"争论是战争" vs "争论是舞蹈"。前者让你想赢、想压倒；后者让你想共舞、想配合。语言只是表层，真正被替换的是你对整件事的感知和你能采取的动作。
场景 · BigCat
① AI／认知科学："心智是计算机"是一个统治了半个世纪的隐喻——它让"记忆""存储""带宽""运行程序"显得天经地义，却悄悄抹掉了情绪、身体与具身。换成"大脑是预测机器"或"心智是生态系统"，研究方向立刻不同。② 你做的每一次技术抽象，都是一次隐喻选择："神经网络""注意力""训练"，既照亮也遮蔽。意识到自己正在用哪个隐喻思考，是 AI 超级个体的一项元能力。
English Summary
Conceptual Metaphor (Lakoff & Johnson) — metaphor is not decoration of language but the structure of thought itself. We can barely think about abstractions (time, argument, ideas) directly; we map them from a concrete source domain onto an abstract target domain. "ARGUMENT IS WAR": we attack positions, defend claims, win or lose. "TIME IS MONEY": we spend, save, waste it. The mapping is systematic yet invisible. Key consequences: (1) the metaphor isn't in the words, it's in the cognitive stance — it pre-decides what you can see and do; (2) change the metaphor and you change the entire space of available actions — imagine "ARGUMENT IS DANCE" and there are no winners, only coordination; (3) this is core evidence for embodied cognition — abstract thought is grounded in sensorimotor experience ("important is big," "intimate is close"). The mind is not a disembodied symbol processor but a body shaped by metaphor.
AI Prompts
中文提示词
我正在思考／讨论 [某个抽象问题或处境]。请帮我做一次「概念隐喻审计」：
① 找出我和别人正在用哪些隐喻来谈它（如"战争""旅程""市场""机器"）；
② 每个隐喻照亮了什么、又悄悄遮蔽了什么？
③ 给我 2 个全新的源域隐喻，并说明换上它之后，会浮现哪些原来看不见的行动选项。
English Prompt
I'm thinking through / discussing [an abstract problem or situation]. Run a "conceptual metaphor audit":
1. Identify the metaphors I and others are using to frame it (e.g. war, journey, market, machine).
2. For each, what does it illuminate and what does it quietly hide?
3. Give me 2 fresh source-domain metaphors and explain which previously-invisible options each one would surface.
① 技术叙事：同一项 AI 技术，框架成"AI 取代人类"还是"AI 增强人类、成就超级个体"，会左右监管、投资与采用的整个走向——叙事之争其实是框架之争。② 工程评审：把"最终一致性"框成"弱一致"（听着像缺陷）还是"高可用、可扩展"（听着像优势），常常在同样的底层机制上得出相反结论。识破这一点，你就能把争论从"谁的词更动听"拉回"底层权衡到底是什么"。
English Summary
Framing Wars — a frame is the default structure for understanding something, and the real battlefield is usually not the facts but which frame becomes default. "Tax relief" pre-loads an entire frame: tax is an affliction, relief is rescue, the taxer a villain — your stance is decided before any argument. The most counter-intuitive law: negating a frame activates it. Tell someone "don't think of an elephant" and they picture one; argue against the opponent's term and you reinforce it each time. The real victory is not winning inside their frame but making your frame the unthinking default — "estate tax" vs "death tax," "membership fee" vs "fine." Same root as anchoring, but deeper: anchoring shifts a number, framing shifts the whole problem space. Frames are neutral tools — usable to reveal or to manipulate — so spotting framing wars is first of all a defense: seeing which frame is doing your thinking for you.
AI Prompts
中文提示词
这是一场关于 [某议题／某产品叙事／某技术辩论] 的争论，相关说法是：[贴上对方的关键措辞]。请帮我打这场框架战争：
① 拆解对方的措辞里预设了哪套框架，把什么当成了理所当然；
② 提醒我哪些"反驳"其实会激活并加固对方的框架；
③ 给我一套自己的命名／框架，让我不在对方的战场上应战，并说明它把焦点引向哪里。
English Prompt
Here's a debate about [an issue / product narrative / technical argument]; the key wording in play is: [paste the opponent's phrasing]. Help me fight this framing war:
1. Unpack which frame their wording presupposes and what it treats as a given.
2. Warn me which "rebuttals" would actually activate and reinforce their frame.
3. Give me my own naming / frame so I don't fight on their battlefield, and explain where it redirects attention.
① 工程：与其问"系统为什么这么慢"（陷入无尽的性能调优），不如重新框定为"哪个 SLO 没被满足、对谁重要"——常常发现根本不必让全局变快。把"模型不够准"重新框定为"任务边界划错了"，往往比再标十万条数据更有效。② 佛学：你深感兴趣的"转念"与般若智慧，内核就是重新框定——同一个境遇，换一个观看它的框架，整片心理解空间随之改变。把无常从"威胁"重新框定为"事物本来的流动"，痛苦的结构就松动了。
English Summary
Reframing — the constructive use of framing: facing a stuck situation, don't try harder inside the old frame; change the frame organizing the problem so new options appear. The classic "slow elevator problem": tenants complain it's slow, every speed-up is too expensive — until someone reframes "the elevator is too slow" into "the wait is too boring" and installs mirrors. The elevator never sped up; the problem changed. Non-trivial: (1) the frame determines the entire solution space — under "speed up," mirrors aren't even a candidate; reframe and they become optimal. That's why reframing is the highest-leverage move — it shifts the space the solutions live in (cf. leverage points, Day 3: the highest leverage is changing the paradigm). (2) It's not spin: the same facts must still hold under the new frame, only reorganized — "positions" into "interests," "split the pie" into "grow the pie." (3) Shadow side: reframing that slides into denying the real problem becomes toxic positivity. The discipline: reframe to open options, not to evade reality.
AI Prompts
中文提示词
我被一个问题卡住了：[描述处境 + 我现在是怎么定义这个问题的]。我在原框架里已经试过 [已试过的方案]，都不行。请帮我重新框定：
① 我现在隐含的问题框架是什么？它把什么默认成了"必须如此"？
② 给我 3 个不同的重新框定，每个都打开一片新的解空间；
③ 提醒我哪些重新框定其实是在逃避真问题，而非打开选项。
English Prompt
I'm stuck on a problem: [describe the situation + how I currently define the problem]. Inside the old frame I've already tried [attempted solutions], none worked. Help me reframe:
1. What is my implicit problem frame, and what does it assume "must be so"?
2. Give me 3 different reframings, each opening a new solution space.
3. Flag which reframings would actually be evading the real problem rather than opening options.
① AI："训练"模型是一个死隐喻，它让 RLHF 听起来像驯兽或教书，掩盖了底层不过是梯度下降在高维曲面上的统计拟合。复活这个隐喻，你对"对齐""模型会不会骗人"的直觉会整个改变——你不是在管教一个学生，而是在塑形一片曲面。技术行话几乎全是死隐喻："云""内存""病毒""防火墙""神经网络"，每一个都在悄悄替你设定预设。② 意识／佛学：你关注的"无我"，恰恰是在复活一个最大的死隐喻——"自我"。把名词"我"还原成它本来的动词性过程（不断生灭的觉知之流），"我究竟是谁"这个问题就被重新打开了。
English Summary
Dead Metaphor — metaphors have a life cycle: vivid at first, then used so often they "die" into literal meaning. "The leg of a table," "the foot of a mountain," "grasp an idea," "the heart of the matter" — no one feels these as figurative anymore. A dead metaphor still carries its original mapping, but the mapping has gone invisible and is taken as reality itself. That's why dead metaphors are far more dangerous than live ones: because they no longer look like metaphors, they escape scrutiny. "Time is money" is so dead we don't notice it shaping our whole relationship to time (idleness as "waste," life as a "ledger"). The cure is to revive the dead metaphor — treat it as a figure of speech again and its smuggled assumptions resurface, open to challenge (same root as "the map is not the territory," Day 43). Worst offender: nominalization — freezing a process into a thing ("self," "mind," "memory"), which manufactures pseudo-questions like "where is my true self?"
AI Prompts
中文提示词
这是我在 [某领域／某段思考／某篇文字] 里反复使用的一组核心词：[列出 3-6 个词]。请帮我做一次「死隐喻复活」：
① 指出其中哪些其实是死隐喻，它们原本的源域映射是什么；
② 每个死隐喻偷偷夹带了哪些预设，正被我当成"现实本身"？
③ 复活或替换其中一两个，看看我对这件事的直觉会如何改变。
English Prompt
Here is a set of core words I keep using in [a domain / a line of thinking / a piece of writing]: [list 3-6 words]. Run a "dead-metaphor revival":
1. Point out which are actually dead metaphors and what their original source-domain mapping was.
2. What assumptions does each smuggle in that I'm now treating as "reality itself"?
3. Revive or replace one or two of them and show how my intuition about the matter would shift.