谭亮Tan Liang LIANG JIAN · 亮见

亮见活动LIANGJIAN EVENTS

与同路人，面对面Face to Face, with Fellow Travelers

思想在纸面上成型，在现场发生。
这里记录亮见的线下足迹：亮见沙龙、企业家闭门交流会、公开课与演讲——每一场的影像与要点。 Ideas take shape on paper; they happen in the room.
This page keeps the record of Liang Jian's live footprint — salons, closed-door forums, open lectures and keynotes, with scenes and takeaways from each.

亮见沙龙北京场 全场 亮见沙龙 · 首场LIANGJIAN SALON · NO.1
2026 · 09 · 08 北京 · 融科资讯中心Beijing · Raycom Infotech Park 亮见 × ttc 联合出品Liang Jian × ttc

谭亮教授 × 企业家闭门交流会Prof. Tan Liang × Entrepreneurs Closed-Door Forum

一场不设标准答案的深度对话 · 近 30 位企业家 · 三小时：AI 时代，一号位的进化地图A deep dialogue without standard answers — nearly 30 entrepreneurs, three hours: the evolution map for No.1 leaders in the AI era

白板手绘 Gartner 曲线 谭亮演讲特写 讲台分享 小组讨论 小组讨论 全场视角 全员合影

当晚主线 · 文章要点KEY THREADS OF THE NIGHT

  1. 先看大势：K 型分化正在走向小写 i——AI 先把企业拉到 80 分，但 80 分是一根线；没有变化本身就是最大的风险。AI 治理从「合不合规」进入「合不合法」。The big picture: from capital-K divergence to a small-i world — AI lifts every firm to 80 points, and that line is the real risk. AI governance has moved from compliance to legality.
  2. 管好「三天」：今天是流水，明天是战略，后天是长期不变的事。AI 时代节奏调整为看三年、想一年、做一季，以 90 天为最小闭环。Manage "three days": today is cash flow, tomorrow is strategy, the day after is what never changes. In the AI era: see three years out, think one year, act in 90-day loops.
  3. A2A 判断：商业模式全面走向 Agent to Agent。大厂同一时间段抢的是「入口」——请回答：你的行业里，AI 时代的入口在哪里？你离它有多远？The A2A thesis: business models are moving to Agent-to-Agent. The giants are racing for the gateway — ask: where is the AI-era gateway in your industry, and how far are you from it?
  4. 1391 地图：1 次范式重构（人和人、人和 Agent、Agent 和 Agent 三种关系），3 大挑战（战略、组织、人），9 问，回到一号位自己的领导力。The 1391 map: one paradigm shift (human-human, human-Agent, Agent-Agent), three systemic challenges, nine questions — and back to the leader's own leadership.
  5. 决策力三层突破：提升输入的质量（多样性、真实性、及时性），让输出闭环，验证假设——把隐性的假设显性化，定期检验。Three layers of decision-making: upgrade input quality (diversity, authenticity, timeliness), close the output loop, and make hidden assumptions explicit — then verify them.
  6. 崩塌生境：客体世界在崩塌，回到主体世界盖楼。一号位真正的领导力是元认知——AI 时代最稀缺的是判断力，判断力背后是专注。Rebuild within: as the external world collapses, return to the inner world. True leadership is meta-cognition — and in the AI era, judgment is scarce; behind judgment sits focus.
阅读全文：《亮见讲台｜AI 时代，一号位的进化地图》（谭亮知行录 · 2026 年 9 月 15 日）↗Read the full recap: "The No.1 Leader's Evolution Map in the AI Era" (Tan Liang's WeChat, Sep 15, 2026) ↗
亮见沙龙上海场 全场 亮见沙龙 · 第二期LIANGJIAN SALON · NO.2
2026 · 09 · 16 上海 · 安福路 201 号Shanghai · Anfu Road 亮见 × ttc 联合出品Liang Jian × ttc

谭亮教授 × 企业家闭门交流会Prof. Tan Liang × Entrepreneurs Closed-Door Forum

一块白板，一支笔，三小时——「向下三层」：AI 时代的一号位决策力模型 · 定向邀请 10–15 位企业创始人与 CEOOne whiteboard, one marker, three hours — "Three Layers Down": the decision-making model for No.1 leaders in the AI era, with an invited circle of 10–15 founders and CEOs

白板讲授「向下三层」 三小时完整板书 讲台分享 现场讲授 小组复盘「四种真实」 与学员合影 现场问答 全员合影

当晚主线 · 文章要点KEY THREADS OF THE NIGHT

  1. SEI：只有进化者：AI 时代没有大师，只有进化者。拉开差距的是人自己的三种能力——Sensing 感知、Judging 决断、Iterating 迭代；SEI 落到一号位身上，就是「向下三层」。SEI: only evolvers: in the AI era there are no masters, only evolvers. What sets people apart is Sensing, Judging and Iterating — and for a No.1 leader, SEI unfolds into the "Three Layers Down."
  2. 第一层 IPO：输入的质量决定输出的质量。破三关——破「单一」：把反对制度化、渠道多元、给讨论配上多空两个 Agent；破「延迟」：数据大屏加 T3 三大指标；破「听不到真话」：分清一手信息与二手结论，让原音重现。Layer 1 · IPO: output quality hinges on input quality. Break three bottlenecks — single-source input (institutionalize dissent, diversify channels, add bull-and-bear Agents), latency (data dashboards plus your T3 metrics), and filtered truth (first-hand signals over second-hand conclusions).
  3. 第二层 OFR：决策做完，工作才算开始。用 Output–Feedback–Result 闭环校验决策质量；深度复盘有两层——调参与换框，复的是自己背后的认知与假设；Agent 可以是你的镜子。Layer 2 · OFR: a decision is the starting line, not the finish. Close the Output–Feedback–Result loop; deep review works at two levels — re-tuning parameters and re-framing your own assumptions — and an Agent can be your mirror.
  4. 第三层 VDA：验证双重假设。个人假设——说出「高质量断言」：基于什么、判断什么、希望看到什么；团队假设——集体拆解关键变量，收敛到三到五个，画出箭头关系，找到那个第一性变量。Layer 3 · VDA: validate dual assumptions. Personal — state a high-quality assertion: what you base on, what you judge, what you expect. Team — surface the critical variables together, converge on three to five, map their arrows, and find the first-principles variable.
  5. 现场真问题：AI 解决不了人性问题，但会放大人性——技术是技术，责任是人。领导力回归主体世界、建立元认知；人机协同分 HITL 与 HOTL 两条路，FDE（前线部署工程师）正在长成新角色；「蒸馏 CEO」之后，组织走向哑铃型——上下文，是一号位最重要的资产。Real questions from the floor: AI cannot solve human nature, but it amplifies it — technology is technology; accountability is human. Leadership returns to the inner world as meta-cognition. Human-AI collaboration splits into HITL and HOTL paths, with the FDE (forward-deployed engineer) emerging; after the "distilled CEO," organizations turn dumbbell-shaped — and context becomes the leader's most important asset.
  6. 四种真实 × 行动卡：四个小组复盘同一题——过去两年最重要的一个决策：当时看见了什么，今天又看见了什么。散场前，每人写下一张行动卡：90 天，用这三层，改变一个具体的决策。Four kinds of truth × action cards: four groups worked one question — your most important decision of the past two years: what did you see then, and what do you see now? Everyone left with an action card: in 90 days, use the three layers to change one concrete decision.
阅读全文：《亮见讲台｜向下三层：AI 时代的一号位决策力模型》（谭亮知行录 · 2026 年 9 月 20 日）↗Read the full recap: "Three Layers Down: The No.1 Leader's Decision Model in the AI Era" (Tan Liang's WeChat, Sep 20, 2026) ↗

亮见沙龙将持续举办，更多场次与回顾，陆续记录于此。Liang Jian Salon continues — future sessions and recaps will be recorded here.