Self-training kit · AI-native organisation · data & sources 自训练手册 · AI 原生组织 · 数据与出处

Once a year. Then a quarter. Now — Tuesday. 以前一年一次。后来一季一次。现在是——星期二。

“It used to be a re-org every year. Then every half year. Then every quarter. Now, even every week, every day.” 「过去是每年重组一次，然后每半年，然后每季度……现在甚至每周一次、每天一次。」

This page is not here to calm you down. Calm is not the goal. When the structure around you changes faster than you can learn it, the only stable thing left is what you decide to be excellent at — even when that thing looks nothing like the boxes moving around you. Below: the clock, the numbers, five anchors, a seven-day drill, and every source so you can check it yourself. 这一页不是来让你平静的。平静不是目的。当周围的结构变化得比你学习它的速度还快，唯一稳定的东西，就是你决定做到极致的那件事——哪怕它和身边挪来挪去的方框完全不一样。以下是：节拍器、数字、五个锚、七天训练，以及全部出处，方便你自己核对。

01The clock节拍器

How often does your structure get redrawn? Tap it. 你的组织多久重画一次？点一下试试。

Same amount of attention, more restarts. The ticks below get denser from left to right — that is the whole story of the last decade of org design. 注意力总量不变，重新开始的次数变多。下面的刻度从左到右越来越密——这就是过去十年组织设计的全部剧情。

×250 / YEAR Daily. The org chart is now a group chat. Focus is the only structure left. 每天。组织架构图变成了一个群聊。专注，是唯一还站得住的结构。

02The numbers数字

What employers actually reported 雇主自己报出来的数字

The World Economic Forum surveyed over 1,000 employers representing more than 14 million workers, across 22 industry clusters and 55 economies, for the Future of Jobs Report 2025. 《2025 未来就业报告》调研了 1,000 多家雇主，覆盖 22 个行业集群、55 个经济体，代表逾 1,400 万名员工。

39%

of the core skills you use at work today are expected to change by 2030.你今天工作中使用的核心技能，预计有 39% 到 2030 年会发生改变。

WEF · Future of Jobs 2025
+78M

net new jobs by 2030 — 170 million created, 92 million displaced. Churn touches 22% of today's jobs.到 2030 年净增 7,800 万个岗位——新增 1.7 亿、消失 9,200 万。变动波及今天 22% 的岗位。

WEF · Future of Jobs 2025
63%

of employers name the skills gap as the number-one barrier to transforming their business.63% 的雇主把「技能缺口」列为业务转型的头号障碍。

WEF · Future of Jobs 2025
86%

expect AI and information-processing technology to transform their business by 2030.预计到 2030 年，AI 与信息处理技术将改变自己的业务。

85%

plan to prioritise upskilling their existing workforce.计划优先为现有员工做技能升级。

70%

call analytical thinking essential in 2025 — seven in ten companies, the top core skill.认为「分析性思维」在 2025 年不可或缺——十家里有七家，排名第一的核心技能。

59%

of workers will need training by 2030 — and 11% are unlikely to receive it.的员工到 2030 年需要接受培训——其中 11% 很可能得不到。

40%

of employers expect to reduce headcount where AI can automate the tasks.的雇主预计，在 AI 能自动化的岗位上会减少人手。

Your slide, delivered 🎁你的那页 PPT，已备好 🎁

Creative thinking is the 4th fastest-growing skill for 2025–2030 — right behind AI & big data, networks & cybersecurity, and technological literacy. And 39% of core skills expire by 2030. 在 2025–2030 年增长最快的技能里，创造性思维排第 4——仅次于「AI 与大数据」「网络与网络安全」「技术素养」。同时，39% 的核心技能会在 2030 年前过期。

There. Creative thinking, standing next to a percentage. Try to make it mean something on Monday. 好了，创造性思维旁边有百分比了。周一记得让它真的产生意义。

~70%

the amount the skill set for the same job is expected to change by 2030 (about 25% has already changed since 2015).同一个岗位所需技能到 2030 年的预计变化幅度（2015 年以来已变化约 25%）。

LinkedIn Economic Graph
66%

faster: the rate at which the skills employers ask for are changing in AI-exposed occupations.更快：在受 AI 影响的职业中，雇主要求的技能变化速度。

PwC · AI Jobs Barometer
78%

of organisations now use AI in at least one business function — up from 55% two years earlier.的组织已在至少一项业务职能中使用 AI——两年前这一比例是 55%。

McKinsey · State of AI
95%

of enterprise GenAI pilots showed no measurable P&L impact. Reorgs are fast; learning is not.的企业生成式 AI 试点，没有可衡量的损益影响。重组很快，学习不快。

MIT Media Lab NANDA · 2025
03Why it hurts为什么会累

Change fatigue isn't weakness. It's arithmetic. 变革疲劳不是脆弱，是算术。

Gartner reports that the average employee lived through 10 planned enterprise changes in 2022, up from 2 in 2016 — while willingness to support enterprise change fell from 74% to 43% over the same period. Every reorg spends attention before it produces anything. If the cadence multiplies and the attention budget doesn't, something gets dropped. Choose what you drop, or the calendar will choose for you. Gartner 的数据显示：普通员工在 2022 年经历了 10 次计划内的组织变革，而 2016 年只有 2 次；同一时期，员工支持变革的意愿从 74% 降到 43%。每一次重组都先花掉注意力，之后才可能产出。如果节奏翻倍而注意力预算没有，一定有东西被放弃。要么你来选放弃什么，要么日历替你选。

Calm is not the goal. Aim is.目的不是平静。是准心。 The point of this page这一页想说的
04The aim准心

Five anchors that don't move when the boxes do 五个不随方框移动的锚

Anchor on the user, not the box锚在用户，不在方框

Haier dissolved its middle management into 4,000+ micro-enterprises that form, merge and disband around user needs. The structure is disposable on purpose; the user's problem is not.海尔把中层拆成 4,000 多个「小微」，围绕用户需求不断组建、合并、解散。结构被有意设计成可抛弃的，用户的问题不是。

HBR · The End of Bureaucracy

Give context, not control给上下文，不给控制

Netflix keeps decisions distributed by over-sharing context — strategy, numbers, trade-offs — so people stay aimed even when reporting lines move.Netflix 靠「过量共享上下文」——战略、数字、取舍——让决策留在一线，即使汇报线变了，方向也不散。

Netflix · Culture memo

The work chart beats the org chart工作图谱胜过组织架构图

Microsoft's 2025 Work Trend Index describes “Frontier Firms”: teams assembled around an outcome, with human and agent capacity mixed — the fixed department is no longer the unit of work.微软《2025 工作趋势指数》描述的「前沿企业」：团队围绕结果临时组建，人与智能体混编——固定部门不再是工作的单位。

Microsoft · Work Trend Index

Re-orient faster than you re-plan重新定向要快过重新规划

John Boyd's OODA loop: the winner isn't the one with the best plan, it's the one who updates their picture of reality fastest. In a weekly reorg, orientation is the skill; the plan is a by-product.Boyd 的 OODA 循环：赢的不是计划最好的人，而是最快更新现实认知的人。在每周重组的环境里，「定向」才是能力，计划只是副产品。

John Boyd · The Essence of Winning and Losing, 1996

Prove the machine can't do it先证明机器做不了

Shopify's 2025 memo made reflexive AI use a baseline expectation and asked teams to show why AI couldn't do a job before requesting more headcount. Amazon's CEO told staff generative AI will shrink the corporate workforce. Your job description is being rewritten by agents — write it first.Shopify 2025 年的内部信把「本能地使用 AI」定为基本要求：申请增编前，先说明为什么 AI 做不了。亚马逊 CEO 也告诉员工，生成式 AI 将缩减公司职能岗位。你的岗位说明正在被智能体改写——不如你先动笔。

Shopify memo, 2025 · Amazon · Jassy on generative AI
05The drill训练

Seven days. About thirty minutes each. Do it during the next reorg. 七天，每天约三十分钟。就在下一次重组期间做。

Not a course. A set of repeatable moves that turn a moving structure into a sharper aim. 这不是课程，是一组可重复的动作：把移动中的结构，换成更清晰的准心。

DAY 1第 1 天

Write your invariant写下你的不变量

One sentence: the value you create that survives any org chart. If it names a team, a title or a tool, rewrite it until it names an outcome.一句话：无论组织怎么变，你创造的那份价值。如果这句话里出现了团队名、职位名或工具名，就重写，直到它说的是「结果」。

Anchor: orientation before planning依据：先定向，再规划
DAY 2第 2 天

Audit your skill decay盘点你的技能衰减

Open the WEF skills-on-the-rise list. Mark three of your skills that are decaying and three that are rising. 39% of core skills change by 2030 — decide which 39% is yours.打开 WEF 的「上升中的技能」清单，标出你自己三项正在衰减、三项正在上升的技能。到 2030 年有 39% 的核心技能会变——你要决定，哪 39% 是你的。

WEF · Future of Jobs 2025
DAY 3第 3 天

Ship one agent交付一个智能体

Pick a task you repeat every week. Hand it end-to-end to an AI tool. Time yourself before and after, and write the number down. A number beats an opinion in every reorg conversation.挑一件你每周都重复的任务，端到端交给 AI 完成。前后各计一次时，把数字写下来。在任何一次重组对话里，数字都比观点管用。

Case: Shopify memo · Microsoft WTI 2025案例：Shopify 内部信 · 微软 WTI 2025
DAY 4第 4 天

Run a 25-minute OODA loop跑一次 25 分钟的 OODA

Take one real open decision. Spend 15 minutes on Orient — what actually changed? — and 10 on Decide and Act. No deck. Notice how much of the old plan was inertia.选一个真实未决的决定：15 分钟做「定向」——到底什么变了？——10 分钟做「决策与行动」。不做 PPT。你会发现，原计划里有多少只是惯性。

John Boyd, 1996
DAY 5第 5 天

Draw your work chart画出你的工作图谱

List who and what actually gets your work done now — people, partners, models, agents. Put it beside the official org chart. Believe the first one.列出现在真正在完成你工作的人和东西——同事、外部伙伴、模型、智能体。把它和官方架构图并排放。相信前者。

Microsoft · Work Trend Index 2025
DAY 6第 6 天

Teach it in ten minutes用十分钟讲给别人

Demo Day 3's automation to your team. Teaching is the cheapest test of whether you actually understood it — and it turns a private trick into team capability, which no reorg can take away.把第 3 天的自动化演示给团队看。「讲一遍」是检验自己是否真懂的最便宜方式，而且它把个人技巧变成团队能力——这是任何一次重组都拿不走的。

Practice: retrieval & teaching方法：提取式学习与教学
DAY 7第 7 天

Cut one thing砍掉一件事

Delete or hand off one recurring commitment. Focus is subtraction, and a weekly reorg will happily add for you but never subtract. Then start again at Day 1 — the loop is the point.删掉或转交一项周期性承诺。专注靠减法，而每周一次的重组只会替你做加法，从不做减法。然后回到第 1 天重新开始——循环本身就是重点。

Repeat weekly. Yes, at reorg cadence.每周重复。对，就按重组的节奏。
06Sources & further reading出处与延伸阅读

Check every number yourself 每个数字，都请自己核对

  1. World Economic Forum — Future of Jobs Report 2025 39% skill change · 170M created / 92M displaced / +78M net · 22% churn · 63% skills-gap barrier · 86% AI transformation · 85% upskilling · 70% analytical thinking · 59% / 11% training gap · creative thinking as 4th fastest-growing skill.39% 技能变化 · 新增 1.7 亿 / 消失 9,200 万 / 净增 7,800 万 · 22% 岗位变动 · 63% 技能缺口障碍 · 86% AI 转型 · 85% 技能升级 · 70% 分析性思维 · 59%/11% 培训缺口 · 创造性思维增长第 4。
  2. World Economic Forum — Publications hub Every edition of the report, plus the chief-economists and future-of-growth series.该报告的历年版本，以及首席经济学家展望等系列。
  3. LinkedIn Economic Graph — Work Change data Skill sets for the same job changing ~70% by 2030; ~25% already changed since 2015.同一岗位所需技能到 2030 年变化约 70%；2015 年以来已变化约 25%。
  4. PwC — Global AI Jobs Barometer Skills demanded in AI-exposed jobs changing 66% faster; wage premium for AI skills.受 AI 影响岗位的技能要求变化快 66%；AI 技能的薪酬溢价。
  5. McKinsey — The State of AI 78% of organisations using AI in at least one function; how few call themselves mature.78% 的组织已在至少一项职能中使用 AI；真正称得上「成熟」的极少。
  6. Microsoft — Work Trend Index 2025 “Frontier Firms”, human–agent teams, and the work chart replacing the org chart.「前沿企业」、人机混编团队，以及取代组织架构图的工作图谱。
  7. Amazon — Andy Jassy on generative AI (2025) Agents will change how work is done and reduce the corporate workforce.智能体将改变工作方式，并缩减公司职能岗位。
  8. Gartner — HR newsroom & change-fatigue research 10 planned enterprise changes per employee in 2022 vs 2 in 2016; willingness to support change 43% vs 74%.员工 2022 年经历 10 次计划内变革，2016 年为 2 次；支持变革意愿 43% 对 74%。
  9. Harvard Business Review — “The End of Bureaucracy”, Hamel & Zanini Haier's RenDanHeYi model: micro-enterprises that continuously reconfigure around users.海尔「人单合一」：围绕用户持续重组的小微组织。
  10. Netflix — Culture memo “Context, not control” as the mechanism for staying aimed in a fast-moving structure.「给上下文，不给控制」——在快速变化的结构中保持方向的机制。
  11. Stanford HAI — AI Index Report Independent baseline on AI adoption, capability and economic impact.关于 AI 采用率、能力与经济影响的独立基准数据。
  12. MIT Media Lab (NANDA) — “The GenAI Divide: State of AI in Business 2025”MIT Media Lab（NANDA）——《生成式 AI 鸿沟：2025 企业 AI 现状》report PDF报告 PDF 95% of enterprise GenAI pilots produced no measurable P&L impact.95% 的企业生成式 AI 试点没有可衡量的损益影响。
  13. Shopify — Tobi Lütke internal memo, April 2025Shopify —— Tobi Lütke 内部信，2025 年 4 月memo内部信 Reflexive AI use as a baseline expectation; teams must show AI can't do the job before adding headcount.「本能地使用 AI」成为基本要求；增编前须证明 AI 做不了这份工作。
  14. John Boyd — “The Essence of Winning and Losing”, 1996John Boyd ——《胜负的本质》，1996briefing讲义 The OODA loop; speed of orientation as the real advantage.OODA 循环；真正的优势在于「定向」的速度。
  15. BCG — “Where's the Value in AI?”, 2024BCG ——《AI 的价值在哪里？》，2024report报告 Around three quarters of companies struggle to scale value from AI beyond pilots.约四分之三的公司难以让 AI 的价值走出试点、规模化落地。

Figures are quoted from the sources above as published; report editions get revised, so open the original before you put a number on a slide. That habit is, itself, day one of the drill. 以上数字均引自上述已发布的来源；报告会有修订，把数字放上 PPT 之前，请先打开原文。这个习惯本身，就是训练的第一天。