long-horizon problem is worth monitoring now, and what small capability-building experiment would produce useful evidence without requiring a rigid commitment?
3. Which recent regret still relies on an imagined best-case alternative, and what future decision rule—if any—remains after restoring the original costs, risks, and information?
4. Where is AI increasing draft volume without improving end-to-end flow, quality, or decision capacity, and which deterministic check or handoff is the actual constraint?
5. What daily air-quality source and household rule would make PM2.5 information usable without confusing AQI with PM2.5 concentration?