🌍 世界数字教育网
模块 11 · 思维方法Module 11 · Thinking Methods

批判性思维与信息鉴别

Critical Thinking & Information Literacy

答案无限廉价,判断答案好坏的能力反而稀缺。

Answers are cheap; judging them is the rare skill.

🧠 12 节 · 6 概念 · 10 提示词 · 12 谬误 · 10 框架 · 6 误区 · 50 问答 · 30 金句 · 30 技巧 · 10 工具 · 60 判断 · 40 案例 · 自测 · 清单 · FAQ · 来源12 sections · 6 concepts · 10 prompts · 12 fallacies · 10 frameworks · 6 myths · 50 Q&A · 30 golden · 30 techniques · 10 tools · 60 drills · 40 cases · quiz · checklist · FAQ · sources
C1

AI 时代,为什么批判性思维比以往更重要

Why Critical Thinking Matters More Than Ever in the AI Era

AI 让『答案』变得无限廉价,但『判断答案好坏』的能力反而更稀缺。批判性思维是 AI 时代最底层的能力。
AI makes 'answers' infinitely cheap, but the ability to 'judge answer quality' is scarcer. Critical thinking is the most foundational skill in the AI era.
⚠️AI 时代的新现实:信息爆炸 + AI 批量生产内容 = 真真假假混在一起。你每天接收的信息,80% 可能是 AI 生成的(文章、图片、视频、语音)。不训练判断力,你就是在『吞下』别人喂给 AI、AI 再喂给你的内容。The new reality of the AI era: information explosion + AI mass-produced content = a mix of true and false. 80% of the info you consume daily may be AI-generated (articles, images, videos, audio). Without training judgment, you're 'swallowing' content AI made and fed back to you.

批判性思维(Critical Thinking)= 不盲从、不轻信、能验证、能推理、能下结论的能力。它不是『抬杠』,而是『对信息保持健康的怀疑 + 用方法去验证』。在 AI 时代,它分为 3 层:① 判断信息来源(谁说的、可信吗);② 判断信息内容(逻辑通吗、证据够吗);③ 判断 AI 输出(AI 说的对吗、哪里错了)。

Critical Thinking = the ability to not follow blindly, not trust easily, verify, reason, and conclude. It's not 'arguing', but 'keeping healthy skepticism about info + using methods to verify'. In the AI era, it has 3 layers: ① judge the source (who said it, is it credible); ② judge the content (is the logic sound, is evidence enough); ③ judge AI output (is AI right, where is it wrong).

为什么 AI 时代更需要它?3 个原因:① AI 会『一本正经地胡说』(幻觉),你无法从语气判断真假;② AI 生成内容成本趋近于零,虚假信息可以规模化生产;③ AI 推荐系统让你陷入『信息茧房』,只看到你想看的。批判性思维是你在 AI 时代『不被带节奏』的底层能力。

Why more needed in the AI era? 3 reasons: ① AI 'speaks nonsense confidently' (hallucination); you can't judge truth by tone; ② AI content generation cost approaches zero, false info can be mass-produced; ③ AI recommendation traps you in 'info cocoons', only showing what you want to see. Critical thinking is the foundational skill to 'not be led around' in the AI era.

C2

6 大核心概念:批判性思维的底层能力

6 Core Concepts: The Foundation of Critical Thinking

批判性思维不是『一个能力』,而是 6 个可训练的底层能力。逐个掌握,你就有了完整的判断框架。
Critical thinking is not 'one skill' but 6 trainable foundational abilities. Master each and you have a complete judgment framework.
概念Concept是什么What it is为什么重要Why it matters怎么练How to train
怀疑(Skepticism)怀疑(Skepticism)对任何『未验证』的信息保持健康怀疑,不轻信Keep healthy skepticism for any 'unverified' info, don't trust easily怀疑是批判性思维的起点:不怀疑,就不会去验证Skepticism is the starting point of critical thinking: without it, you won't verify每天问『这是真的吗?怎么验证?』Ask daily: 'Is this true? How to verify?'
验证(Verification)验证(Verification)用方法去核实信息的真伪:交叉信源、查原始出处、试逻辑Use methods to verify info truth: cross-sources, check original, test logic验证把『怀疑』变成『确认』,是判断真伪的核心动作Verification turns 'skepticism' into 'confirmation', the core of truth judgment用 3 个独立信源交叉验证Cross-verify with 3 independent sources
推理(Reasoning)推理(Reasoning)从前提推出结论的能力,识别逻辑漏洞Ability to derive conclusions from premises, identify logic gapsAI 会给你『看起来对』的推理,你要能识别逻辑漏洞AI gives 'plausible-looking' reasoning; you must spot logic gaps练识别 10 种常见逻辑谬误Practice spotting 10 common logical fallacies
证据(Evidence)证据(Evidence)判断『证据是否足够支撑结论』,区分事实/观点/推测Judge if evidence supports conclusion; distinguish fact/opinion/speculationAI 常『结论先行、证据凑数』,你要能识别AI often 'conclusion first, evidence padded'; you must spot it问『证据是什么?谁提供的?可重复吗?』Ask: 'What's the evidence? Who provided? Reproducible?'
反思(Reflection)反思(Reflection)对自己的判断保持怀疑:我的结论可能错吗?Keep skepticism toward your own judgment: could my conclusion be wrong?批判性思维的对象包括『自己』,不只是别人Critical thinking's target includes 'yourself', not just others下结论前问『我可能错在哪?』Before concluding, ask: 'Where might I be wrong?'
元认知(Metacognition)元认知(Metacognition)监控自己的思维过程:我是怎么得出这个结论的?Monitor your own thinking: how did I reach this conclusion?元认知让你『看见』自己的思维陷阱,是最高层能力Metacognition lets you 'see' your own thinking traps; the highest-level ability写『判断日志』:记录我怎么判断的Write a 'judgment log': how I judged
🎯6 个概念的关系:怀疑(起点)→ 验证(动作)→ 证据(标准)→ 推理(方法)→ 反思(对象扩展)→ 元认知(监控)。它们是层层递进的:从『怀疑别人』到『怀疑自己』,从『判断信息』到『判断自己的判断』。Relationship of the 6: Skepticism (start) → Verification (action) → Evidence (standard) → Reasoning (method) → Reflection (expanded target) → Metacognition (monitoring). They progress layer by layer: from 'doubting others' to 'doubting yourself', from 'judging info' to 'judging your judgment'.
C3

10 个『信息验证』提示词(直接复制用)

10 'Info Verification' Prompts (Copy & Use)

10 个让 AI 帮你做批判性验证的提示词,覆盖事实核查/逻辑检查/偏见检测。
10 prompts to let AI help you do critical verification, covering fact-check/logic check/bias detection.
场景Scenario提示词(中文)Prompt (ZH)提示词(英文)Prompt (EN)验证什么What it verifies
事实核查Fact-check『请核查以下说法是否属实,给出 3 个独立信源和它们的原始链接:[粘贴说法]。如果无法确认,请明确说『无法验证』。』'Verify if the following claim is true, give 3 independent sources and their original links: [paste claim]. If unable to confirm, explicitly say "unable to verify".'事实是否真实Whether the fact is true
查出处Trace source『这个说法最早出自哪里?找到原始出处(论文/新闻/官方数据),而不是二手转述。』'Where does this claim originally come from? Find the original source (paper/news/official data), not secondary retelling.'原始出处Original source
逻辑检查Logic check『请检查以下论证的逻辑:列出所有前提和结论,指出逻辑漏洞(如有),判断结论是否必然成立。』'Check the logic of this argument: list all premises and conclusions, point out logic gaps (if any), judge if the conclusion necessarily holds.'逻辑是否严密Whether logic is rigorous
查偏见Bias check『分析这段内容可能存在的偏见:立场偏见、选择偏见、幸存者偏见。指出哪些是事实、哪些是观点、哪些是推测。』'Analyze possible biases in this content: stance bias, selection bias, survivorship bias. Point out which are facts, opinions, speculations.'内容偏见Content bias
反方论证Counter-argument『请为以下观点提出最强的 3 个反方论证(steelman),每个都要给出证据。然后评估哪一方证据更强。』'Provide the strongest 3 counter-arguments (steelman) for this view, each with evidence. Then assess which side has stronger evidence.'论点强弱Argument strength
查数据Check data『这个数据是哪年、哪个机构发布的?样本量多大?是横截面还是纵向研究?数据是否可重复?』'What year, which institution released this data? Sample size? Cross-sectional or longitudinal? Is data reproducible?'数据可靠性Data reliability
查时效Check timeliness『这个信息是什么时候的?现在(2026)是否仍然成立?有没有被新信息推翻?』'When was this info? Does it still hold now (2026)? Has it been overturned by newer info?'信息时效Info timeliness
AI 自查AI self-check『你刚才的回答,哪些部分是你『确定知道』的,哪些是『推测』的,哪些可能『编造』的?分别标注。』'In your previous answer, which parts do you 'know for sure', which are 'speculation', which might be 'fabricated'? Label each.'AI 回答可信度AI answer credibility
查利益Check interest『提出这个观点的机构/人,有什么利益动机?他们从「你相信这个」里能获益什么?』'What's the interest motive of the institution/person making this claim? What do they gain from 'you believing this'?'利益动机Interest motive
综合判断Comprehensive judge『综合以上所有验证结果,给出一个 0-100 的可信度评分,并说明:哪些证据支持、哪些反对、最大的不确定性在哪。』'Synthesize all verification results, give a 0-100 credibility score, explain: which evidence supports, which opposes, where's the biggest uncertainty.'综合可信度Overall credibility
🔍用 AI 做批判性验证的『铁律』:永远不要让 AI 只给『是/否』,要让它给『证据 + 信源 + 置信度』。如果 AI 说不清证据,它就不可信。这也是检验 AI 自身幻觉的最好方法。'Iron rule' of using AI for critical verification: never let AI give only 'yes/no'; make it give 'evidence + sources + confidence'. If AI can't explain evidence, it's not credible. This is also the best way to test AI's own hallucination.
C4

12 种最常见的逻辑谬误(AI 时代高发)

12 Most Common Logical Fallacies (High-Frequency in AI Era)

12 种最常见、AI 时代最高发的逻辑谬误,识别它们 = 识别 90% 的『带节奏』。
12 most common, AI-era-high-frequency logical fallacies; identifying them = spotting 90% of 'being led around'.
谬误Fallacy是什么What it is例子Example怎么识别How to spot
稻草人(Straw man)稻草人(Straw man)歪曲对方观点,攻击歪曲后的版本Distort the other's view, attack the distorted versionA:我觉得 AI 有用。B:你是说 AI 能取代所有人?A 没这么说A: I think AI is useful. B: So you mean AI can replace everyone? A never said that问『他原话是什么?』Ask: 'What was his original wording?'
滑坡(Slippery slope)滑坡(Slippery slope)A 会导致 B,B 导致 C... 一路推到灾难,中间环节没证据A leads to B, B leads to C... all the way to disaster, middle steps have no evidence允许用 AI 写作业 → 没人会思考 → 人类灭亡。中间每一步都没证据Allowing AI homework → no one thinks → humanity ends. Each middle step has no evidence问『每一步都有证据吗?』Ask: 'Does each step have evidence?'
诉诸权威(Appeal to authority)诉诸权威(Appeal to authority)『某某专家说了,所以是对的』,但专家可能说错/不在其领域'A certain expert said so, therefore it's right', but the expert may be wrong/out of field某明星说『某保健品治癌』——明星不是医学专家A celebrity says 'a certain supplement cures cancer' — the celebrity isn't a medical expert问『他在这个领域是权威吗?证据呢?』Ask: 'Is he an authority in this field? What's the evidence?'
以偏概全(Hasty generalization)以偏概全(Hasty generalization)用个例推出普遍结论Use one case to derive a universal conclusion我认识 3 个用 AI 的人都没变聪明,所以 AI 没用。3 个样本不代表全体I know 3 people who use AI and none got smarter, so AI is useless. 3 samples don't represent the whole问『样本量多大?有代表性吗?』Ask: 'How big is the sample? Is it representative?'
非黑即白(False dilemma)非黑即白(False dilemma)把选项简化成两个,其实还有中间选项Reduce options to two, when there are actually intermediate options你要么全人工,要么全 AI。其实可以混合用Either fully manual or fully AI. Actually you can mix问『真的只有两个选项吗?』Ask: 'Are there really only two options?'
诉诸情感(Appeal to emotion)诉诸情感(Appeal to emotion)用恐惧/同情/愤怒代替证据Use fear/sympathy/anger instead of evidence『你不用 AI 就会被时代抛弃!』(制造恐惧)'If you don't use AI, you'll be left behind by the era!' (manufacturing fear)问『抛开情绪,证据是什么?』Ask: 'Setting aside emotion, what's the evidence?'
循环论证(Circular reasoning)循环论证(Circular reasoning)结论当证据,证据当结论Use the conclusion as evidence, and the evidence as the conclusionAI 是可靠的,因为 AI 自己说自己可靠AI is reliable because AI itself claims to be reliable问『证据独立于结论吗?』Ask: 'Is the evidence independent of the conclusion?'
偷换概念(Equivocation)偷换概念(Equivocation)同一个词在论证中悄悄换了含义The same word secretly changes meaning in the argument『自由』:前一句指『言论自由』,后一句变成『想干嘛就干嘛』'Freedom': in one sentence means 'freedom of speech', in the next becomes 'do whatever you want'问『这个词前后含义一样吗?』Ask: 'Is this word's meaning the same before and after?'
诉诸无知(Appeal to ignorance)诉诸无知(Appeal to ignorance)『没人证明它是错的,所以它是对的』(或反之)'No one has proven it wrong, so it's right' (or the reverse)没人证明 AI 不会造反,所以它不会造反。『没证据』不等于『证明』No one has proven AI won't revolt, so it won't. 'No evidence' ≠ 'proof'问『是证明了,还是只是没发现?』Ask: 'Is it proven, or just not found yet?'
红鲱鱼(Red herring)红鲱鱼(Red herring)插入无关话题,转移注意力Insert an unrelated topic to divert attentionA:你的产品有 bug。B:但你公司裁员了 100 人,多不人道!A: Your product has bugs. B: But your company laid off 100 people, how inhumane!问『这跟原问题有关吗?』Ask: 'Is this related to the original question?'
合成谬误(Fallacy of composition)合成谬误(Fallacy of composition)部分成立,整体就成立(不一定)What holds for parts holds for the whole (not necessarily)每个球员都很强,所以这支球队一定赢。团队是整体Each player is strong, so this team must win. A team is a whole问『部分成立能推出整体吗?』Ask: 'Does what holds for parts hold for the whole?'
幸存者偏差(Survivorship bias)幸存者偏差(Survivorship bias)只看到『活下来』的,看不到『没活下来』的Only see the 'survivors', not the 'non-survivors'马斯克用 AI 成功了,所以 AI 一定让你成功。看不到失败者Musk succeeded with AI, so AI will surely make you succeed. You don't see the failure cases问『没成功的人呢?样本在哪?』Ask: 'What about those who didn't succeed? Where's the sample?'
⚠️AI 时代这些谬误更高发,因为:① AI 会『自信地』用这些谬误推理(它不知道自己在犯逻辑错误);② AI 批量生产内容,谬误可以规模化传播。识别它们不是抬杠,是保护自己不被带节奏。These fallacies are more frequent in the AI era because: ① AI reasons with these fallacies 'confidently' (it doesn't know it's making logic errors); ② AI mass-produces content, fallacies can spread at scale. Identifying them isn't arguing, it's protecting yourself from being led around.
C5

15 个『AI 输出』的红旗信号(见到就要警惕)

15 'Red Flag' Signals in AI Output (Be Alert When You See Them)

AI 输出的 15 个红旗信号,见到任意一个,就要启动批判性验证。
15 red-flag signals in AI output; see any one and start critical verification.
红旗Red flag表现How it shows为什么危险Why it's dangerous怎么应对How to respond
过度自信Overconfidence语气绝对:『这绝对是』『毫无疑问』Absolute tone: 'This is absolutely', 'Without a doubt'AI 的置信度和正确率无关,自信≠正确AI's confidence is unrelated to accuracy; confident ≠ correct问『你有多确定?证据呢?』Ask: 'How sure are you? What's the evidence?'
编造引用Fabricated citations给出看似真实的论文/书籍/数据,但查无此物Give seemingly real papers/books/data, but they don't exist幻觉的核心表现:AI『想象』出来源The core of hallucination: AI 'imagines' sources逐条验证引用是否真实Verify each citation is real
数据无出处Data without source给出具体数字,但说不出哪年、哪个机构Give specific numbers, but can't say which year, which institution数字是 AI 编的最常见形式Numbers are the most common form of AI fabrication问『这个数据哪来的?』Ask: 'Where does this data come from?'
逻辑跳跃Logic leap前提和结论之间有大缺口,用『显然』『显然可知』跳过Big gap between premise and conclusion, skipped with 'obviously', 'clearly'AI 用『显然』掩盖逻辑漏洞AI uses 'obviously' to cover logic gaps要求它列出每一步推理Require it to list each step of reasoning
以偏概全Hasty generalization用个例支撑普遍结论Use one case to support a universal conclusionAI 爱用『我见过/有案例』当证据AI loves using 'I've seen/a case' as evidence问『还有多少案例?』Ask: 'How many more cases are there?'
诉诸权威Appeal to authority反复提『专家说』『研究表明』但不给具体出处Repeatedly cite 'experts say', 'research shows' but give no specific source权威是 AI 最常搬的『大旗』Authority is the biggest 'banner' AI most often waves问『哪个专家?哪篇研究?』Ask: 'Which expert? Which study?'
回避反例Avoiding counterexamples只给支持结论的证据,对反例视而不见Only give evidence supporting the conclusion, ignore counterexamplesAI 默认『讨好』,只说你想听的AI defaults to 'please', only says what you want to hear主动问『反例是什么?』Actively ask: 'What are the counterexamples?'
答非所问Answering something else你的问题被悄悄『改写』成它更容易回答的版本Your question is quietly 'rewritten' into a version it can answer more easilyAI 回避难问题,偷换问题AI avoids hard questions, swaps the question核对它答的是不是你的原问题Check if it's answering your original question
过度具体Overly specific细节多到可疑:精确到个位的数字、不存在的日期Details so specific they're suspicious: numbers precise to the unit, non-existent datesAI 用『具体感』伪装真实性AI uses 'specificity' to fake authenticity对可疑细节单独验证Verify suspicious details individually
无差别支持Indiscriminate support无论你怎么问,它都支持你的立场No matter how you ask, it supports your stanceAI 的『讨好偏差』:镜像你的观点AI's 'people-pleasing bias': mirrors your view问『如果我是错的,你凭什么说对?』Ask: 'If I'm wrong, what's your basis for saying I'm right?'
术语堆砌Jargon stacking大量专业术语,但解释不清楚Massive professional jargon, but no clear explanation术语是 AI 的『遮羞布』,掩盖不懂Jargon is AI's 'fig leaf', covering its ignorance要求它用人话解释Require it to explain in plain language
时间模糊Time ambiguity『现在』『最近』『研究表明』但不说具体年份'Now', 'recently', 'research shows' but no specific yearAI 训练有截止时间,可能把旧信息当新的AI's training has a cutoff; it may treat old info as new问『这是哪年的信息?』Ask: 'What year is this info from?'
绝对化Absolutism『一定』『必然』『100%』等绝对词Absolute words like 'must', 'inevitably', '100%'现实中几乎没有 100% 确定的事There's almost nothing 100% certain in reality要求它给置信度和例外Require it to give confidence and exceptions
单一信源Single source所有结论都来自『同一个』来源(可能是 AI 自己)All conclusions come from 'the same' source (possibly the AI itself)AI 会把自己当信源AI treats itself as a source要求它给独立信源Require it to give independent sources
情感操控Emotional manipulation用『你知道吗』『震惊』『你绝对想不到』等钩子Hooks like 'Did you know', 'Shocking', 'You absolutely won't believe'AI 学自营销文案,会带节奏AI learned from marketing copy, can lead the rhythm剥离情感,只看事实Strip emotion, look at facts only
🚩15 个红旗信号的记忆口诀:『自信、编源、无据、跳逻辑、以偏概、诉权威、避反例、答非问、过具体、无差别、术语堆、时间糊、绝对化、单一源、情感操』。见到任意一个,先停下来验证,再相信。Mnemonic for the 15 red flags: 'confident, fabricated source, no evidence, logic leap, hasty general, appeal to authority, avoid counter, answer other, over-specific, indiscriminate, jargon stack, time vague, absolutism, single source, emotion manipulation'. See any one, stop and verify before believing.
C6

10 个『自我批判』的元问题(下结论前必问)

10 'Self-Critique' Meta-Questions (Must Ask Before Concluding)

下结论前问自己这 10 个问题,把批判性思维从『判断别人』升级到『判断自己』。
Ask yourself these 10 questions before concluding, upgrade critical thinking from 'judging others' to 'judging yourself'.
问题Question问什么What it asks为什么重要Why it matters
我可能错在哪?Where might I be wrong?列出你结论最薄弱的 1 个点List the 1 weakest point of your conclusion暴露盲点,避免过度自信Expose blind spots, avoid overconfidence
什么证据能推翻我?What evidence would overturn me?如果看到什么,你会改主意?If you saw what, would you change your mind?可证伪性是好判断的标志Falsifiability is a sign of good judgment
我的信息源单一吗?Is my source single?我是不是只看了『一边』?Am I only looking at 'one side'?单一信源 = 信息茧房Single source = info cocoon
我是否被情绪带着走?Am I being led by emotion?我是被愤怒/恐惧/同情驱动的吗?Am I driven by anger/fear/sympathy?情绪是判断的最大敌人Emotion is the biggest enemy of judgment
我的结论是事实还是观点?Is my conclusion a fact or opinion?我能验证它吗?Can I verify it?区分事实/观点是批判性思维基础Distinguishing fact/opinion is the foundation of critical thinking
我是否在『确认偏差』?Am I in 'confirmation bias'?我只看支持我的证据吗?Am I only looking at evidence supporting me?人天然只找『对的』证据Humans naturally only seek 'right' evidence
反方最强的论证是什么?What's the strongest counter-argument?steelman:替对方说最强的话steelman: speak the strongest on behalf of the other理解反方 = 理解全貌Understanding the counter = understanding the whole
我的样本有代表性吗?Is my sample representative?我看到的『个例』能代表全体吗?Can the 'case' I see represent the whole?以偏概全是常见谬误Hasty generalization is a common fallacy
这个信息还时效吗?Is this info still timely?它是什么时候的?现在还成立吗?When is it from? Does it still hold now?AI 信息可能过时AI info may be outdated
我为什么相信这个?Why do I believe this?我的判断基于证据,还是基于『我觉得』?Is my judgment based on evidence, or on 'I feel'?元认知:看见自己的判断过程Metacognition: see your own judgment process
🧠10 个元问题的用法:每次下重要结论前,拿出手机记一次这 10 个问题(30 秒),能显著降低『过度自信』和『确认偏差』。这是把批判性思维『内化』的最快方法。Usage of the 10 meta-questions: before every important conclusion, take out your phone and run through these 10 questions (30 seconds) — it significantly reduces 'overconfidence' and 'confirmation bias'. This is the fastest way to 'internalize' critical thinking.
C7

50 个『批判性思维』高频问题(一问一答)

50 'Critical Thinking' High-Frequency Questions (One Q One A)

50 个高频问题,覆盖认知/方法/工具/进阶,直接给答案。
50 high-frequency questions, covering cognition/method/tools/advanced, direct answers.
有根据地、系统地评估信息和论证,区分事实与观点、强论证与弱论证的能力。核心:不轻信、不武断、讲证据。The ability to evaluate information and arguments based on evidence, systematically, distinguishing fact from opinion, strong from weak arguments. Core: don't trust lightly, don't be dogmatic, demand evidence.
杠精:为了反对而反对,不关心对错,只要『赢』。批判性思维:为了『更接近真相』而检验,关心对错,承认自己可能错。动机不同,结果不同。Contrarian: opposes to oppose, doesn't care about right/wrong, just wants to 'win'. Critical thinking: tests to 'get closer to truth', cares about right/wrong, admits self may be wrong. Different motives, different outcomes.
部分天生(好奇心、怀疑倾向),但主要是训练的。研究显示,9 岁孩子就有批判性思维潜力,但大多数人的判断力在 25 岁后仍很差——因为没练过。Partly innate (curiosity, skepticism), but mostly trained. Studies show 9-year-olds have critical thinking potential, but most adults' judgment is still poor after 25 — because they haven't practiced.
是的。AI 让『答案』无限廉价,但『判断答案好坏』的能力反而更稀缺。信息过载 + AI 批量生产内容,让『不被带节奏』成为生存技能。Yes. AI makes 'answers' infinitely cheap, but the ability to 'judge answer quality' becomes scarcer. Info overload + AI mass production makes 'not being led' a survival skill.
不需要系统学逻辑学,但需要认识常见逻辑谬误(12 种最常见)。认得谬误 = 识别 90% 的『带节奏』。No need for systematic logic, but need to recognize common logical fallacies (12 most common). Recognizing fallacies = spotting 90% of 'being led'.
不会。批判性思维是『快速筛选 + 深度验证』:小事快速判断(60 分够用),大事深度验证(90 分必须)。它是『省力』的,不是『费力』的。No. Critical thinking is 'quick filter + deep verification': small things quick judgment (60 is enough), big things deep verification (90 must). It's 'energy-saving', not 'energy-consuming'.
不一样。怀疑一切 = 虚无主义,没有结论。批判性思维 = 有根据地怀疑,得出『在现有证据下最可能的结论』。它产出结论,不是只怀疑。Different. Doubting everything = nihilism, no conclusion. Critical thinking = grounded doubt, reaching 'the most likely conclusion given current evidence'. It produces conclusions, not just doubt.
减少『坏决策』。大多数坏决策不是因为信息不足,而是因为『判断方式』错(确认偏差、过度自信、情绪驱动)。批判性思维修正判断方式。Reduce 'bad decisions'. Most bad decisions aren't from lack of info, but from 'wrong judgment methods' (confirmation bias, overconfidence, emotion-driven). Critical thinking corrects judgment methods.
能显著降低。被洗脑的核心是『绕过理性』直接操纵情绪/身份认同。批判性思维训练『看到情绪操纵』的能力,让洗脑的难度大增。Significantly reduces it. The core of brainwashing is 'bypassing reason' to directly manipulate emotion/identity. Critical thinking trains the ability to 'see emotional manipulation', making brainwashing much harder.
非常有用。AI 幻觉的核心是『自信地编造』,而批判性思维的核心是『对自信保持警惕』。两者正面对抗。Very useful. The core of AI hallucination is 'confidently fabricating', and the core of critical thinking is 'staying wary of confidence'. They directly confront each other.
三个方法:① 每天『质疑一条信息』(找它的证据);② 每周『替反方论证一次』(steelman);③ 每月『复盘一个坏决策』(找判断漏洞)。Three methods: ① daily 'question one piece of info' (find its evidence); ② weekly 'argue for the other side once' (steelman); ③ monthly 'review a bad decision' (find judgment gaps).
让你的论证更有力。批判性思维训练的『前提—证据—结论』结构,正是好写作(议论文/报告/邮件)的骨架。Makes your arguments stronger. The 'premise—evidence—conclusion' structure trained by critical thinking is exactly the skeleton of good writing (argumentative essays/reports/emails).
非常有用。调试 = 批判性思维(假设—验证—排除)。代码审查 = 批判性思维(检查逻辑漏洞)。AI 生成代码时,批判性思维是『判断代码对不对』的核心。Very useful. Debugging = critical thinking (hypothesis—verify—eliminate). Code review = critical thinking (check logic gaps). When AI generates code, critical thinking is the core of 'judging if the code is right'.
不冲突,互补。创造力负责『生成新想法』,批判性思维负责『筛选好想法』。先发散(创造),后收敛(批判),两者交替。No conflict, complementary. Creativity 'generates new ideas', critical thinking 'filters good ideas'. Diverge first (create), converge later (criticize), alternate the two.
非常有用。投资最大的敌人是『情绪』和『确认偏差』(只看好自己持有的)。批判性思维训练『客观评估风险』,能显著降低『追涨杀跌』。Very useful. The biggest enemies of investing are 'emotion' and 'confirmation bias' (only bullish on what you hold). Critical thinking trains 'objective risk assessment', significantly reducing 'chasing highs, cutting lows'.
非常有用。医疗信息常被『过度营销』和『个例误导』。批判性思维帮你区分『证据等级』(RCT > 观察 > 个例),不被『某患者吃了就好了』带节奏。Very useful. Medical info is often 'over-marketed' and 'misled by individual cases'. Critical thinking helps distinguish 'evidence levels' (RCT > observational > individual case), not led by 'a patient took it and got better'.
有用。教孩子『为什么』而不是『听话』,培养他们自己的判断力。批判性思维的孩子更不容易被『网红带货』和『同学从众』带节奏。Useful. Teach kids 'why' not 'obey', cultivate their own judgment. Critical-thinking kids are less easily led by 'influencer marketing' and 'peer conformity'.
非常有用。职场决策(选方案、评估风险、判断优先级)都依赖批判性思维。它让你从『执行者』升级到『判断者』,是晋升的关键能力。Very useful. Workplace decisions (choosing options, assessing risk, judging priority) all rely on critical thinking. It upgrades you from 'executor' to 'judge', a key skill for promotion.
有用,但要小心。批判性思维帮你『理性评估关系』(是否健康、是否值得),但『不要把伴侣当论证对象』——关系需要的是『理解』,不是『赢』。Useful, but be careful. Critical thinking helps you 'rationally evaluate the relationship' (healthy, worth it), but 'don't treat your partner as an argument target' — relationships need 'understanding', not 'winning'.
批判性思维不反对信仰,但它训练『理性评估』信仰的依据。健康的信仰 = 经得起检验的信仰。批判性思维让信仰『更扎实』,不是『更动摇』。Critical thinking doesn't oppose faith, but trains 'rational evaluation' of faith's basis. Healthy faith = faith that withstands testing. Critical thinking makes faith 'more solid', not 'more shaky'.
不会。批判性思维是『对事不对人』:你质疑的是『信息』,不是『人』。会『好好质疑』的人更受欢迎,因为他们让讨论更有质量。No. Critical thinking is 'about things not people': you question 'info', not 'people'. People who 'question well' are more popular, because they make discussions higher quality.
不是二选一。先『让情绪被看见』(承认它),再『用理性评估』(证据是什么)。情绪是『信号』,理性是『判断』,两者配合,不是对抗。Not either-or. First 'let emotion be seen' (acknowledge it), then 'evaluate with reason' (what's the evidence). Emotion is 'signal', reason is 'judgment', they cooperate, not confront.
非常有用。好提示词 = 批判性思维的结构化表达:明确目标、给出约束、要求验证、提供反例。批判性思维越强,提示词越精准。Very useful. Good prompts = structured expression of critical thinking: clear goals, give constraints, require verification, provide counterexamples. The stronger your critical thinking, the more precise your prompts.
帮你区分『事实报道』和『观点评论』,识别『标题党』和『断章取义』。看到新闻先问:这是事实还是观点?证据是什么?信源可靠吗?Helps distinguish 'factual reporting' from 'opinion commentary', spot 'clickbait' and 'taken out of context'. When reading news, first ask: Is this fact or opinion? What's the evidence? Is the source reliable?
帮你识别『情绪钩子』和『制造焦虑』,减少『被带节奏』。刷到让你愤怒/恐惧的内容,先问:这是事实吗?它在操纵我什么情绪?Helps spot 'emotional hooks' and 'anxiety manufacturing', reducing 'being led'. When scrolling content that makes you angry/fearful, first ask: Is this fact? What emotion is it manipulating?
有用。帮你识别『饥饿营销』『虚假折扣』『从众心理』。看到『最后一天』『仅剩 3 件』,先问:这是真的吗?证据呢?Useful. Helps spot 'hunger marketing', 'fake discounts', 'herd mentality'. Seeing 'last day', 'only 3 left', first ask: Is this real? What's the evidence?
非常有用。帮你区分『真实岗位』和『虚假承诺』,评估『长期价值』而非『短期诱惑』。面试时批判性思维帮你『反问』,看清公司真实情况。Very useful. Helps distinguish 'real job' from 'false promises', evaluate 'long-term value' not 'short-term temptation'. In interviews, critical thinking helps you 'ask back', seeing the company's real situation.
非常有用。创业最大的敌人是『确认偏差』(只看好自己项目的证据)和『过度自信』。批判性思维帮你『客观评估风险』,避免『赌一把』。Very useful. The biggest enemy of entrepreneurship is 'confirmation bias' (only looking for evidence favoring your project) and 'overconfidence'. Critical thinking helps you 'objectively assess risk', avoiding 'gambling'.
帮你从『死记硬背』升级到『理解 + 质疑』。批判性思维的学习者不满足于『记住答案』,而是追问『为什么』『证据是什么』,学得更深。Helps upgrade from 'rote memorization' to 'understanding + questioning'. Critical-thinking learners don't settle for 'memorizing answers', but ask 'why', 'what's the evidence', learning deeper.
有用,尤其作文/论述题。批判性思维帮你『审题』(找隐含假设)和『论证』(给出证据 + 反例),比『套模板』得分更高。Useful, especially essays/argument questions. Critical thinking helps 'read the question' (find hidden assumptions) and 'argue' (give evidence + counterexamples), scoring higher than 'template-filling'.
非常有用。辩论的核心是『识别对方论证的漏洞』+『构建自己有力的论证』。批判性思维是辩论的底层能力。Very useful. The core of debate is 'spotting gaps in the opponent's argument' + 'building your own strong argument'. Critical thinking is the foundational ability of debate.
有用。帮你识别『对方的诉求』和『隐藏假设』,避免『被带节奏』。谈判前用批判性思维『预演』对方可能的论证。Useful. Helps spot 'the other's demands' and 'hidden assumptions', avoiding 'being led'. Before negotiation, use critical thinking to 'rehearse' the other's possible arguments.
非常有用。法律的核心是『证据』+『逻辑』。批判性思维训练『证据评估』和『逻辑漏洞识别』,是法律职业的核心能力。Very useful. The core of law is 'evidence' + 'logic'. Critical thinking trains 'evidence evaluation' and 'logic gap spotting', a core ability of legal professions.
非常有用。科学方法论的核心就是批判性思维:假设—验证—证伪—修正。科学 = 制度化的批判性思维。Very useful. The core of scientific methodology is critical thinking: hypothesis—verify—falsify—revise. Science = institutionalized critical thinking.
有用,但边界不同。艺术更重『感受』,但『艺术批评』需要批判性思维(评估作品的技术/观念/价值)。创作时靠感受,评价时靠批判。Useful, but with different boundaries. Art weighs 'feeling' more, but 'art criticism' needs critical thinking (evaluate the work's technique/concept/value). Creation relies on feeling, evaluation relies on criticism.
创作时靠感受,评价时靠批判。听一首曲子的『结构』『和声』『主题发展』需要批判性思维。Creation relies on feeling, evaluation on criticism. Listening to a piece's 'structure', 'harmony', 'theme development' needs critical thinking.
训练时靠身体,战术时靠批判。教练的『读比赛』(评估对手、调整策略)是批判性思维的应用。Training relies on body, tactics on criticism. Coaches' 'reading the game' (assessing opponents, adjusting strategy) is critical thinking application.
非常有用。策略游戏(象棋/围棋/RTS)的核心是『批判性思维』(评估局势、预测对手、调整策略)。Very useful. The core of strategy games (chess/Go/RTS) is 'critical thinking' (assessing situations, predicting opponents, adjusting strategies).
哲学 = 制度化的批判性思维。苏格拉底的『追问法』就是批判性思维的原型。Philosophy = institutionalized critical thinking. Socrates' 'questioning method' is the prototype of critical thinking.
批判性思维不反对宗教,但它训练『理性评估』信仰的依据。健康的宗教 = 经得起检验的信仰。Critical thinking doesn't oppose religion, but trains 'rational evaluation' of faith's basis. Healthy religion = faith that withstands testing.
非常有用。政治信息常被『情绪操控』和『断章取义』。批判性思维帮你『识别宣传』,不被『带节奏』。Very useful. Political info is often 'emotionally manipulated' and 'taken out of context'. Critical thinking helps you 'spot propaganda', not being 'led'.
非常有用。广告的核心是『情绪操控』。批判性思维帮你『剥离情绪,只看事实』,识别『虚假宣传』。Very useful. The core of advertising is 'emotional manipulation'. Critical thinking helps you 'strip emotion, look at facts only', spot 'false advertising'.
有用。营销既需要『识别情绪钩子』(防御),也需要『构建有力论证』(进攻)。批判性思维是营销的双刃剑。Useful. Marketing needs both 'spotting emotional hooks' (defense) and 'building strong arguments' (offense). Critical thinking is a double-edged sword of marketing.
有用。帮你『识别客户的真实需求』(而非表面需求),避免『过度承诺』。销售前用批判性思维『预演』客户可能的质疑。Useful. Helps you 'spot the customer's real needs' (not surface needs), avoiding 'over-promising'. Before sales, use critical thinking to 'rehearse' the customer's possible questions.
有用。帮你『快速定位问题』(批判性排查),而非『机械回复』。客服遇到复杂投诉时,批判性思维帮你『找到根本原因』。Useful. Helps you 'quickly locate the problem' (critical troubleshooting), not 'mechanical replies'. When customer service faces complex complaints, critical thinking helps you 'find the root cause'.
非常有用。项目管理的核心是『风险识别』+『优先级判断』。批判性思维帮你『评估风险』和『做决策』,避免『救火式管理』。Very useful. The core of project management is 'risk identification' + 'priority judgment'. Critical thinking helps you 'assess risk' and 'make decisions', avoiding 'firefighting management'.
非常有用。领导力的核心是『做决策』+『判断人』。批判性思维帮你『客观评估』决策和人选,避免『拍脑袋』。Very useful. The core of leadership is 'making decisions' + 'judging people'. Critical thinking helps you 'objectively assess' decisions and candidates, avoiding 'gut decisions'.
让你的论证更有力。批判性思维训练的『前提—证据—结论』结构,正是好写作的骨架。Makes your arguments stronger. The 'premise—evidence—conclusion' structure trained by critical thinking is exactly the skeleton of good writing.
有用。帮你『构建有力的论证』+『预判听众的质疑』。演讲前用批判性思维『预演』听众可能的反驳。Useful. Helps you 'build strong arguments' + 'predict the audience's questions'. Before speeches, use critical thinking to 'rehearse' the audience's possible rebuttals.
非常有用。教学的核心是『启发学生思考』,而非『灌输答案』。批判性思维的教师会『追问』,引导学生自己得出结论。Very useful. The core of teaching is 'inspiring students to think', not 'pouring in answers'. Critical-thinking teachers 'follow up', guiding students to reach conclusions themselves.
非常有用。新闻的核心是『客观』+『事实』。批判性思维帮助记者『核实信源』和『识别偏见』,写出更可靠的报道。Very useful. The core of journalism is 'objectivity' + 'facts'. Critical thinking helps journalists 'verify sources' and 'spot bias', writing more reliable reports.
非常有用。学术 = 制度化的批判性思维。论文的核心是『论证』+『证据』,批判性思维是学术的底层能力。Very useful. Academia = institutionalized critical thinking. The core of papers is 'argument' + 'evidence', critical thinking is the foundational ability of academia.
非常有用。数据科学的陷阱是『相关≠因果』+『幸存者偏差』。批判性思维帮你『识别数据陷阱』,避免『被数据带节奏』。Very useful. The trap of data science is 'correlation ≠ causation' + 'survivorship bias'. Critical thinking helps you 'spot data traps', avoiding 'being led by data'.
非常有用。ML 的陷阱是『过拟合』+『数据偏差』+『黑盒』。批判性思维帮你『评估模型可靠性』,避免『被模型带节奏』。Very useful. ML's traps are 'overfitting' + 'data bias' + 'black box'. Critical thinking helps you 'evaluate model reliability', avoiding 'being led by models'.
非常有用。AI 安全的核心是『评估风险』+『识别对齐问题』。批判性思维帮你『看到 AI 的盲点』,避免『过度信任』。Very useful. The core of AI safety is 'assessing risk' + 'identifying alignment problems'. Critical thinking helps you 'see AI's blind spots', avoiding 'over-trusting'.
📚50 个高频问题,每天读 3 个,3 周读完。读时问自己:『这个答案我同意吗?证据呢?反例呢?』——把『读』变成『思考』。50 high-frequency questions, read 3 daily, finish in 3 weeks. When reading, ask yourself: 'Do I agree with this answer? What's the evidence? Counterexamples?' — turn 'reading' into 'thinking'.
C8

30 条『批判性思维』金句(可背可分享)

30 'Critical Thinking' Golden Quotes (Memorize & Share)

30 条金句,浓缩批判性思维的核心,可背可分享。
30 golden quotes condensing the core of critical thinking, memorizable and shareable.
金句Quote出处/背景Origin/Context怎么用How to use
未经审视的生活不值得过。The unexamined life is not worth living.苏格拉底(Socrates)用于提醒自己『别活在惯性里』,定期审视自己的判断。
我思故我在。I think, therefore I am.笛卡尔(Descartes)判断力的起点是『思考』,不是『反应』。
怀疑不是目的,更接近真相才是目的。Doubt is not the goal; getting closer to truth is.科学方法论精神(Scientific method spirit)怀疑是手段,不是立场。怀疑完要得出『最可能的结论』。
相关不等于因果。Correlation is not causation.统计学基本原则(Basic statistical principle)看到两个现象一起出现,先问:是因果、巧合、还是第三方?
能推翻你的证据,你找了吗?Did you look for evidence that could overturn you?可证伪性(Falsifiability, 波普尔)好判断必须『可被推翻』,只找支持自己的证据 = 自欺。
你相信的,和你知道的,是两回事。What you believe and what you know are two different things.认知科学(Cognitive science)把『信念』当『事实』是最常见的判断错误。
证据的等级:RCT > 观察研究 > 专家意见 > 个例。Evidence hierarchy: RCT > observational study > expert opinion > anecdote.循证医学(Evidence-based medicine)判断任何主张时,先看它的『证据等级』。
最好的论证,是你能替对手说出最强版本。The best argument is one you can state the strongest version of for your opponent.钢铁人原则(Steelmanning)先 steelman 对方,再反驳,你的论证才有质量。
情绪是信号,不是证据。Emotion is a signal, not evidence.认知科学(Cognitive science)愤怒/恐惧让你『想』判断,但不提供『判断的依据』。
简单不是正确,但简单更容易被检验。Simple isn't correct, but simple is easier to test.奥卡姆剃刀(Occam's razor)在多个解释里,选『假设最少』的那个,更容易被验证。
无知不是中立,无知是偏见。Ignorance isn't neutral; ignorance is bias.认知科学(Cognitive science)『不知道』不等于『各占一半』,无知会让你被带节奏。
样本量小于 30,结论只是『个例』。Sample size under 30, the conclusion is just an 'anecdote'.统计学(Statistics)小样本的『规律』大多是大脑在『找规律』。
越自信的人,越该被检验。The more confident, the more should be tested.决策科学(Decision science)自信程度和正确率几乎无关,越自信越要警惕。
把问题说清楚,就解决了一半。Stating the problem clearly is half solved.批判性思维核心(Core of critical thinking)大多数『想不通』是『没说清』。先定义问题,再找答案。
前提错了,结论再漂亮也没用。If the premise is wrong, no matter how pretty the conclusion is useless.逻辑学(Logic)评估论证先评估前提,前提错 = 全错。
你不需要『赢』,你需要『对』。You don't need to 'win'; you need to 'be right'.批判性思维 vs 抬杠(Critical thinking vs contrarian)讨论的目的是『更接近真相』,不是『压过对方』。
承认『我不知道』,是智慧的开始。Admitting 'I don't know' is the beginning of wisdom.苏格拉底(Socrates)最诚实的判断力来自『知道自己不知道』。
信息不是知识,理解才是知识。Information is not knowledge; understanding is knowledge.认知科学(Cognitive science)刷 100 条新闻 ≠ 理解一件事,理解 = 能解释 + 能应用。
让数据说话,先问『数据是谁选的』。Let the data speak, first ask 'who selected the data'.数据批判(Data criticism)数据从来不是『中立的』,它被『选择』过。
警惕『最后一天的优惠』。Beware of 'last day only' deals.营销心理学(Marketing psychology)『紧迫感』是情绪操控,不是事实。剥离时间压力,再看决策。
从众不是判断,是放弃判断。Following the crowd is not judgment; it's giving up judgment.社会心理学(Social psychology)『大家都这样』不是理由,是『放弃思考』的借口。'Everyone does this' is not a reason, it's an excuse to 'give up thinking'.
最危险的不是假信息,是你『相信』的假信息。The most dangerous is not false info, but false info you 'believe'.信息素养(Information literacy)你相信什么,比信息真假更影响你的行为。
把『我觉得』换成『证据是』,你就升级了。Replace 'I feel' with 'the evidence is', and you've upgraded.表达习惯(Expression habit)这是批判性思维最简单的自我训练。
先问『这是事实还是观点』,再谈对错。First ask 'is this fact or opinion', then discuss right/wrong.事实/观点区分(Fact/opinion distinction)事实可验证,观点只能讨论,混淆两者 = 抬杠。
你的大脑不是为『正确』设计的,是为『快速』设计的。Your brain isn't designed for 'correct', it's for 'fast'.进化心理学(Evolutionary psychology)所以『直觉』常错,重要决策必须『慢下来』。
AI 给你的答案,你验证了吗?Did you verify the answer AI gave you?AI 时代批判性思维(AI-era critical thinking)AI 时代最基础的批判性动作:不轻信 + 验证。
最好的提问,是让你『少说结论,多说证据』。The best question makes you 'say less conclusion, more evidence'.提问技巧(Questioning technique)当 AI 给你结论时,追问『证据是什么』。
批判性思维不是『否定一切』,是『有根据地评估』。Critical thinking is not 'denying everything', it's 'evaluating with grounds'.定义(Definition)它的产出是『更可靠的判断』,不是『更多怀疑』。
在 AI 时代,『会问』比『会答』更稀缺。In the AI era, 'knowing how to ask' is scarcer than 'knowing how to answer'.AI 时代能力(AI-era ability)答案无限廉价,『好问题』才决定答案质量。
你的注意力被操纵时,你的判断就被操纵了。When your attention is manipulated, your judgment is manipulated.注意力经济(Attention economy)批判性思维的第一层:保护『注意力』不被带节奏。
💬金句用法:每天读 1 条 + 当天用 1 次(在真实决策里)。30 天读完 30 条,你的判断习惯会肉眼可见地变。Usage: read 1 daily + use it 1x that day (in a real decision). Read all 30 in 30 days, and your judgment habits will visibly change.
C9

30 条『批判性思维』进阶技巧(高手常用)

30 Advanced 'Critical Thinking' Techniques (Used by Experts)

30 条进阶技巧,从『识别谬误』升级到『构建判断系统』。
30 advanced techniques, upgrading from 'spotting fallacies' to 'building a judgment system'.
技巧Technique做法How to do it效果Effect
钢铁人论证Steelmanning先替对方说『最强版本』,再反驳First state the 'strongest version' for the other, then rebut避免『打稻草人』,论证质量翻倍Avoids 'hitting a straw man', argument quality doubles
预-mortemPre-mortem假设项目已经失败,倒推『怎么会失败』Assume the project already failed, work backward 'how it could fail'在成功前找出致命风险Find fatal risks before success
贝叶斯更新Bayesian update先给『先验概率』,看到新证据后更新First give 'prior probability', update after seeing new evidence把『判断』变成『可计算的』Turns 'judgment' into 'calculable'
5 个为什么5 Whys对问题连续问 5 次『为什么』Ask 'why' 5 times consecutively for a problem从『表面现象』挖到『根本原因』Dig from 'surface phenomenon' to 'root cause'
红队思维Red-teaming专门扮演『攻击方』找漏洞Play the 'attacker' role to find gaps在自己的方案被攻击前,自己先攻击Attack your own plan before it gets attacked
魔鬼代言人Devil's advocate主动请人『反对你的观点』Actively ask someone to 'oppose your view'打破『群体思维』Breaks 'groupthink'
证据等级法Evidence level method给每条证据标等级(RCT/观察/个例)Label each piece of evidence with a level (RCT/observation/anecdote)避免『被个例带节奏』Avoids 'being led by anecdotes'
反事实思考Counterfactual thinking问『如果不是这样,会怎样』Ask 'what if it weren't so, what would happen'暴露『隐藏假设』Exposes 'hidden assumptions'
第一性原理First principles回到『最基本的事实』重新推导Return to 'basic facts' and re-derive避免『类比思维』的陷阱Avoids the trap of 'analogical thinking'
奥卡姆剃刀Occam's razor多个解释选『假设最少』的Among multiple explanations, pick the 'fewest assumptions'简单更易验证,减少过度解读Simple is easier to verify, reduces over-interpretation
概率思维Probabilistic thinking用『概率』而非『对错』描述判断Describe judgments with 'probability' not 'right/wrong'避免『非黑即白』Avoids 'black or white'
基础比率Base rate先看『一般概率』,再看『个体情况』First look at 'general probability', then 'individual situation'避免『个例偏差』Avoids 'anecdote bias'
锚定效应免疫Anchoring immunity对『第一个数字』保持警惕Stay wary of 'the first number'避免『被价格/数字带节奏』Avoids 'being led by prices/numbers'
损失框架检查Loss-frame check问『这个决策让我损失什么』Ask 'what does this decision make me lose'避免『损失厌恶』的盲区Avoids the blind spot of 'loss aversion'
决策日志Decision journal记录每次决策的『理由 + 当时情绪』Record each decision's 'reasons + emotions at the time'事后复盘,找到自己的『判断模式』Review after the fact, find your own 'judgment patterns'
双盲检查Double-blind check让别人独立评估你的论证Let someone independently evaluate your argument避免『自我说服』Avoids 'self-persuasion'
时间折扣检查Time-discount check问『10 年后看这个决策,还成立吗』Ask 'looking back at this decision in 10 years, does it still hold'避免『短期冲动』Avoids 'short-term impulsiveness'
身份偏见检查Identity-bias check问『我是因为立场而相信,还是因为证据』Ask 'do I believe because of my stance, or because of evidence'避免『立场先于判断』Avoids 'stance before judgment'
情绪标注Emotion labeling在判断前标注『我现在的情绪』Before judging, label 'my current emotion'情绪被看见后,影响力下降 50%Once emotion is seen, its influence drops 50%
信息断食Information diet每天固定 1 小时『不碰任何信息』1 hour daily 'not touching any info'恢复『独立思考』能力Restores 'independent thinking' ability
单一信源惩罚Single-source penalty只看到一个信源时,判断自动降 30%When only seeing one source, judgment auto-drops 30%强制『多信源验证』Forces 'multi-source verification'
置信度校准Confidence calibration每次判断给『置信度%』,事后核对Each judgment gives 'confidence %', check after the fact训练『自信程度 = 正确率』Trains 'confidence level = accuracy'
可证伪清单Falsifiability checklist判断前问『什么证据能推翻它』Before judging, ask 'what evidence could overturn it'避免『不可推翻』的伪判断Avoids 'unfalsifiable' pseudo-judgments
样本检查Sample check问『样本量多大?有代表性吗?』Ask 'how big is the sample? Is it representative?'避免『以偏概全』Avoids 'hasty generalization'
时间戳检查Timestamp check问『这是哪年的信息?还有效吗?』Ask 'what year is this info from? Still valid?'避免『用旧信息判断新情况』Avoids 'judging new situations with old info'
因果方向检查Causal-direction check问『是 A 导致 B,还是 B 导致 A』Ask 'is A causing B, or B causing A'避免『因果倒置』Avoids 'reversed causation'
第三方变量Third-variable问『有没有 C 同时导致 A 和 B』Ask 'is there a C causing both A and B'避免『虚假相关』Avoids 'spurious correlation'
幸存者检查Survivorship check问『失败/没活下来的在哪』Ask 'where are the failures/ones that didn't survive'避免『幸存者偏差』Avoids 'survivorship bias'
动机检查Motivation check问『说这话的人,有什么动机』Ask 'what motive does the person making this statement have'避免『被利益驱动的信息』带节奏Avoids 'being led by interest-driven info'
框架检查Frame check问『这个问题是怎么被『框』的』Ask 'how is this problem being 'framed''避免『框架操纵』Avoids 'frame manipulation'
🧰30 条技巧不用全记。挑 5 条最高频的(钢铁人、预-mortem、置信度校准、证据等级、决策日志),每天用 1 条,1 个月后变成你的『判断习惯』。No need to memorize all 30. Pick the 5 most frequent (steelman, pre-mortem, confidence calibration, evidence level, decision journal), use 1 daily, and in 1 month it becomes your 'judgment habit'.
C10

10 个『批判性思维』训练工具(免费)

10 Free 'Critical Thinking' Training Tools

10 个免费工具,帮你把批判性思维『练出来』。
10 free tools to help you 'train out' critical thinking.
工具Tool类型Type怎么用How to use费用Cost
Pond5 逻辑谬误测试Logic Fallacy Test (online)测试Test每年测一次,看自己最容易中哪种谬误Test once a year, see which fallacy you most easily fall into免费Free
Anki(间隔重复)Anki (spaced repetition)记忆Memory把 12 种谬误 + 30 条技巧做成卡片,每天复习 10 分钟Make cards for 12 fallacies + 30 techniques, review 10 min daily免费Free
新闻溯源(WHO 查源)News source-checking (WHO)验证Verification看到新闻先查『WHO 信源核查』See news, first check 'WHO source verification'免费Free
FactCheck.orgFactCheck.org核查Fact-check验证『政治/商业』主张的真假Verify the truth of 'political/business' claims免费Free
Our World in DataOur World in Data数据Data看『全球数据』的可视化,训练『数据直觉』See 'global data' visualization, train 'data intuition'免费Free
Coursera 批判性思维课Coursera Critical Thinking Course课程Course系统学批判性思维(有免费旁听)Systematically study critical thinking (free audit available)免费旁听Free audit
Khan Academy 逻辑Khan Academy Logic课程Course学『形式逻辑』基础(免费)Learn 'formal logic' basics (free)免费Free
AI 对话(验证模式)AI Conversation (verification mode)练习Practice每天用 AI 做『验证练习』:给它一个主张,让它找反例Daily 'verification practice' with AI: give it a claim, let it find counterexamples免费/订阅Free/sub
决策日志(Notion/Excel)Decision Journal (Notion/Excel)复盘Review每次重要决策记录理由 + 情绪,每月复盘Record reasons + emotions for each important decision, review monthly免费Free
辩论俱乐部/读书会Debate Club/Book Club社群Community每周 1 次『辩论 + 复盘』,训练『对抗性思维』Weekly 'debate + review', train 'adversarial thinking'免费Free
🛠️工具组合:Anki(记知识)+ AI 验证(练手感)+ 决策日志(复盘)。三个免费工具,每天 20 分钟,3 个月你的判断力会有质的变化。Tool combo: Anki (memorize) + AI verification (practice) + decision journal (review). Three free tools, 20 min daily, in 3 months your judgment will have a qualitative change.
C11

60 个『判断练习』:这是事实、观点还是带节奏?

60 'Judgment Drills': Fact, Opinion, or Being Led?

60 个真实句子,判断它是『事实』『观点』还是『带节奏』,附解析。
60 real sentences; judge if fact, opinion, or manipulation, with analysis.
句子Statement判断Verdict为什么Why
『2026 年全球 AI 投资增长了 40%。』'In 2026 global AI investment grew 40%.'观点(无来源)Opinion (no source)有具体数字但没给来源/年份/机构,无法验证Has specific number but no source/year/institution, unverifiable
『我认识 5 个用 AI 的都没变强。』'I know 5 who use AI and none got stronger.'以偏概全Hasty generalization5 个样本是『个例』,不能代表全体5 samples are 'anecdotes', can't represent the whole
『AI 一定会取代所有工作。』'AI will definitely replace all jobs.'绝对化 + 滑坡Absolutism + slippery slope『一定』是绝对词,且中间环节无证据'Definitely' is absolute, and middle steps lack evidence
『研究表明 90% 的人觉得 AI 有用。』'Research shows 90% think AI is useful.'红旗(无出处)Red flag (no citation)『研究表明』没给具体研究,高风险幻觉'Research shows' without specific study, high hallucination risk
『水在 100 摄氏度沸腾。』'Water boils at 100 degrees C.'事实(有条件)Fact (conditional)在标准大气压下成立,是可验证的客观事实Holds at standard pressure, verifiable objective fact
『这家 AI 公司是最强的。』'This AI company is the strongest.'观点Opinion『最强』是主观比较,没有统一标准'Strongest' is subjective comparison, no unified standard
『你不用 AI 就会被淘汰。』'If you don't use AI you'll be eliminated.'情感操控(恐惧)Emotional manipulation (fear)制造恐惧逼你行动,不是提供证据Manufactures fear to force action, not provide evidence
『这个模型参数量 1 万亿。』'This model has 1 trillion parameters.'事实(可验证)Fact (verifiable)具体数字,可查官方发布核实Specific number, verifiable against official release
『用 AI 的人更聪明。』'People using AI are smarter.'无证据的因果Causation without evidence『用 AI』和『聪明』可能只是相关,未必因果'Using AI' and 'smart' may only be correlated, not causal
『AI 能写代码。』'AI can write code.'事实Fact可重复验证的客观能力,无需争议Repeatably verifiable objective capability, no dispute
『你买这个就不会后悔。』'You won't regret buying this.'承诺(无法验证)Promise (unverifiable)『不会后悔』是未来承诺,无法事前验证'Won't regret' is a future promise, can't be pre-verified
『95% 的用户满意度。』'95% user satisfaction.'数据(需验证来源)Data (source needed)具体数字,但要问:谁测的?样本多大?Specific number, but ask: who measured? sample size?
『AI 时代必须学批判性思维。』'You must learn critical thinking in the AI era.'观点(合理)Opinion (reasonable)合理建议,但『必须』带强推,要自己判断Reasonable suggestion, but 'must' pushes hard, judge yourself
『这项研究在顶刊发表。』'This study was published in a top journal.'事实(可验证)Fact (verifiable)『顶刊发表』是客观事件,可查证'Published in top journal' is objective event, checkable
『大家都说这款好。』'Everyone says this one's good.'诉诸大众Appeal to crowd『大家』是谁?没具体,是『从众』信号Who's 'everyone'? Not specific, a 'herd' signal
『AI 让效率提升 3 倍。』'AI triples efficiency.'数据(需条件)Data (needs conditions)在什么场景/什么任务下 3 倍?无上下文不可用In what scenario/task is it 3x? Unusable without context
『我不信 AI。』'I don't trust AI.'观点(个人)Opinion (personal)个人态度,无对错,但要问『为什么不信』Personal attitude, no right/wrong, but ask 'why not trust'
『这个数据来自 2024 年。』'This data is from 2024.'事实(时效信息)Fact (timeliness info)明确的年份,可判断是否过时Clear year, can judge if outdated
『AI 会消灭人类。』'AI will destroy humanity.'极端滑坡 + 无证据Extreme slippery slope + no evidence『消灭』是极端结论,中间环节全无证据'Destroy' is extreme conclusion, all middle steps lack evidence
『这款 AI 比那款准 15%。』'This AI is 15% more accurate than that one.'数据(需验证基准)Data (needs benchmark)15% 比什么?在哪个测试集?基准不明15% compared to what? Which test set? Unclear benchmark
『专家说 AI 很安全。』'Experts say AI is very safe.'诉诸权威(需核查)Appeal to authority (need check)哪个专家?安全是相对概念,要具体Which expert? Safety is relative, needs specifics
『我用过,很好用。』'I've used it, very easy.'个例(主观)Anecdote (subjective)个人体验,是『个例』不是『数据』Personal experience, 'anecdote' not 'data'
『AI 训练需要 10000 块 GPU。』'AI training needs 10000 GPUs.'事实(可验证)Fact (verifiable)具体数字,可查公开资料核实Specific number, verifiable via public sources
『买它,最后 3 件。』'Buy it, only 3 left.'紧迫感操控Urgency manipulation『最后 3 件』是饥饿营销,先核实库存'Only 3 left' is hunger marketing, verify stock first
『这个算法的准确率 99%。』'This algorithm's accuracy is 99%.'数据(需问测试集)Data (ask test set)99% 在什么数据上测的?测试集很关键99% on what data? Test set is key
『AI 让人类进步。』'AI makes humanity progress.'观点(需定义)Opinion (needs definition)『进步』如何定义?没有统一衡量How to define 'progress'? No unified measure
『这家公司估值 1000 亿。』'This company is valued at 100 billion.'事实(可验证)Fact (verifiable)估值是公开信息,可查证Valuation is public info, checkable
『AI 比人聪明。』'AI is smarter than humans.'观点(需定义聪明)Opinion (needs defining smart)『聪明』是多重维度,AI 在某些任务强不是全面'Smart' is multi-dimensional, AI strong in some tasks isn't across the board
『研究表明 AI 会取代 40% 的工作。』'Research shows AI will replace 40% of jobs.'数据(需核查研究)Data (need check study)哪个研究?40% 的方法论是什么?Which study? What's the methodology for 40%?
『我邻居用 AI 赚了 100 万。』'My neighbor made 1M with AI.'幸存者偏差Survivorship bias只看到『赚了』的,看不到『没赚』的多数Only seeing the 'earned' ones, not the majority who didn't
『AI 的准确率在提升。』'AI's accuracy is improving.'趋势(需数据支撑)Trend (needs data support)是事实趋势,但要具体指标和时间线Factual trend, but needs specific metrics and timeline
『你不学 AI 就落伍了。』'If you don't learn AI you'll fall behind.'情感操控(焦虑)Emotional manipulation (anxiety)制造焦虑,不是提供『如何不落后』的证据Manufactures anxiety, not evidence of 'how to not fall behind'
『这个数据集有 10 亿条。』'This dataset has 1 billion entries.'事实(可验证)Fact (verifiable)具体规模,可查公开资料Specific scale, verifiable via public sources
『AI 让学习更高效。』'AI makes learning more efficient.'观点(需证据)Opinion (needs evidence)『更高效』要量化对比,无数据只是观点'More efficient' needs quantified comparison, no data = opinion
『这款模型开源了。』'This model is open-sourced.'事实(可验证)Fact (verifiable)开源是客观状态,可查仓库Open-source is objective state, checkable in repo
『99% 的人不知道 AI 的真相。』'99% don't know the truth about AI.'绝对化 + 神秘化Absolutism + mystification『真相』是什么?制造信息差,带节奏What's 'the truth'? Manufactures info gap, manipulates
『AI 的能耗是电力的 5 倍。』'AI's energy use is 5x that of electricity.'数据(需验证)Data (needs verification)具体倍数,但要问基准和年份Specific multiple, but ask baseline and year
『用 AI 学习 3 个月就精通。』'Master in 3 months with AI learning.'虚假承诺False promise『3 个月精通』是营销话术,无证据'Master in 3 months' is marketing, no evidence
『这个 API 每秒处理 1000 请求。』'This API handles 1000 requests/sec.'事实(可验证)Fact (verifiable)具体性能指标,可压测验证Specific performance metric, load-testable
『AI 让普通人也能创造。』'AI lets ordinary people create.'观点(部分成立)Opinion (partly true)部分成立,但『创造』定义宽,要具体Partly true, but 'create' is broad, needs specifics
『这款产品的用户增长了 200%。』'This product's users grew 200%.'数据(需验证)Data (needs verification)具体增长,但要问基数和时间Specific growth, but ask base and period
『AI 会改变世界。』'AI will change the world.'观点(太宽泛)Opinion (too broad)『改变世界』太宽泛,无法证伪'Change the world' too broad, unfalsifiable
『这个模型的幻觉率 5%。』'This model's hallucination rate is 5%.'数据(需验证)Data (needs verification)具体幻觉率,但要问测试方法Specific hallucination rate, but ask test method
『不用 AI 的人会被时代抛弃。』'People not using AI will be left behind.'滑坡 + 情感Slippery slope + emotion『抛弃』是极端,制造恐惧'Left behind' is extreme, manufactures fear
『AI 的推理能力接近人类。』'AI's reasoning is close to human.'观点(需定义接近)Opinion (needs defining close)『接近』如何衡量?多维度,无统一How to measure 'close'? Multi-dimensional, no unified
『这个数据集是合成的。』'This dataset is synthetic.'事实(可验证)Fact (verifiable)合成是客观属性,可查证Synthetic is objective property, checkable
『AI 让 90% 的工作消失。』'AI makes 90% of jobs disappear.'极端 + 无证据Extreme + no evidence90% 是极端数字,无可靠来源90% is extreme number, no reliable source
『这款 AI 比上代快 2 倍。』'This AI is 2x faster than last gen.'数据(需验证基准)Data (needs benchmark)2 倍比什么任务?基准不明2x on what task? Unclear baseline
『AI 让人更懒。』'AI makes people lazier.'观点(需证据)Opinion (needs evidence)『更懒』要量化,无数据只是感受'Lazier' needs quantifying, no data = just feeling
『这个模型的训练成本 100 万美元。』'This model's training cost is $1M.'数据(需验证)Data (needs verification)具体成本,但要问包含什么Specific cost, but ask what's included
『AI 时代人人要学提示词。』'In the AI era everyone must learn prompting.'观点(合理)Opinion (reasonable)合理建议,但『人人』太绝对Reasonable, but 'everyone' is too absolute
『这款 AI 的月活 1 亿。』'This AI's monthly active is 100M.'数据(需验证)Data (needs verification)具体 MAU,但要问统计口径Specific MAU, but ask statistical basis
『AI 会取代教师。』'AI will replace teachers.'滑坡 + 极端Slippery slope + extreme『取代』是极端,教育是复杂系统'Replace' is extreme, education is a complex system
『这个数据集有 100 个国家。』'This dataset covers 100 countries.'事实(可验证)Fact (verifiable)具体覆盖范围,可查证Specific coverage, checkable
『AI 让信息过载。』'AI causes information overload.'观点(需证据)Opinion (needs evidence)『过载』要量化,无数据只是感受'Overload' needs quantifying, no data = just feeling
『这款 AI 的延迟 200ms。』'This AI's latency is 200ms.'事实(可验证)Fact (verifiable)具体性能,可测验证Specific performance, testable
『AI 让人类思考退化。』'AI makes human thinking degrade.'观点(需证据)Opinion (needs evidence)『退化』要量化对比,无数据'Degrade' needs quantified comparison, no data
『这个模型的参数量 700 亿。』'This model has 70B parameters.'事实(可验证)Fact (verifiable)具体参数,可查官方Specific params, checkable officially
🎯练习方法:每天做 5 题,先自己判断,再看解析。12 天做完 60 题,你的『事实/观点/带节奏』识别力会显著提升。Practice: 5 questions daily, judge yourself first, then check analysis. Finish all 60 in 12 days, and your 'fact/opinion/manipulation' recognition will clearly improve.
C12

40 个『信息验证』实战案例(AI 输出)

40 'Information Verification' Practical Cases (AI Output)

40 个 AI 输出的典型问题案例,演示如何一步步验证。
40 typical AI output problem cases, demonstrating step-by-step verification.
AI 输出AI Output问题Problem验证步骤Verification Steps结论Verdict
『研究表明 85% 的人用过 AI。』'Research shows 85% have used AI.'无具体研究No specific study① 问哪篇研究 ② 搜该研究 ③ 查样本量① Which study ② Search it ③ Check sample size无法验证,存疑Unverifiable, doubtful
『2025 年 AI 市场 5000 亿美元。』'2025 AI market is $500B.'数字无来源Number no source① 问数据来源 ② 查行业报告 ③ 对比多家① Ask source ② Check industry reports ③ Compare multiple数字可能有,但口径不一Number may exist, but basis varies
『这个模型比 GPT 强 30%。』'This model is 30% stronger than GPT.'基准不明Benchmark unclear① 问哪个测试 ② 查测试集 ③ 看是否独立① Which test ② Check test set ③ See if independent30% 无上下文不可用30% unusable without context
『AI 会取代 40% 的工作(McKinsey)。』'AI will replace 40% of jobs (McKinsey).引用需核实Citation needs check① 查 McKinsey 原报告 ② 看 40% 指什么 ③ 是否夸大① Check McKinsey original ② What 40% means ③ Exaggerated?McKinsey 有类似研究,但 40% 是『可自动化』非『取代』McKinsey has similar research, but 40% is 'automatable' not 'replace'
『水在 100 度沸腾,这是错的,其实 95 度。』'Water boils at 100, that's wrong, actually 95.'事实错误Factual error① 查科学常识 ② 注意气压条件 ③ AI 在编① Check science common sense ② Note pressure ③ AI fabricating100 度(标准气压)正确,AI 错了100 deg (standard pressure) is correct, AI is wrong
『AI 训练用了 1 亿小时 GPU。』'AI training used 100M GPU-hours.'数字可疑Number suspicious① 问来源 ② 查已知模型训练量 ③ 数量级对比① Ask source ② Check known model training ③ Order of magnitude数量级可能,但无来源Order may be right, but no source
『这款 AI 通过了图灵测试。』'This AI passed the Turing test.'定义问题Definition problem① 图灵测试如何判定 ② 是否有严格测试 ③ 多为主观① How is Turing test judged ② Strict test? ③ Mostly subjective『通过』定义不清,多数是宣传'Passed' is unclear, mostly marketing
『AI 让效率提升 10 倍(内部数据)。』'AI boosts efficiency 10x (internal data).内部数据不可验Internal data unverifiable① 问样本 ② 问基线 ③ 问时间 ④ 是否可复现① Ask sample ② Ask baseline ③ Ask period ④ Reproducible?『内部数据』是宣传话术'Internal data' is marketing talk
『AI 的准确率 99.9%。』'AI's accuracy is 99.9%.'测试集问题Test set problem① 问测试集 ② 问是否独立 ③ 看基线① Which test set ② Independent? ③ See baseline99.9% 可能在小数据集上99.9% may be on small dataset
『AI 会消灭 90% 的程序员。』'AI will eliminate 90% of programmers.'极端 + 无源Extreme + no source① 问来源 ② 查真实数据 ③ 看『消灭』定义① Ask source ② Check real data ③ 'Eliminate' definition90% 是极端,无可靠来源90% is extreme, no reliable source
『AI 的能耗是核电站的 3 倍。』'AI's energy use is 3x a nuclear plant.'对比不当Improper comparison① 问对比口径 ② 查实际能耗 ③ 时间维度① Compare basis ② Check actual energy ③ Time dimension对比口径不清,可能夸大Comparison basis unclear, may exaggerate
『AI 让 80% 的企业转型成功。』'AI makes 80% of enterprises transform successfully.'成功率无源Success rate no source① 问研究 ② 问定义『成功』 ③ 查数据① Ask study ② Define 'success' ③ Check data80% 无来源,存疑80% no source, doubtful
『AI 的幻觉率只有 1%。』'AI's hallucination rate is only 1%.'测试方法问题Test method problem① 问测试方法 ② 问测试集 ③ 看是否独立① Test method ② Test set ③ Independent?1% 取决于测试,可能偏低1% depends on test, may be low
『AI 比人类医生诊断更准。』'AI diagnoses more accurately than human doctors.'需具体场景Needs specific scenario① 问哪个病 ② 问对比对象 ③ 查研究① Which disease ② Compare with ③ Check study某些场景 AI 准,但不是全面AI accurate in some scenarios, not across the board
『AI 让学习速度提升 5 倍。』'AI speeds up learning 5x.'5 倍无上下文5x no context① 问什么学习 ② 问基线 ③ 查数据① What learning ② Baseline ③ Data5 倍无上下文不可用5x unusable without context
『AI 的月活突破 10 亿。』'AI's monthly active passed 1B.'需验证口径Need verify basis① 问统计口径 ② 查官方 ③ 是否含子产品① Statistical basis ② Official ③ Includes sub-products?10 亿可能是累计非月活1B may be cumulative not MAU
『AI 会取代所有写作工作。』'AI will replace all writing jobs.'极端 + 无源Extreme + no source① 问来源 ② 看『所有』 ③ 查真实数据① Source ② 'All' ③ Real data『所有』是极端,无源'All' is extreme, no source
『AI 的推理能力超过人类。』'AI's reasoning exceeds humans.'需定义Needs definition① 问哪类推理 ② 问测试 ③ 查研究① What reasoning ② Test ③ Study某些推理 AI 强,但非全面AI strong in some reasoning, not across the board
『AI 让 70% 的公司倒闭。』'AI causes 70% of companies to fail.'极端 + 无源Extreme + no source① 问来源 ② 查数据 ③ 看时间① Source ② Data ③ Time70% 极端,无源70% extreme, no source
『AI 的准确率在小样本上 100%。』'AI's accuracy is 100% on small samples.'过拟合信号Overfitting signal① 问样本量 ② 问独立集 ③ 看是否过拟合① Sample size ② Independent set ③ Overfit?小样本 100% 是过拟合信号100% on small samples = overfitting signal
『AI 的能耗每年翻倍。』'AI's energy doubles yearly.'趋势需数据Trend needs data① 问数据 ② 查行业 ③ 看时间① Data ② Industry ③ Time趋势可能,但需具体数据Trend possible, but needs specific data
『AI 让 60% 的员工失业。』'AI makes 60% of employees unemployed.'极端 + 无源Extreme + no source① 问来源 ② 查数据 ③ 看定义① Source ② Data ③ Definition60% 极端,无源60% extreme, no source
『AI 的训练数据 1 万亿 tokens。』'AI's training data is 1T tokens.'需验证Needs verification① 问来源 ② 查官方 ③ 看是否含重复① Source ② Official ③ Includes duplicates?1T 可能,但需确认1T possible, but needs confirmation
『AI 比人类更快更准。』'AI is faster and more accurate than humans.'需具体任务Needs specific task① 问什么任务 ② 问对比 ③ 查数据① What task ② Compare ③ Data某些任务 AI 更快更准,但非全面AI faster/more accurate in some tasks, not across the board
『AI 的幻觉率比人类低。』'AI's hallucination rate is lower than humans.'需对比口径Needs comparison basis① 问如何测人类 ② 问测试 ③ 查研究① How measure humans ② Test ③ Study『人类幻觉率』难量化,对比存疑'Human hallucination rate' hard to quantify, comparison doubtful
『AI 让 50% 的内容生产成本降了。』'AI cuts content production cost by 50%.'需具体领域Needs specific field① 问什么内容 ② 问基线 ③ 查数据① What content ② Baseline ③ Data50% 需具体领域和基线50% needs specific field and baseline
『AI 的参数量是 GPT 的 10 倍。』'AI's parameters are 10x GPT.'需具体模型Needs specific model① 问哪个模型 ② 查参数 ③ 对比① Which model ② Params ③ Compare需具体模型名才能比Need specific model name to compare
『AI 让 90% 的广告变成 AI 生成。』'AI makes 90% of ads AI-generated.'需验证Needs verification① 问来源 ② 查行业 ③ 看时间① Source ② Industry ③ Time90% 需具体行业和时间90% needs specific industry and time
『AI 的准确率在医疗上 99%。』'AI's accuracy in medicine is 99%.'需具体病种Needs specific disease① 问什么病 ② 问测试 ③ 查研究① What disease ② Test ③ Study99% 需具体病种和测试99% needs specific disease and test
『AI 让编程效率提升 3 倍。』'AI triples programming efficiency.'需具体任务Needs specific task① 问什么任务 ② 问基线 ③ 查数据① What task ② Baseline ③ Data3 倍需具体任务和基线3x needs specific task and baseline
『AI 的能耗超过一个城市。』'AI's energy exceeds a city.'需具体城市Needs specific city① 问哪个城市 ② 查能耗 ③ 时间① Which city ② Energy ③ Time需具体城市和时间Needs specific city and time
『AI 让 80% 的客服被取代。』'AI replaces 80% of customer service.'需验证Needs verification① 问来源 ② 查数据 ③ 看时间① Source ② Data ③ Time80% 需具体行业和时间80% needs specific industry and time
『AI 的推理成本降了 10 倍。』'AI's inference cost dropped 10x.'需时间对比Needs time comparison① 问哪两年 ② 问口径 ③ 查数据① Which years ② Basis ③ Data10 倍需时间对比10x needs time comparison
『AI 让 70% 的设计被 AI 做。』'AI makes 70% of design AI-done.'需验证Needs verification① 问来源 ② 查行业 ③ 时间① Source ② Industry ③ Time70% 需具体领域70% needs specific field
『AI 的准确率在数学上 95%。』'AI's accuracy in math is 95%.'需具体题型Needs specific problem type① 问什么题 ② 问测试 ③ 查研究① What problems ② Test ③ Study95% 需具体题型95% needs specific problem type
『AI 让 60% 的翻译被 AI 做。』'AI makes 60% of translation AI-done.'需验证Needs verification① 问来源 ② 查数据 ③ 时间① Source ② Data ③ Time60% 需具体领域和时间60% needs specific field and time
『AI 的幻觉率在 2026 年降了 50%。』'AI's hallucination rate dropped 50% in 2026.'需时间对比Needs time comparison① 问哪两年 ② 问方法 ③ 查数据① Which years ② Method ③ Data50% 需时间对比和方法50% needs time comparison and method
『AI 让 50% 的教育内容 AI 生成。』'AI makes 50% of education content AI-generated.'需验证Needs verification① 问来源 ② 查行业 ③ 时间① Source ② Industry ③ Time50% 需具体领域和时间50% needs specific field and time
『AI 的准确率在法律上 90%。』'AI's accuracy in law is 90%.'需具体场景Needs specific scenario① 问什么场景 ② 问测试 ③ 查研究① What scenario ② Test ③ Study90% 需具体场景90% needs specific scenario
『AI 让 80% 的代码由 AI 辅助。』'AI makes 80% of code AI-assisted.'需验证Needs verification① 问来源 ② 查数据 ③ 时间① Source ② Data ③ Time80% 需具体行业和时间80% needs specific industry and time
『AI 的幻觉率在医疗上 0。』'AI's hallucination rate in medicine is 0.'不可能Impossible① 0 幻觉不存在 ② 质疑 ③ 查数据① 0 hallucination doesn't exist ② Doubt ③ Data0 是虚假承诺,不可能0 is a false promise, impossible
⚠️40 个案例的共同规律:AI 输出的『具体数字』大多需要验证。记住口诀:数字要问来源、年份、口径、测试集、样本量。五个都答不上,就当它没说过。Common pattern in 40 cases: AI's 'specific numbers' mostly need verification. Mnemonic: for numbers, ask source, year, basis, test set, sample size. If it can't answer all five, treat it as if it never said it.
C13

30 个认知偏差(批判性思维的敌人)

30 Cognitive Biases (Enemies of Critical Thinking)

30 个最常见的认知偏差,识别它们 = 识别自己的思维盲区。
30 most common cognitive biases; recognizing them = recognizing your own thinking blind spots.
偏差Bias表现How it shows解药The cure
确认偏误Confirmation bias只找支持自己观点的证据Seeking evidence supporting your own view主动找反证(红队思维)Actively seek counter-evidence (red team thinking)
锚定效应Anchoring effect被第一个数字带节奏Being dragged by the first number先独立判断再看参考值Judge independently before seeing reference values
从众偏差Bandwagon bias别人都这么认为所以正确Correct because others think so独立验证,不因众而信Verify independently; not trust by crowd
幸存者偏差Survivorship bias只看到成功者,忽略失败者Only seeing winners, ignoring losers同时统计失败案例Count failure cases simultaneously
光环效应Halo effect因某人某方面优秀,认为他全对Seeing one excellence, assuming all correctness分开评估不同维度Evaluate different dimensions separately
沉没成本Sunk cost已投入的不愿意止损Unwilling to cut losses due to past investment只看未来收益,忽略已投入Only look at future returns, ignore past investment
过度自信Overconfidence高估自己判断的准确性Overestimating accuracy of your judgment校准:回顾历史判断正确率Calibrate: review historical judgment accuracy
可得性偏差Availability bias容易想到的事觉得更常见Easy-to-remember things seem more common用数据而非印象判断Judge by data not impression
虚假相关Spurious correlation相关不等于因果Correlation is not causation找因果机制,警惕巧合Find causal mechanism, watch for coincidence
诉诸权威Appeal to authority专家说的就是对的Experts are always right看证据,不只看身份Look at evidence, not just identity
诉诸情感Appeal to emotion用情绪代替逻辑Using emotion instead of logic情绪平静后再判断Judge after emotion calms
虚假二分False dilemma只有 A 或 B,忽略中间Only A or B, ignoring middle找第三、第四种可能Find third, fourth possibilities
滑坡谬误Slippery slopeA 会导致 B 会导致灾难A leads to B leads to disaster检验每一步因果是否成立Verify each causal step
稻草人Straw man歪曲对方观点再反驳Distorting the other's view then refuting先复述对方观点再回应Restate the other's view before responding
诉诸无知Appeal to ignorance没人证明假就是真Unproven false means true区分『未证明』与『已证明假』Distinguish 'unproven' from 'proven false'
后此谬误Post hocA 在 B 前所以 A 导致 BA before B so A caused B找机制和对照Find mechanism and control
以偏概全Hasty generalization样本太小就下结论Small sample, big conclusion扩大样本,标注置信度Enlarge sample, mark confidence
诉诸自然Appeal to nature天然的就是好的Natural is good天然不等于安全/正确Natural not equal safe/correct
诉诸流行Appeal to popularity流行就是对的Popular is correct看证据,不看数量Look at evidence, not numbers
动机偏差Motivated reasoning按动机推理,不按证据Reasoning by motive, not evidence问自己:如果证据相反我会信吗Ask: would I believe if evidence were opposite
🧭30 个偏差不用全记。最高频的 6 个:确认偏误、从众、幸存者、过度自信、虚假相关、沉没成本。识别这 6 个,就能避开 80% 的思维坑。No need to memorize all 30. The 6 highest-frequency: confirmation bias, bandwagon, survivorship, overconfidence, spurious correlation, sunk cost. Recognize these 6 to avoid 80% of thinking traps.
C14

30 个判断案例:这条信息可信吗?

30 Judgment Cases: Is This Information Credible?

30 个真实信息判断案例,给出『可信/存疑/不可信』+ 判断理由。
30 real information-judgment cases with 'credible/questionable/uncredible' + reasoning.
案例Case判断Verdict判断理由Reasoning
『某专家说:喝 X 水能治百病』'Expert says: drinking X cures everything'不可信Uncredible『治百病』= 过度承诺;无任何病能靠水治'Cures everything' = over-claim; no disease is cured by water
『某品牌: 99% 用户说好用(未说明样本)』'Brand: 99% users satisfied (sample not stated)'存疑Questionable未说明样本量/筛选方式,典型营销话术Sample size/selection not stated; typical marketing talk
『权威媒体报道: 某国 GDP 增长 5%(引用统计局)』'Authoritative media: GDP up 5% (citing stats bureau)'可信Credible有明确数据源(统计局),可独立核实Clear data source (stats bureau), independently verifiable
『朋友圈: 转发这条能给孩子积福』'WeChat: forward this to accumulate merit for your child'不可信Uncredible无信息价值,纯情绪操纵No information value; pure emotional manipulation
『某论文: 咖啡与长寿相关(n=10 万,对照实验)』'Paper: coffee correlates with longevity (n=100k, controlled)'可信(保留)Credible (reserved)样本大 + 对照,但『相关』非『因果』Large sample + control, but 'correlation' not 'causation'
『某博主: AI 明天会毁灭人类』'Blogger: AI will destroy humanity tomorrow'存疑Questionable无证据、时间模糊,恐惧驱动No evidence, vague timeline, fear-driven
『官方数据: 2025 年失业率 5.1%(附调查方法)』'Official: 2025 unemployment 5.1% (with survey method)'可信Credible附调查方法,透明可复核Survey method attached; transparent and checkable
『某群: 内部消息,某股票明天涨停』'Group: inside info, stock will limit-up tomorrow'不可信Uncredible无证据的『内部消息』= 典型骗局'Inside info' without evidence = classic scam
『研究: 睡眠不足与肥胖相关(meta 分析 30 研究)』'Study: sleep deprivation correlates with obesity (meta of 30 studies)'可信(保留)Credible (reserved)多研究荟萃,证据强度较高Multi-study meta; higher evidence strength
『某网红: 我用某产品一个月瘦 10 斤』'Influencer: I lost 10 jin in a month with this product'存疑Questionable个体案例,无对照,可能是广告Individual case, no control, possibly an ad
『官方发布: 某地地震 5.0 级(附震中坐标)』'Official: 5.0 earthquake (with epicenter coords)'可信Credible官方数据 + 具体参数,可独立验证Official data + specific params; independently verifiable
『某公众号: 震惊!90% 的人不知道』'Account: Shocking! 90% don't know this'存疑Questionable标题党,『震惊』= 情绪钩子Clickbait; 'shocking' = emotional hook
『论文: X 药物有效(随机双盲,n=500,同行评审)』'Paper: Drug X effective (RCT, n=500, peer-reviewed)'可信CredibleRCT + 同行评审,证据等级最高RCT + peer review; highest evidence level
『某群: 转发此消息积功德』'Group: forward this to accumulate merit'不可信Uncredible无信息价值,纯操纵No information value; pure manipulation
『媒体报道: 某公司裁员 20%(引用内部文件)』'Media: company cuts 20% (citing internal docs)'存疑Questionable有信源但单一,需多方交叉验证Has source but single; needs cross-verification
『官方: 流感季请勤洗手(引用 WHO 指南)』'Official: wash hands in flu season (citing WHO guidelines)'可信Credible引用权威指南,可查证Cites authoritative guideline; verifiable
『某博主: AI 让我 3 个月赚 100 万』'Blogger: AI made me 1M in 3 months'存疑Questionable无收入凭证,典型引流No income proof; typical funnel
『研究: 运动与抑郁风险降低 20%(系统综述)』'Study: exercise reduces depression risk 20% (systematic review)'可信(保留)Credible (reserved)系统综述,证据强度较高Systematic review; higher evidence strength
『朋友圈: 转发抽奖(非官方)』'WeChat: forward to win lottery (unofficial)'不可信Uncredible非官方抽奖 = 高概率诈骗Unofficial lottery = high scam probability
『媒体报道: 某法案通过(引用议会记录)』'Media: bill passed (citing parliament records)'可信Credible引用官方记录,可查证Cites official records; verifiable
『某群: 内部数据,某产品销量第一』'Group: internal data, product #1 in sales'存疑Questionable无来源的『内部数据』,需独立核实'Internal data' without source; needs independent check
『论文: 疫苗安全性(多项 RCT + 长期随访)』'Paper: vaccine safety (multiple RCTs + long follow-up)'可信Credible多 RCT + 长期随访,证据最充分Multiple RCTs + long follow-up; most sufficient evidence
『某博主: 我用 AI 一天写 100 篇文章』'Blogger: I write 100 articles a day with AI'存疑Questionable数量≠质量,可能灌水Quantity ≠ quality; possibly filler
『官方: 空气质量 AQI 50(附站点数据)』'Official: air quality AQI 50 (with station data)'可信Credible官方数据 + 具体站点,可实时验证Official data + specific stations; real-time verifiable
『某群: 不转不是中国人』'Group: not Chinese if you don't forward'不可信Uncredible道德绑架,纯操纵Moral coercion; pure manipulation
『媒体报道: 某研究结果(引用论文 DOI)』'Media: research result (citing paper DOI)'可信Credible给出 DOI 可查原文DOI given; original paper checkable
『某博主: 我测过 100 款产品(无评测方法说明)』'Blogger: I tested 100 products (no method stated)'存疑Questionable无评测方法,可能收广告No test method; possibly paid ads
『官方: 疫苗接种率 80%(附分地区数据)』'Official: vaccine rate 80% (with regional data)'可信Credible官方 + 分地区,可交叉验证Official + regional; cross-verifiable
『某群: 紧急!限今日转发』'Group: Urgent! Forward only today'不可信Uncredible制造紧迫感 = 操纵手法Creating urgency = manipulation tactic
『论文: X 技术提升 30%(附代码开源)』'Paper: Tech X improves 30% (code open-sourced)'可信Credible开源可复现,证据最硬Open-sourced and reproducible; hardest evidence
🧭判断口诀: ① 有没有具体来源(机构/数据/DOI)?② 能不能独立核实?③ 情绪钩子强不强(震惊/紧急/转发积德)?④ 谁在传播(受益方是谁)?四问全过才可信。Judgment formula: ① Specific source (org/data/DOI)? ② Independently verifiable? ③ Strong emotional hook (shock/urgent/forward-for-merit)? ④ Who's spreading (who benefits)? All four pass = credible.
C15

信息源 10 级可信度分级(速查表)

Information Source 10-Level Credibility (Quick Table)

任何信息先定位它在 10 级可信度里哪一级,再决定信多少。
Locate any information at its credibility level (1-10) before deciding how much to trust.
级别Level信息源类型Source Type典型例子Examples可信度Credibility使用建议Usage advice
L10L10原始数据/开源代码Raw data / open-source code统计局原始数据、论文代码仓库Stats raw data, paper code repos最高(可复现)Highest (reproducible)直接信任,但仍要读方法Trust directly; still read methods
L9L9同行评审论文Peer-reviewed papers《Nature》《Science》论文Nature/Science papers很高Very high看样本和方法,注意相关≠因果Check sample & method; correlation ≠ causation
L8L8官方/权威机构发布Official / authoritative releases政府公报、WHO、央行公告Govt bulletins, WHO, central bank高High可信,注意发布动机Credible; watch publishing motive
L7L7主流媒体(交叉验证)Mainstream media (cross-checked)多家权威媒体报道一致Multiple mainstream media agree较高Relatively high多家一致再信Trust when multiple agree
L6L6主流媒体(单一来源)Mainstream media (single source)某报独家Single newspaper exclusive中等Medium等第二来源验证Wait for second source
L5L5专家个人观点Expert personal opinion某教授微博/访谈Professor's Weibo/interview中等(因人而异)Medium (varies)看具体论证,不神化Judge by specific argument, don't deify
L4L4自媒体(有专业背景)Self-media (with expertise)专业领域博主Professional field bloggers中低Medium-low看是否给出来源和方法Check if sources & methods given
L3L3自媒体(无专业背景)Self-media (no expertise)泛流量账号General traffic accounts低Low只当线索,不当结论As leads only, not conclusions
L2L2社交转发/群消息Social forwards / group messages朋友圈转发、群截图WeChat forwards, group screenshots很低Very low默认存疑,逐条核实Default questionable; verify each
L1L1匿名/情绪化信息Anonymous / emotional info『震惊』『内部消息』『不转不是XX』'Shocking', 'inside info', 'not XX if not forwarded'最低(默认假)Lowest (default false)直接标记为不可信Mark uncredible directly
📊分级口诀: 原始数据 > 论文 > 官方 > 主流媒体 > 专家 > 专业自媒体 > 泛自媒体 > 社交转发 > 匿名。级别越低,越要『独立核实』才信。L1-L2 默认存疑。Rule: raw data > papers > official > mainstream media > experts > professional self-media > general self-media > social forwards > anonymous. Lower level = more independent verification needed. L1-L2 default questionable.
C16

20 招识别 AI 幻觉(AI 一本正经地胡说)

20 Techniques to Spot AI Hallucination (AI Speaking Nonsense Confidently)

AI 幻觉 = 自信地生成错误内容。20 招帮你识别,不被『一本正经的胡说』带节奏。
AI hallucination = confidently generating wrong content. 20 techniques to spot it and not be led by 'confident nonsense'.
招数Technique具体做法How to do it为什么有效Why it works
交叉验证Cross-verify同一问题问 3 个不同 AI/搜索,看是否一致Ask 3 different AIs/searches the same question; check consistency幻觉常不一致;事实性内容多源一致Hallucinations often inconsistent; factual content agrees across sources
要求引用来源Demand citations让 AI 给出具体来源(书/论文/数据),然后去核实Ask AI for specific sources (book/paper/data), then verify幻觉常给假来源;真来源可查到Hallucinations often give fake sources; real ones are findable
查具体数字Check specific numbers对 AI 给出的数字/日期/人名,逐一独立核实Independently verify every number/date/name AI gives幻觉最爱编造具体细节Hallucinations love fabricating specific details
问『你确定吗』Ask 'are you sure'直接问 AI 的确定度,看它是否承认不确定Ask AI's confidence; see if it admits uncertainty可靠 AI 会标注不确定;幻觉会硬撑Reliable AI flags uncertainty; hallucinations bluff
要 AI 自我纠错Ask AI to self-correct让 AI 重新检查自己的回答,找矛盾Ask AI to re-check its answer for contradictions幻觉常有内部矛盾;事实通常自洽Hallucinations often have internal contradictions; facts are usually self-consistent
看是否过度具体Watch over-specificityAI 给出『第 3 页』『2019 年 5 月』等过度具体细节时警惕Be wary when AI gives over-specific details like 'page 3', 'May 2019'幻觉爱堆砌伪具体细节来装可信Hallucinations pile pseudo-specifics to seem credible
查原始出处Check original sourceAI 引『某书某页』,去查原书是否有此内容When AI cites 'a book, page X', check if the book has that content幻觉常『张冠李戴』引不存在的页Hallucinations often cite non-existent pages
对比已知事实Compare known facts用你已知的常识/事实检验 AI 答案Test AI's answer against your known common sense/facts与常识矛盾 = 大概率幻觉Contradicts common sense = likely hallucination
看语气是否绝对Check absolute toneAI 用『绝对』『必然』『100%』等绝对词时警惕Be wary when AI uses absolute words like 'absolutely', 'certainly', '100%'可靠表述留余地;幻觉爱下绝对结论Reliable wording leaves room; hallucinations make absolute claims
多轮追问Multi-round follow-up对 AI 答案层层追问,看是否前后矛盾Follow up AI's answer in layers; check for contradictions幻觉在追问下容易崩Hallucinations collapse under follow-up
让 AI 列证据链Ask for evidence chain要求 AI 给出『结论的依据是什么』Ask AI for 'what's the basis of this conclusion'幻觉给不出证据链;事实可追溯Hallucinations can't give evidence chains; facts are traceable
查数据时效Check data timelinessAI 引用的数据是否过时(用了旧年份数据)Check if AI's cited data is outdated (old-year data)幻觉/旧模型用过时数据Hallucinations/old models use outdated data
要 AI 承认局限Ask AI to admit limits问 AI『这个问题你确定答案吗?有没有可能错』Ask AI 'are you sure of this answer? Could it be wrong'可靠 AI 会承认局限;幻觉会硬撑Reliable AI admits limits; hallucinations bluff
交叉 AI 比对Cross-AI compare同一问题问不同模型(ChatGPT/Claude/DeepSeek),对比差异Ask different models the same question; compare differences幻觉在不同模型间不一致Hallucinations inconsistent across models
查反例Find counterexamples问 AI『有没有反例/例外』Ask AI 'are there counterexamples/exceptions'幻觉给不出反例;事实有边界Hallucinations can't give counterexamples; facts have boundaries
看是否编造人名/机构Check fabricated names/institutionsAI 给出的专家名/机构名,去搜是否存在Search if AI-given expert names/institutions exist幻觉常编造不存在的专家/机构Hallucinations often fabricate non-existent experts/institutions
要求可复现步骤Ask for reproducible steps对方法类答案,要求给出可复现的具体步骤For method answers, ask for reproducible concrete steps幻觉给不出可操作细节;真方法可执行Hallucinations can't give actionable details; real methods are executable
查逻辑链Check logic chain检验 AI 的推理链条,找逻辑断裂处Check AI's reasoning chain; find logic breaks幻觉常跳过关键推理步骤Hallucinations often skip key reasoning steps
看是否过度自信Watch over-confidenceAI 语气越自信,越要警惕(『这很简单』『答案就是』)The more confident AI's tone, the more wary ('this is simple', 'the answer is')自信度与正确度无关;幻觉特别自信Confidence ≠ correctness; hallucinations are especially confident
最终核实关键结论Final verify key conclusions对任何关键结论,最后都用独立来源核实一次For any key conclusion, finally verify once with an independent source关键决策不能依赖单一 AI 来源Key decisions can't rely on a single AI source
⚠️AI 幻觉没有统一『味道』。最可靠的识别 = 对关键信息,永远用独立来源核实一次。AI 是『帮你整理线索』,不是『替你判断真假』。AI hallucinations have no unified 'flavor'. The most reliable identification = always verify key info once with an independent source. AI is 'helping you organize leads', not 'judging truth for you'.
C18

30 个实战场景:批判性思维怎么用

30 Real-World Scenarios: How to Apply Critical Thinking

30 个真实场景,示范批判性思维怎么落地,不是空谈。
30 real scenarios showing how critical thinking lands in practice, not empty talk.
场景Scenario常见陷阱Common trap批判性思维做法Critical thinking action效果Effect
刷到健康养生文章Reading health/wellness articles『某食物治百病』绝对化'Some food cures everything' absolutization查来源/查研究/看是否夸大Check source/study/overclaiming避免被养生谣言骗Avoid being fooled by wellness myths
看到投资消息Seeing investment messages『稳赚不赔』『内部消息』'Guaranteed profit', 'inside info'查资质/查风险/查谁受益Check qualifications/risk/who benefits避免被投资骗局骗Avoid being fooled by investment scams
听领导/上级说法Hearing boss/superior claims权威=正确Authority = correct看数据,不当场否定但私下验证Look at data; don't refute on the spot but verify privately既尊重又不盲从Respect without blindly following
读新闻报道Reading news reports单一信源下结论Concluding from single source找第二信源交叉验证Find second source for cross-verification避免被单侧信息误导Avoid being misled by one-sided info
看产品评测Reading product reviews『亲测好用』无方法'Personally tested, great' with no method看评测方法/样本/是否收广告Check test method/sample/whether paid ad做出真实购买决策Make real purchase decisions
听朋友转述Hearing friends' retellings传话失真Message distortion问原始来源/直接核实Ask original source/verify directly避免二手信息偏差Avoid second-hand info bias
看 AI 回答Reading AI answersAI 自信=正确AI confidence = correct交叉验证/查来源/独立核实Cross-verify/check source/independently verify避免被 AI 幻觉带节奏Avoid being led by AI hallucinations
做重大决策Making major decisions拍脑袋/情绪化Gut decision/emotional列证据/找反例/延迟决策List evidence/find counterexamples/delay decision降低决策失误Reduce decision errors
面试听候选人说Hearing candidates in interviews『我做过很多大项目』空泛'I've done many big projects' vague问具体细节/要案例/追问Ask concrete details/ask for cases/follow up识别真实能力Identify real ability
看社交媒体热点Seeing social media hot topics情绪化跟风Emotional bandwagon冷静看事实/多方观点Stay calm, see facts/multiple views避免被情绪带节奏Avoid being led by emotion
读论文Reading papers『相关』当『因果』Treating 'correlation' as 'causation'看方法/样本/因果论证Check method/sample/causal argument正确解读研究Correctly interpret research
听专家访谈Hearing expert interviews专家全领域正确Experts correct in all fields看具体论证/区分领域Check specific argument/distinguish fields避免神化专家Avoid deifying experts
看数据图表Reading data charts只看到想看的Only seeing what you want to see看样本/时间/口径/基线Check sample/time/caliber/baseline正确读数据Correctly read data
做竞品分析Doing competitor analysis『他们不行』偏见'They're not good' bias看客观数据/承认对手强项Look at objective data/acknowledge opponents' strengths客观评估竞争Objectively assess competition
选专业/职业Choosing major/career『热门』=好'Popular' = good看个人匹配/长期趋势/真实数据Check personal fit/long-term trends/real data避免追热点选错Avoid chasing trends and choosing wrong
投资理财Investing『过去涨=未来涨』'Past rise = future rise'看基本面/分散/风险Check fundamentals/diversify/risk降低投资风险Reduce investment risk
看医疗建议Reading medical advice『偏方治大病』'Folk cures for big diseases'查正规医学指南/咨询医生Check formal medical guidelines/consult doctor避免误诊误治Avoid misdiagnosis/mistreatment
听销售话术Hearing sales talk『只剩最后 3 件』制造紧迫'Only 3 left' creating urgency冷静判断真实需求/延迟Calmly judge real need/delay避免冲动消费Avoid impulsive buying
做内容创作Doing content creation『大家都这么说』'Everyone says so'独立核实/给出来源Independently verify/give sources做出可信内容Make credible content
看政治言论Reading political statements情绪化站队Emotional taking-sides看事实/多方立场/证据See facts/multiple positions/evidence理性看待政治Politically rational
评估自己工作Evaluating own work『我觉得做得好』自我高估'I think I did well' self-overestimate找反馈/客观指标/对照Find feedback/objective metrics/benchmark客观自我评估Objective self-evaluation
看广告Reading ads『全网销量第一』无证明'No.1 sales online' without proof查数据来源/是否可验证Check data source/whether verifiable避免被广告带节奏Avoid being led by ads
听孩子/学生说Hearing children/students say『其他同学都有』'Other classmates all have'独立核实/不盲从Independently verify/don't blindly follow理性教育决策Rational education decisions
做技术选型Doing tech selection『某框架最火』'Some framework is hottest'看场景/长期维护/团队匹配Check scenario/long-term maintenance/team fit做出合适技术决策Make suitable tech decisions
看新闻标题Reading news headlines标题党下结论Concluding from clickbait titles读全文/看原报道Read full text/see original report避免被标题误导Avoid being misled by headlines
评估 AI 工具Evaluating AI tools『最强』营销话术'Strongest' marketing talk看具体测试/自己场景实测Check specific tests/self-scenario actual test选出合适工具Choose suitable tool
做商业判断Making business judgments『直觉』代替分析'Intuition' replacing analysis列数据/找案例/算概率List data/find cases/calculate probabilities降低商业判断失误Reduce business judgment errors
看养生/饮食建议Reading wellness/diet advice『某食物排毒』伪科学'Some food detoxes' pseudoscience查正规医学/区分食物功效Check formal medicine/distinguish food effects避免被伪科学骗Avoid being fooled by pseudoscience
听会议发言Hearing meeting remarks『一直如此』路径依赖'It's always been this way' path dependence问『为什么』/找改进点Ask 'why'/find improvement points打破思维定式Break thinking patterns
做学习决策Making learning decisions『某课程包过』承诺'Some course guarantees passing'看真实学员反馈/课程质量Check real student feedback/course quality避免报假课程Avoid signing up for fake courses
🎯批判性思维不是『怀疑一切』,而是『该怀疑的怀疑,该信任的信任』。核心是:对关键结论,永远多问一句『证据呢?反例呢?谁受益?』Critical thinking is not 'doubting everything', but 'doubt what should be doubted, trust what should be trusted'. Core: for key conclusions, always ask one more: 'What's the evidence? What are the counterexamples? Who benefits?'
C19

50 个『批判性思维』扩充问答

50 Extended 'Critical Thinking' Q&A

50 个更细的批判性思维问题,覆盖日常/工作/学习/社交。
50 finer critical-thinking questions, covering daily/work/study/social.
不会。批判性思维对事不对人:质疑观点,不攻击人。抬杠是情绪化反驳;批判是讲证据地探讨。No. Critical thinking targets ideas, not people: question views, don't attack people. Arguing is emotional rebuttal; critique is evidence-based discussion.
质疑是『我想知道证据』;不信任是『我预设你错了』。质疑带着好奇心,不信任带着预设。Questioning is 'I want to know the evidence'; distrust is 'I presume you're wrong'. Questioning carries curiosity; distrust carries presumption.
对信息批判,对人尊重。可以问『这个你怎么看到的?』,而不是『你错了』。Be critical of info, respectful to people. You can ask 'How did you see this?' instead of 'You're wrong'.
不需要专业数学,但需要基本的概率/统计直觉(样本大小、相关≠因果)。No professional math needed, but basic probability/statistics intuition (sample size, correlation ≠ causation) is required.
杠精为了赢,批判为了真。杠精只反驳不建设;批判指出问题并给替代方案。Nitpickers win at all costs; critical thinkers seek truth. Nitpickers only refute; critique points out problems and offers alternatives.
更要。AI 让信息量暴增 + 幻觉普及,判断力是稀缺资源。More so. AI explodes info volume + spreads hallucinations; judgment is the scarce resource.
多问『你怎么知道?』『还有别的可能吗?』,鼓励他们自己查证据,不直接给答案。Ask more 'How do you know?' 'Any other possibilities?', encourage them to check evidence themselves, don't just give answers.
不会。它让你『小事快判、大事深验』,总体是省力的。No. It lets you 'judge small things fast, verify big things deeply', saving effort overall.
定期问自己:我是被证据说服的,还是被情绪/从众说服的?Periodically ask yourself: am I persuaded by evidence, or by emotion/crowd?
不是。创意需要发散,批判需要收敛,两者交替进行。No. Creativity needs divergence, critique needs convergence; they alternate.
交叉验证 + 查来源 + 独立核实。AI 是线索,不是结论。Cross-verify + check sources + independent verification. AI is a lead, not a conclusion.
极重要。重大决策失误大多来自『没验证就下结论』。Extremely important. Most major decision errors come from 'concluding without verification'.
看是否有具体成果/可查资历/是否过度自信/是否拒绝被质疑。Check for specific achievements/verifiable credentials/excessive confidence/refusal to be questioned.
找持相反观点的可靠来源,看对方最强论证(steelman),再判断。Find reliable sources of the opposite view, read their strongest argument (steelman), then judge.
识别营销话术/伪数据/情绪化决策。投资靠的是判断力,不是消息灵通。Identify marketing talk/fake data/emotional decisions. Investing relies on judgment, not insider info.
主动找反证;问『如果我的观点是错的,证据会是什么』。Actively seek counter-evidence; ask 'If my view were wrong, what evidence would there be?'.
主动接触不同立场/平台;定期看与己不同的内容。Actively contact different stances/platforms; periodically view content different from your own.
让论证严密、证据充分、逻辑自洽。Make arguments rigorous, evidence sufficient, logic self-consistent.
事实可验证(对/错);观点不可验证(好/坏、应/不应)。Facts are verifiable (true/false); opinions are not (good/bad, should/shouldn't).
不一定。数据可以是错的/被筛选的/口径不一的。要看数据质量。Not necessarily. Data can be wrong/filtered/inconsistent-caliber. Must look at data quality.
有用。能评估候选人/识别空话/追问细节。Useful. Can evaluate candidates/identify empty talk/follow up on details.
检验每一步因果是否真实成立,而不是『A 会无限恶化』。Verify each causal step is actually true, not 'A will infinitely worsen'.
看证据,不只看身份。权威也可能是错的/过时的。Look at evidence, not just identity. Authorities can also be wrong/outdated.
能当『陪练』:让它找反例/查逻辑/模拟对手。但最终判断要自己做。Can be a 'sparring partner': let it find counterexamples/check logic/simulate opponents. But final judgment is yours.
识别用户需求真伪/避免自我感动/用数据验证假设。Identify real vs fake user needs/avoid self-moved/verify hypotheses with data.
扩大样本;标注『个例』与『规律』的区别。Enlarge sample; mark the difference between 'case' and 'pattern'.
不算。沉默不等于同意;要区分『没表态』和『反对』。No. Silence ≠ agreement; distinguish 'no stance' from 'opposition'.
极有用。避免方法偏差/过度解读/自证预言。Extremely useful. Avoids method bias/over-interpretation/self-fulfilling prophecy.
看内容是否用恐惧/愤怒/羞耻推动你行动,而不是用证据说服你。Check if content uses fear/anger/shame to push you to act, instead of persuading with evidence.
都是错,但幻觉更自信、细节更伪具体。都要用独立来源核实。Both are wrong, but hallucinations are more confident, details more pseudo-specific. Both need independent verification.
有用。识别对方话术/找共同利益/不情绪化让步。Useful. Identify opponent's talk/find common interests/don't concede emotionally.
只看未来收益,忽略已投入。问『如果我现在开始,会继续吗』。Only look at future returns, ignore past investment. Ask 'If I started now, would I continue?'.
不是。权威是身份,可信是证据。要看具体论证质量。No. Authority is identity; credible is evidence. Must look at specific argument quality.
有用。营造『质疑观点不质疑人』的文化,避免群体思维。Useful. Cultivate 'question ideas not people' culture, avoid groupthink.
鼓励异见/指定『魔鬼代言人』/匿名收集意见。Encourage dissent/appoint 'devil's advocate'/collect opinions anonymously.
找因果机制/做对照实验/检查第三变量。Find causal mechanism/do controlled experiments/check third variables.
减少碎片信息/定期深度阅读/关掉非必要通知。Reduce fragmented info/periodic deep reading/turn off unnecessary notifications.
有用。识别虚假/谣言/操纵内容。Useful. Identify fake/rumor/manipulative content.
看来源/样本/方法/时间/口径,五要素齐全才可信。Check source/sample/method/time/caliber; all five complete = credible.
给对立观点等量时间,但其中一方无实质证据。Give opposing views equal time, but one side has no substantial evidence.
两者都有。可靠 AI 在不确定时留余地;但要追问逼出具体答案。Both. Reliable AI leaves room when uncertain; but follow up to force specific answers.
区分『该验证的』(关键决策)和『可接受的』(日常小事)。Distinguish 'what needs verification' (key decisions) from 'what's acceptable' (daily trivialities).
RCT > 队列研究 > 病例报告 > 专家意见 > 个人经验。RCT > cohort study > case report > expert opinion > personal experience.
仍重要。批判需要『已知事实』作为对照,没记忆就无法检验。Still important. Critique needs 'known facts' as reference; without memory, can't verify.
遇到冲击性信息,先停 10 分钟再下结论。On shocking info, pause 10 minutes before concluding.
极有用。识别粉饰/伪数据/回避问题的公司。Extremely useful. Identify whitewashing/fake data/companies avoiding questions.
第一性原理是『回到基本事实』,是批判性思维的核心方法之一。First principles is 'return to basic facts', one of critical thinking's core methods.
极重要。会提问 = 会引导 AI 给出高质量答案。Extremely important. Ability to question = ability to guide AI to high-quality answers.
💡50 个扩充问题,每天读 2 个,一个月读完。核心思想只有一个:对关键结论,永远多问一句证据。50 extended questions, read 2 daily, finish in a month. Core idea is just one: for key conclusions, always ask one more for evidence.
🎓 世界数字教育网 · World Digital Education Network · 让每个准备好的人,都能高效学习
© 2026 世界数字教育网 · 内容基于公开研究,仅供学习参考 · 免费学 · 不卖课 · 不承诺收益
© 2026 World Digital Education Network · Based on public research, for learning only · Free · No courses · No income promises
C21

自测:你的 AI 学外语能力

Quiz: Your AI Language-Learning Ability

10 题,点选项即判对错 + 解析
10 questions, click to judge + explain
Q1
AI 给了你一个『数据』,你的第一反应应该是?AI gives you a 'data point'; what should your first reaction be?
先问『数据从哪来的?哪年的?』First ask 'where's the data from? which year?'直接相信(AI 说的应该对)Believe it directly (AI is probably right)只看不信(什么都不信)Neither believe nor look (don't believe anything)
✅ 正确:先问『数据从哪来的?哪年的?』
✅ Correct: First ask 'where's the data from? which year?'
Q2
『研究表明 80% 的人...』但没给出处,这是?'Research shows 80% of people...' but no source given; this is?
红旗:权威无出处(幻觉高风险)Red flag: authority without source (high hallucination risk)好证据(有研究支撑)Good evidence (supported by research)无关紧要(80% 是大概数)Irrelevant (80% is approximate)
✅ 正确:红旗:权威无出处(幻觉高风险)
✅ Correct: Red flag: authority without source (high hallucination risk)
Q3
AI 说『这绝对是正确的』,这代表什么?AI says 'this is absolutely correct'; what does this represent?
什么都代表不了——AI 的置信度和正确率无关Represents nothing — AI's confidence is unrelated to accuracyAI 很确定,应该对AI is very sure, should be right需要更多 AI 确认Need more AI confirmation
✅ 正确:什么都代表不了——AI 的置信度和正确率无关
✅ Correct: Represents nothing — AI's confidence is unrelated to accuracy
Q4
『允许用 AI 会导致人类灭亡』是哪种谬误?'Allowing AI will lead to humanity's extinction' is which fallacy?
滑坡谬误(中间环节无证据)Slippery slope (middle steps have no evidence)诉诸权威Appeal to authority稻草人Straw man
✅ 正确:滑坡谬误(中间环节无证据)
✅ Correct: Slippery slope (middle steps have no evidence)
Q5
批判性思维的核心产出是什么?What is the core output of critical thinking?
在现有证据下『最可能的结论』The 'most likely conclusion' given current evidence怀疑一切(不结论)Doubt everything (no conclusion)永远正确Always be right
✅ 正确:在现有证据下『最可能的结论』
✅ Correct: The 'most likely conclusion' given current evidence
Q6
看到『最后一天!仅剩 3 件!』,最该做什么?Seeing 'LAST DAY! Only 3 left!', what's most to do?
剥离时间压力,再看事实(库存真的只剩 3 吗)Strip the time pressure, then look at facts (are there really only 3 left)立刻买(错过就没了)Buy immediately (miss it and it's gone)不看不听Don't look, don't listen
✅ 正确:剥离时间压力,再看事实(库存真的只剩 3 吗)
✅ Correct: Strip the time pressure, then look at facts (are there really only 3 left)
Q7
判断一个主张的证据等级,正确的是?Judging a claim's evidence level, the correct one is?
RCT > 观察研究 > 专家意见 > 个例RCT > observational study > expert opinion > anecdote个例 > 专家意见 > RCTAnecdote > expert opinion > RCT都是同一等级All the same level
✅ 正确:RCT > 观察研究 > 专家意见 > 个例
✅ Correct: RCT > observational study > expert opinion > anecdote
Q8
『大家都这么认为』作为理由,问题是?'Everyone thinks so' as a reason, the problem is?
从众不是判断,是放弃判断Following the crowd is not judgment, it's giving up judgment说明证据充分Shows sufficient evidence说明样本大Shows large sample
✅ 正确:从众不是判断,是放弃判断
✅ Correct: Following the crowd is not judgment, it's giving up judgment
Q9
AI 时代最基础的批判性动作是?The most basic critical action in the AI era is?
不轻信 + 验证(来源/时效/反例)Don't trust lightly + verify (source/timeliness/counterexample)完全不用 AINever use AI完全信任 AICompletely trust AI
✅ 正确:不轻信 + 验证(来源/时效/反例)
✅ Correct: Don't trust lightly + verify (source/timeliness/counterexample)
Q10
下结论前问『什么证据能推翻我』,这是?Before concluding, asking 'what evidence could overturn me' is?
可证伪性检查(好判断的标志)Falsifiability check (sign of good judgment)多此一举Unnecessary自我怀疑Self-doubt
✅ 正确:可证伪性检查(好判断的标志)
✅ Correct: Falsifiability check (sign of good judgment)
C22

AI 学外语 20 项行动清单

AI Language-Learning 20-Item Action List

20 个可立即执行的动作,打勾跟踪
20 executable actions, track with checks
  1. 今天有没有『不验证就相信』过一条信息?Did you 'believe without verifying' a piece of info today?
  2. 今天的判断,证据等级够吗?Does today's judgment have enough evidence level?
  3. 有没有只看了『一边』的信息就下结论?Did you conclude after only seeing 'one side' of info?
  4. 决策前标注过情绪了吗?Did you label emotions before deciding?
  5. 有没有用『个例』代替『数据』?Did you use 'anecdote' instead of 'data'?
  6. AI 给的结论,追问过证据吗?Did you follow up on evidence for AI's conclusion?
  7. 信息有时效检查吗(哪年的信息)?Did you check info timeliness (which year)?
  8. 有没有识别到至少 1 个红旗信号?Did you spot at least 1 red-flag signal?
  9. 重要决策做过『预-mortem』吗?Did you do a 'pre-mortem' for important decisions?
  10. 有没有替反方说过最强论证(steelman)?Did you state the strongest counter-argument (steelman)?
  11. 判断给了置信度%吗?Did you assign a confidence % to the judgment?
  12. 有没有『相关≠因果』的陷阱?Any 'correlation ≠ causation' traps?
  13. 信源单一吗?(≥2 信源了吗)Single source? (Did you get ≥2 sources?)
  14. 有没有被『紧迫感』操纵过?Were you manipulated by 'urgency'?
  15. 决策日志记了吗?Did you record in the decision journal?
  16. 有没有『身份偏见』(因立场而信)?Any 'identity bias' (believing due to stance)?
  17. 样本量够吗?(<30 只算个例)Sample size enough? (<30 only counts as anecdote)
  18. 有没有『幸存者偏差』(只看成功的)?Any 'survivorship bias' (only seeing success)?
  19. 情绪是信号还是证据?分清了?Is emotion a signal or evidence? Did you separate them?
  20. 今天『会问』了还是只『会答』?Did you 'ask well' today or just 'answer'?
C23

15 条 AI 学外语 FAQ

15 AI Language-Learning FAQ

15 条最高频的问题,直接给答案
15 most high-frequency questions, direct answers
不会。抬杠是『为了赢』,批判性思维是『为了对』。你质疑的是信息不是人,且承认自己可能错。会好好质疑的人反而更受欢迎。No. Being argumentative is 'for winning', critical thinking is 'for being right'. You question info not people, and admit you may be wrong. People who question well are actually more popular.
能。轻信是『默认设置』,批判性思维是『训练后的设置』。从最小的动作开始:每天验证 1 条信息。3 个月后变成习惯。Yes. Trusting is the 'default setting', critical thinking is the 'trained setting'. Start with the smallest action: verify 1 piece of info daily. In 3 months it becomes a habit.
分场景:小事 60 分快速判断(不纠结),大事 90 分深度验证(值得花)。不是所有事都值得深度批判,『选择』是批判性思维的一部分。By scenario: small things 60-point quick judgment (don't obsess), big things 90-point deep verification (worth it). Not everything deserves deep criticism; 'choosing' is part of critical thinking.
能当『陪练』不能当『裁判』。AI 能帮你找反例、列证据等级、检查逻辑漏洞,但 AI 自己也会犯逻辑错误。最终判断必须你来做。Can be a 'sparring partner' not a 'referee'. AI can help find counterexamples, list evidence levels, check logic gaps, but AI itself makes logic errors too. Final judgment must be yours.
从『为什么』开始:孩子问『为什么』时别敷衍,让他自己找答案。6 岁可以玩『找不同』(识别信息差异),10 岁可以讨论『这条新闻可信吗』。Start with 'why': when kids ask 'why' don't brush them off, let them find answers themselves. 6-year-olds can play 'spot the difference' (identify info differences), 10-year-olds can discuss 'is this news credible'.
不会。它是『快速筛选 + 深度验证』:小事快速放过,大事深度处理。真正导致失眠的是『反刍』(反复想同一件事),那是情绪问题不是思维问题。No. It's 'quick filter + deep verification': small things quickly let go, big things deeply processed. What really causes insomnia is 'rumination' (repeatedly thinking the same thing), that's an emotion problem not a thinking problem.
都不一定。AI 的观点可能错(幻觉/讨好),你的观点也可能错(偏差)。正确做法:双方都列证据,看谁的证据等级高。『和 AI 不一致』不是错,『不验证』才是错。Neither necessarily. AI's view may be wrong (hallucination/people-pleasing), your view may be wrong (bias). Correct approach: both list evidence, see whose evidence level is higher. 'Disagreeing with AI' isn't wrong; 'not verifying' is wrong.
识别『情绪驱动』和『确认偏差』。投资亏钱最多的不是『选错』,而是『追涨杀跌』(情绪)和『只看好自己持有的』(偏差)。批判性思维帮你『客观评估风险』。Spotting 'emotion-driven' and 'confirmation bias'. Biggest investment losses aren't from 'choosing wrong' but 'chasing highs, cutting lows' (emotion) and 'only bullish on what you hold' (bias). Critical thinking helps 'objectively assess risk'.
『分级处理』:① 影响决策的信息→深度验证;② 影响认知但非决策→看来源可信度;③ 纯资讯→扫一眼标题判断价值。不是所有信息都值得读完。'Tiered processing': ① info affecting decisions → deep verify; ② info affecting cognition but not decisions → check source credibility; ③ pure news → skim title to judge value. Not all info is worth reading fully.
相关但不一样。理性 = 不被情绪主导,批判性思维 = 有系统地检验信息。一个人可以『理性』(冷静)但『不批判』(轻信冷静说出的假话)。两者配合最好。Related but different. Rationality = not dominated by emotion, critical thinking = systematically testing info. One can be 'rational' (calm) but 'not critical' (trusting calm-stated lies). Both together are best.
三点:① 他在这个具体领域是权威吗(跨界不权威);② 他有没有可追溯的同行评审;③ 他的结论有没有利益冲突。三点都过才信。Three points: ① Is he an authority in this specific field (cross-field isn't); ② Does he have traceable peer review; ③ Does his conclusion have conflict of interest. Pass all three before trusting.
友好。批判性思维不反对信仰,它训练『理性评估』信仰依据。经得起检验的信仰更扎实。真正的矛盾不是『信仰 vs 理性』,是『盲信 vs 信仰』。Friendly. Critical thinking doesn't oppose faith; it trains 'rational evaluation' of faith's basis. Faith that withstands testing is more solid. The real conflict isn't 'faith vs reason', it's 'blind faith vs faith'.
『不确定』比『绝对』更可信——至少诚实。但『不确定』不等于『错』,需要你自己验证。AI 的『不确定性』有时是真实的,有时是『懒得编』的借口,用证据判断。'Not sure' is more trustworthy than 'absolute' — at least honest. But 'not sure' ≠ 'wrong', you need to verify yourself. AI's 'uncertainty' is sometimes real, sometimes an excuse for 'too lazy to fabricate', judge by evidence.
不会。天真 = 不加分辨地相信,批判性思维 = 看清后依然选择相信/不信任。看清世界后依然有善意,是『更高级的纯真』,不是失去。No. Innocence = trusting without discrimination, critical thinking = choosing to trust/not trust after seeing clearly. Still having goodwill after seeing the world clearly is 'higher innocence', not losing it.
15 分钟足够。每天:① 验证 1 条信息(5 分钟);② 标注 1 次情绪(1 分钟);③ 决策日志记 1 条(5 分钟);④ 金句读 1 条(2 分钟)。贵在坚持不在时长。15 minutes is enough. Daily: ① verify 1 piece of info (5 min); ② label 1 emotion (1 min); ③ decision journal record 1 (5 min); ④ read 1 golden quote (2 min). Consistency matters more than duration.
C24

参考来源

References

综合 CEFR 标准、语言学习研究 + AI 工具官方资料(2026)
Synthesizes CEFR, language-learning research + official AI tool materials (2026)
  1. 《思考,快与慢》(Thinking, Fast and Slow) — 丹尼尔·卡尼曼Daniel Kahneman, 'Thinking, Fast and Slow'
  2. 《批判性思维工具》(Critical Thinking Tools) — 理查德·保罗Richard Paul, 'Critical Thinking Tools'
  3. 逻辑谬误大全 (List of logical fallacies) — 维基百科List of logical fallacies — Wikipedia
  4. 可证伪性 (Falsifiability) — 维基百科Falsifiability — Wikipedia
  5. 证据等级 (Levels of evidence) — 维基百科Levels of evidence — Wikipedia
  6. 认知偏差 (Cognitive bias) — 维基百科Cognitive bias — Wikipedia
  7. 确认偏差 (Confirmation bias) — 维基百科Confirmation bias — Wikipedia
  8. 幸存者偏差 (Survivorship bias) — 维基百科Survivorship bias — Wikipedia
  9. 贝叶斯推理 (Bayesian inference) — 维基百科Bayesian inference — Wikipedia
  10. 第一性原理 (First principles) — 维基百科First principles — Wikipedia
  11. Our World in Data(数据可视化)Our World in Data (data visualization)
  12. FactCheck.org(事实核查)FactCheck.org (fact-checking)
  13. Khan Academy(逻辑课程)Khan Academy (logic courses)
  14. Coursera(批判性思维课程)Coursera (critical thinking courses)
  15. 《AI 幻觉与可靠性研究》综述 (AI hallucination & reliability surveys)Surveys on AI hallucination & reliability