AI 素养:数字时代的读写能力
AI Literacy: The New Literacy of the Digital Age
AI 素养 = 理解 AI 能做什么、不能做什么、如何安全高效地用它。
AI literacy = knowing what AI can and can't do, and how to use it safely and well.
什么是 AI 素养?为什么 2026 年人人必修
What is AI Literacy? Why It's a Core Skill in 2026
2023 年,世界经济论坛把「AI 素养」列入未来工作技能清单;2024 年,欧盟《AI 法案》要求成员国把 AI 素养纳入义务教育;2025 年,全球 97% 的头部企业把「会用 AI」写进员工能力模型。AI 素养已经不是『加分项』,而是像读写能力一样的『基础能力』。
In 2023, the World Economic Forum listed AI literacy among future work skills; in 2024, the EU AI Act required member states to include AI literacy in compulsory education; in 2025, 97% of leading companies wrote 'can use AI' into their competency models. AI literacy is no longer a bonus — it is a basic capability, like reading and writing.
但『会用 AI』有三层含义:会用工具(打字问问题)、会用方法(提示工程+验证+组合)、会用思维(判断边界、守住伦理、保持自主思考)。大多数人停留在第一层,真正拉开差距的是第三层。
But 'using AI' has three layers: using tools (typing questions), using methods (prompting + verification + combining), and using thinking (judging boundaries, holding ethics, keeping independent thought). Most people stay at layer one; the real gap is created at layer three.
| 层级 | Layer | 定义 | Definition | 典型表现 | Typical behavior |
|---|---|---|---|---|---|
| 第 1 层:操作 | L1: Operation | 会把 AI 当搜索引擎用 | Uses AI as a search engine | 『帮我写一段话』『翻译一下』 | 'Write me a paragraph','translate this' |
| 第 2 层:方法 | L2: Method | 掌握提示工程与验证 | Masters prompting & verification | 会拆解任务、给上下文、交叉验证 | Decomposes tasks, gives context, cross-checks |
| 第 3 层:思维 | L3: Thinking | 理解原理、边界与伦理 | Understands principles, boundaries, ethics | 知道何时不用 AI、如何保持判断力 | Knows when NOT to use AI, keeps judgment |
大语言模型是如何工作的(通俗版)
How Large Language Models Actually Work (Plain Edition)
一句话原理
One-Sentence Principle
大语言模型本质上是一个『超级预测机』:它读过海量文本,学会了『前面这些字后面最可能接什么字』。它不是『理解』,而是『概率预测』。这个区别决定了它的一切优点和缺点。
A large language model is essentially a 'super-predictor': it has read massive text and learned 'what word most likely follows these words'. It does not 'understand' — it does 'probability prediction'. This distinction determines all its strengths and weaknesses.
| 概念 | Concept | 通俗解释 | Plain explanation | 对你意味着什么 | What it means for you |
|---|---|---|---|---|---|
| Token | Token | 文本切成的『字块』,中文约 1 字=1~2 个 token | Text split into chunks; Chinese ≈1 char = 1~2 tokens | 长文会超出模型的『上下文窗口』,需要分段喂 | Long text exceeds the context window — feed it in chunks |
| 上下文窗口 | Context window | 模型一次能『记住』的文本总量,主流 12.8 万~100 万 token | Total text a model can 'remember' at once; mainstream 128K~1M tokens | 把关键信息放在开头和结尾,中间容易被忽略 | Put key info at the beginning and end; the middle gets ignored |
| 温度 | Temperature | 控制输出随机性:0=最确定,1+ 更有创意 | Controls randomness: 0 = most deterministic, 1+ more creative | 要准确答案调低,要创意调高 | Lower for accuracy, higher for creativity |
| 训练数据 | Training data | 模型『读过的书』,有截止日期 | The 'books' the model read — with a cutoff date | 它不知道截止日之后发生的事,可能给过时信息 | It doesn't know events after the cutoff date — may give outdated info |
| 幻觉 | Hallucination | 模型编造不存在的事实,但语气很自信 | Model invents facts with confident tone | 必须对关键信息交叉验证 | Always cross-verify critical info |
| 对齐 | Alignment | 训练模型『说人话、守规矩』的过程 | Training the model to 'speak human, follow rules' | 它被要求礼貌、拒绝危险请求——但也可能过度拒绝 | It's asked to be polite and refuse risky requests — but may over-refuse |
- 『记住』不等于『理解』:模型可以背下整本词典,但不认识『苹果』是水果'Remembering' ≠ 'understanding': a model can memorize a whole dictionary but not know an apple is a fruit
- 『上下文』是工作记忆:它不会把这次对话内容存进长期记忆(除非产品做了记忆功能)'Context' is working memory: it does not store this conversation into long-term memory (unless the product adds memory features)
- 『概率』是双刃剑:让答案流畅自然,也让它敢编造'Probability' is a double-edged sword: it makes answers fluent and natural, but also makes it dare to fabricate
AI 的能力边界:哪些事擅长、哪些事不行
AI Capability Boundaries: What It's Good At, What It Can't Do
| 任务类型 | Task type | AI 表现 | Performance | 正确用法 | Right way |
|---|---|---|---|---|---|
| 知识问答 | Knowledge Q&A | ⭐⭐⭐⭐ 强,但有幻觉风险 | Strong but hallucination risk | 问它当『起点』,再自己查证 | Ask it as a 'starting point', then verify yourself |
| 写作起草 | Writing drafts | ⭐⭐⭐⭐ 快且流畅,但同质化 | Fast and fluent but homogenized | 让它出初稿,你改结构和灵魂 | Have it produce a first draft, you reshape the soul |
| 代码生成 | Code generation | ⭐⭐⭐⭐ 常见库很熟,边缘场景会错 | Good with common libraries, wrong in edge cases | 逐段审查 + 跑测试,别整段照抄 | Review each segment + run tests, don't copy wholesale |
| 翻译 | Translation | ⭐⭐⭐⭐ 通用文本好,专业术语要核对 | Good for general text, check technical terms | 通用直译 + 术语人工核对 | Direct translate + manual term check |
| 数据分析 | Data analysis | ⭐⭐⭐ 会写代码但可能算错,要验证 | Writes code but may compute wrong, verify | 给它数据让它写代码,你验结果 | Give it data to write code, you verify the result |
| 实时信息 | Real-time info | ⭐⭐ 训练有截止日期,除非联网 | Training has a cutoff, unless it has web access | 重大事件先查权威来源,再用 AI 整理 | For major events, check authoritative sources first, then let AI organize |
| 数学/逻辑 | Math & logic | ⭐⭐⭐ 简单题强,多步推理易断链 | Strong on simple, breaks on multi-step reasoning | 复杂推理让它『一步步想』并自己复查 | For complex reasoning, have it 'think step by step' and double-check |
| 创意 | Creativity | ⭐⭐⭐ 组合已有,不真正『原创』 | Combines existing, not truly 'original' | 用 AI 扩脑暴,人类定调 | Use AI to brainstorm, humans set the tone |
| 情绪/心理 | Emotion & psychology | ⭐⭐ 会共情话术但无真正理解 | Empathy scripts but no real understanding | 不能替代专业心理咨询,仅辅助 | Cannot replace professional counseling, only auxiliary |
| 个人决策 | Personal decisions | ⭐⭐⭐ 能给多角度分析,但不能替你负责 | Can give multi-angle analysis but can't take responsibility for you | 把它当『顾问团』,决定权在你 | Treat it as an 'advisory panel'; the decision is yours |
提示工程 10 个核心技巧(把 AI 用出 3 倍效果)
10 Core Prompting Techniques (Get 3× from AI)
| 技巧 | Technique | 做法 | How to do it | 反例→正例 | Wrong → Right |
|---|---|---|---|---|---|
| 1 给角色 | 1. Role | 告诉 AI 它是谁:『你是资深数据分析师』 | Tell it who to be: 'You're a senior data analyst' | 『写报告』→『你是十年经验的财务顾问,写一份…』 | 'write a report' → 'you're a 10-yr financial advisor, write...' |
| 2 给上下文 | 2. Context | 背景、目标、读者、限制一次给全 | Give background, goal, audience, constraints at once | 『翻译这句』→『这是给技术读者的论文摘要,正式体,译成…』 | 'translate this' → 'this is a paper abstract for tech readers, formal tone, translate into...' |
| 3 拆任务 | 3. Decompose | 大任务拆成小步骤,逐个完成 | Break big tasks into small steps, one by one | 『帮我做完整营销方案』→『先列目标,再列渠道,最后写预算』 | 'give me a full marketing plan' → 'first list goals, then channels, then budget' |
| 4 给例子 | 4. Examples | 少样本示例,让 AI 模仿格式 | Few-shot examples so AI mimics the format | 描述 10 遍不如给 2 个例子 | Describing 10 times < giving 2 examples |
| 5 要步骤 | 5. Chain-of-thought | 『请一步步思考并展示推理』 | 'Please think step by step and show reasoning' | 直接要答案 → 要过程,错误率大降 | Direct answer → ask for the process, error rate drops a lot |
| 6 定格式 | 6. Format | 指定输出:表格/列表/JSON/Markdown | Specify output: table/list/JSON/Markdown | 『整理这些』→『用三列表格:名称|价格|库存』 | 'organize these' → 'in a 3-col table: name|price|stock' |
| 7 要边界 | 7. Boundaries | 告诉它『不知道就说不知道』 | Tell it 'say you don't know if you don't know' | 放任它编 → 明确要求标注不确定 | Letting it fabricate → explicitly require marking uncertainty |
| 8 迭代 | 8. Iterate | 别指望一次完美,追问和修改 | Don't expect perfection in one shot; follow up and revise | 一次不满意就放弃 → 基于上一轮追问 | Giving up after one unsatisfying answer → follow up on the last round |
| 9 用变量 | 9. Variables | 把重复的部分做成模板,只换变量 | Make repeated parts into templates, only swap variables | 每次重写长提示 → 做模板填空 | Rewriting long prompts each time → make a fill-in template |
| 10 自我验证 | 10. Self-check | 『请检查上面的回答有没有错误并修正』 | 'Please check the answer above for errors and correct' | 直接采用 → 让它自审一遍 | Using it directly → have it self-review once |
幻觉检测与验证:8 种方法
Hallucination Detection & Verification: 8 Methods
| 方法 | Method | 怎么做 | How to do it | 适用场景 | When to use |
|---|---|---|---|---|---|
| 1 来源核对 | 1. Source check | 让它给出处,你逐个打开验证 | Have it cite sources, you open and verify each | 学术、数据、历史事实 | Academic, data, historical facts |
| 2 交叉验证 | 2. Cross-verify | 同一个问题问 2~3 个模型,看是否一致 | Ask 2~3 models the same question, see if consistent | 关键结论、数字 | Critical conclusions, numbers |
| 3 反事实测试 | 2. Counterfactual | 故意问一个不存在的事实,看它是否硬编 | Deliberately ask about a non-existent fact, see if it fabricates | 测试某个模型的可靠性 | Testing a model's reliability |
| 4 让它自证 | 4. Self-prove | 『请给出推理过程和数据来源』 | 'Please give the reasoning process and data sources' | 数学、逻辑、因果 | Math, logic, causality |
| 5 缩小范围 | 5. Narrow down | 模糊问题改具体,减少它发挥空间 | Turn vague questions specific, reduce its free play | 开放性大问题 | Open-ended big questions |
| 6 要置信度 | 6. Confidence | 『你对这个答案有多确定?0-100 分』 | 'How confident are you? 0-100' | 需要判断可信度 | When you need to judge credibility |
| 7 反向提问 | 7. Reverse ask | 『这个说法有什么反例或争议?』 | 'What counterexamples or controversies exist for this claim?' | 单一观点、争议话题 | Single viewpoints, controversial topics |
| 8 事实/观点分离 | 8. Fact/Opinion split | 让它标注哪些是事实、哪些是推断 | Have it label which are facts, which are inferences | 综合分析、预测 | Comprehensive analysis, predictions |
数据安全与隐私:给 AI 喂什么、不喂什么
Data Security & Privacy: What to Feed AI, What Not to
| 数据类型 | Data type | 能给 AI 吗? | Can it be given to AI? | 原因 | Reason |
|---|---|---|---|---|---|
| 公开信息 | Public info | ✅ 可以 | Yes | 本身已公开,无风险 | Already public, no risk |
| 匿名化数据 | Anonymized data | ✅ 可以 | Yes | 去标识后风险低 | Low risk after de-identification |
| 公司机密 | Company secrets | ❌ 不可以 | No | 会泄露到训练/日志中 | Leaked into training/logs |
| 客户个人信息 | Customer PII | ❌ 不可以 | No | 违反隐私法规(GDPR/个保法) | Violates privacy laws (GDPR/PIPL) |
| 医疗/财务记录 | Medical/financial records | ❌ 不可以 | No | 高度敏感,需脱敏 | Highly sensitive, must be de-identified |
| 内部未发布内容 | Internal unreleased content | ❌ 不可以 | No | 可能被竞争对手间接获取 | May be indirectly obtained by competitors |
| 密码/密钥 | Passwords/keys | ❌ 绝对不 | Never | 任何情况下都不给 | Never, under any circumstance |
| 未公开合同 | Unpublished contracts | ❌ 不可以 | No | 商业机密 | Trade secrets |
- 免费版 vs 商用版:免费版通常用你的数据改进模型,商用版承诺不用于训练(但要看条款)Free vs commercial: free versions usually use your data to improve the model; commercial versions promise not to train on it (but read the terms)
- 本地部署最安全:数据不出内网,但成本高Local deployment is safest: data stays on the intranet, but costly
- 提示词也算数据:你写的长提示里如果含机密,同样会泄露Prompts are also data: if your long prompt contains secrets, they leak too
伦理与偏见:AI 不中立,它带着偏见出生
Ethics & Bias: AI Is Not Neutral — It's Born with Bias
AI 的偏见不是『坏』,而是『统计』:它从人类写的文本里学习,而人类文本本身充满偏见(性别、种族、地域、刻板印象)。所以 AI 会『继承』这些偏见,并在新场景里放大它们。这不是 AI 的错,是数据源的错,但后果由使用者承担。
AI bias is not 'bad', it's 'statistical': it learns from text written by humans, and human text is full of bias (gender, race, region, stereotypes). So AI 'inherits' these biases and amplifies them in new scenarios. This isn't AI's fault, it's the data source's fault, but the consequences are borne by the user.
| 偏见类型 | Bias type | 表现 | Manifestation | 如何缓解 | How to mitigate |
|---|---|---|---|---|---|
| 性别偏见 | Gender bias | 『工程师』常联想男性,『护士』常联想女性 | 'Engineer' often associated with male, 'nurse' with female | 要求多角度、检查代词使用 | Require multiple angles, check pronoun usage |
| 地域偏见 | Regional bias | 对发达地区描述更正面,欠发达地区刻板 | More positive descriptions of developed regions, stereotypes for underdeveloped | 对比多地表述、查证来源 | Compare multi-region descriptions, verify sources |
| 确认偏见 | Confirmation bias | 倾向给『你想听到的』答案 | Tends to give 'the answer you want to hear' | 让它先给反面论据 | Have it give counterarguments first |
| 近因偏见 | Recency bias | 过度依赖训练数据里的近期事件 | Over-relies on recent events in training data | 标注时间、要求考虑长期趋势 | Mark time, require considering long-term trends |
| 精英偏见 | Elitism bias | 偏好名校/知名机构,忽视草根方案 | Prefers elite schools/famous institutions, ignores grassroots solutions | 要求它列非主流方案 | Require it to list non-mainstream solutions |
| 语言偏见 | Language bias | 英语内容质量明显高于其他语言 | English content quality clearly higher than other languages | 重要内容多语言交叉验证 | Cross-verify important content in multiple languages |
多模态 AI:文字之外的新能力
Multimodal AI: New Capabilities Beyond Text
| 能力 | Capability | 2025 年成熟度 | 2025 maturity | 典型应用 | Typical uses |
|---|---|---|---|---|---|
| 图像理解 | Image understanding | ⭐⭐⭐⭐ 成熟 | Mature | 看图回答、OCR、图表解读 | Image Q&A, OCR, chart interpretation |
| 图像生成 | Image generation | ⭐⭐⭐⭐ 成熟 | Mature | 设计、插画、产品图 | Design, illustration, product images |
| 语音识别 | Speech recognition | ⭐⭐⭐⭐⭐ 非常成熟 | Very mature | 会议转写、字幕 | Meeting transcription, subtitles |
| 语音合成 | Speech synthesis | ⭐⭐⭐⭐ 成熟 | Mature | 配音、朗读、客服 | Dubbing, reading, customer service |
| 视频生成 | Video generation | ⭐⭐⭐ 快速成熟 | Rapidly maturing | 短视频、宣传片、演示 | Short videos, promos, demos |
| 代码执行 | Code execution | ⭐⭐⭐⭐ 成熟 | Mature | 跑代码、数据分析、自动化 | Running code, data analysis, automation |
| Agent 操作 | Agent operations | ⭐⭐⭐ 成长中 | Growing | 自动浏览网页、填表、多步任务 | Auto browsing, form filling, multi-step tasks |
| 实时对话 | Real-time conversation | ⭐⭐⭐⭐ 成熟 | Mature | 语音助手、陪练、面试模拟 | Voice assistant, sparring, interview simulation |
AI Agent:从工具到助手的跃迁
AI Agents: The Leap from Tool to Assistant
传统 AI 是你问它答。AI Agent 是你给目标,它自己拆解、规划、执行、检查、修正,像一个能动手的实习生。它能浏览网页、调用工具、写代码、运行程序、把结果交给你。
Traditional AI is you ask, it answers. An AI Agent is you give a goal, it decomposes, plans, executes, checks, and corrects by itself, like an intern who can actually work. It can browse the web, call tools, write code, run programs, and deliver results.
| Agent 能力 | Agent capability | 示例 | Example | 人该做什么 | What the human should do |
|---|---|---|---|---|---|
| 任务拆解 | Task decomposition | 帮我做行业报告,自动拆成搜资料、读、写、排 | 'Make an industry report' auto-decomposes into search, read, write, format | 定目标、定验收标准 | Set goals, set acceptance criteria |
| 工具调用 | Tool calling | 自动调搜索、计算器、API | Auto-calls search, calculator, APIs | 给它授权边界 | Give it authorization boundaries |
| 自我检查 | Self-check | 写完自查格式、数据、逻辑 | After writing, self-checks format, data, logic | 抽查关键部分 | Spot-check key parts |
| 记忆 | Memory | 记住之前对话的上下文和偏好 | Remembers prior conversation context and preferences | 定期校准它的记忆 | Periodically calibrate its memory |
| 多 Agent 协作 | Multi-Agent collaboration | 一个搜索、一个写作、一个校对 | One searches, one writes, one proofreads | 设计分工和流程 | Design division of labor and workflow |
AI 素养的 7 个水平:你在哪一级?
7 Levels of AI Literacy: Which Level Are You At?
| 级别 | Level | 特征 | Characteristics | 典型行为 | Typical behavior |
|---|---|---|---|---|---|
| L0 排斥 | L0 Aversion | 认为 AI 没用或有害,从不使用 | Thinks AI is useless or harmful, never uses it | 只用传统方法,拒绝一切 AI 工具 | Uses only traditional methods, rejects all AI tools |
| L1 尝鲜 | L1 Novice | 偶尔用,当高级搜索引擎 | Occasionally uses, as an advanced search engine | 问简单问题,不深究 | Asks simple questions, doesn't dig deeper |
| L2 熟练 | L2 Proficient | 日常使用,掌握基本提示技巧 | Daily use, masters basic prompting techniques | 写邮件、翻译、写文案都用 AI | Uses AI for emails, translation, copywriting |
| L3 方法 | L3 Method | 掌握提示工程,会验证、会组合 | Masters prompting, verifies, combines | 拆解任务、交叉验证、多模型对比 | Decomposes tasks, cross-verifies, compares models |
| L4 理解 | L4 Understanding | 懂原理、知边界、能判断幻觉 | Understands principles, knows boundaries, judges hallucinations | 知道何时该用、何时不该用、何时要怀疑 | Knows when to use, when not to, when to doubt |
| L5 创造 | L5 Creation | 用 AI 创造新价值,设计工作流 | Creates new value with AI, designs workflows | 搭建 Agent 流程、自动化、产品原型 | Builds Agent workflows, automation, product prototypes |
| L6 专家 | L6 Expert | 能教别人、能评估模型、能定标准 | Can teach others, evaluate models, set standards | 做 AI 培训、模型评测、企业 AI 战略 | Does AI training, model evaluation, enterprise AI strategy |
AI 素养自评:15 道题测测你自己
AI Literacy Self-Assessment: 15 Questions to Test Yourself
快速自评(每题 1 到 3 分,满分 45)
Quick Self-Assessment (1-3 points each, max 45)
| 题目 | Question | 1 分 | 1 pt | 3 分 | 3 pts |
|---|---|---|---|---|---|
| 我会用 AI 写邮件和文案 | I can use AI to write emails and copy | 会 | Can | 熟练且快 | Skilled and fast |
| 我会给 AI 完整的背景上下文 | I give AI full background context | 偶尔 | Occasionally | 每次都给 | Every time |
| 我会让 AI 一步步展示推理 | I have AI show step-by-step reasoning | 不会 | Never | 复杂题都会 | Always for complex |
| 我会对 AI 的输出做交叉验证 | I cross-verify AI output | 不验证 | Don't | 关键信息都验证 | Verify all critical info |
| 我能识别 AI 的幻觉(编造) | I can identify AI hallucinations | 很少 | Rarely | 基本能 | Basically always |
| 我了解大模型的基本原理 | I understand LLM basics | 不了解 | Don't | 能解释 | Can explain |
| 我遵守数据安全红线(不喂机密) | I follow data security red lines | 有时违反 | Sometimes | 从不违反 | Never violate |
| 我会用多模态(图像/语音/视频) | I use multimodal (image/audio/video) | 没用过 | Never | 经常用 | Frequently |
| 我理解 AI 的偏见并会应对 | I understand AI bias and cope | 不理解 | Don't | 能识别并规避 | Can identify and avoid |
| 我会设计 AI 工作流或 Agent 流程 | I design AI workflows or Agent flows | 不会 | Never | 能搭建 | Can build |
| 我会评估不同模型的优劣 | I evaluate different models' pros and cons | 不评估 | Don't | 能对比 | Can compare |
| 我能教别人用 AI | I can teach others to use AI | 不能 | Can't | 能系统教 | Can teach systematically |
| 我知道 AI 能做和不能做的边界 | I know AI's do and don't boundaries | 模糊 | Vague | 清晰 | Clear |
| 我警惕过度依赖 AI,保持自主思考 | I guard against over-reliance | 有时依赖 | Sometimes depend | 始终自主 | Always independent |
| 我跟进 AI 新技术并应用到工作 | I follow AI new tech and apply to work | 不跟进 | Don't | 持续跟进 | Continuously follow |
自测:你的 AI 素养水平
Quiz: Your AI Literacy Level
20 项 AI 素养自检清单
20-Item AI Literacy Self-Check List
- ✓我会给 AI 完整的角色和背景I give AI full role and context
- ✓我会把大任务拆成小步骤喂给 AII break big tasks into small steps for AI
- ✓复杂推理我会让 AI 一步步展示过程For complex reasoning I have AI show step-by-step
- ✓我会对关键信息做交叉验证I cross-verify critical info
- ✓我会检查 AI 的引用是否真实存在I check AI citations actually exist
- ✓我知道 AI 的训练截止日期,警惕过时信息I know AI's training cutoff, wary of outdated info
- ✓我会让 AI 标注不确定和争议的地方I have AI mark uncertainty and controversy
- ✓我不把机密和个人信息喂给 AII don't feed secrets/PII to AI
- ✓我了解 token、上下文、温度的基本含义I understand basic meaning of token/context/temperature
- ✓我会根据任务调温度和格式I adjust temperature and format by task
- ✓我会识别 AI 的性别和地域偏见I recognize AI's gender/regional bias
- ✓我会让 AI 给反面论据,避免确认偏见I have AI give counterarguments to avoid confirmation bias
- ✓我用多模态(图像/语音)辅助学习I use multimodal (image/audio) to aid learning
- ✓我会用 Agent 做多步任务并设检查点I use Agents for multi-step tasks with checkpoints
- ✓关键节点我会人工复核 Agent 的输出I manually review Agent output at key nodes
- ✓我警惕过度依赖,保持独立判断I guard against over-reliance, keep independent judgment
- ✓我会对比 2 个以上模型再下结论I compare 2+ models before concluding
- ✓我能向别人解释 AI 的基本原理I can explain AI basics to others
- ✓我跟进 AI 新技术并应用到工作I follow AI new tech and apply to work
- ✓我建立了自己的 AI 使用红线清单I built my own AI usage red-line list
15 条高频疑问(AI 素养)
15 High-Frequency Questions (AI Literacy)
参考来源(AI 素养)
References (AI Literacy)
- World Economic Forum - Future of Jobs Report 2025 (AI literacy as core skill): weforum.org
- Microsoft - Work Trend Index 2024 (76% of workers use generative AI): microsoft.com/worklab
- European Union - AI Act 2024 (AI literacy in compulsory education): artificialintelligenceact.eu
- OpenAI - GPT-4 Technical Report (model capabilities and limitations): openai.com
- Anthropic - Claude Model Card (bias and safety evaluation): anthropic.com
- Google - Gemini Technical Report (multimodal capabilities): deepmind.google
- Stanford HAI - AI Index Report 2025 (AI adoption statistics): aiindex.stanford.edu
- Pew Research Center - AI and the Public 2024 (public perception and trust): pewresearch.org
- OECD - AI and Future of Work 2025 (skills gap analysis): oecd.org
- Deloitte - State of Generative AI in Enterprise 2025: deloitte.com
- MIT Sloan - AI Literacy Framework 2024: mitsloan.mit.edu
- Brookings Institution - AI and Democracy 2024 (bias and misinformation): brookings.edu
- NIST - AI Risk Management Framework 2024: nist.gov
- Common Sense Media - AI and Children 2025: commonsensemedia.org
- IEEE - Ethics of AI 2024 (bias mitigation guidelines): ieee.org
- Accenture - The State of AI 2025 (enterprise adoption): accenture.com
- Harvard Business Review - AI Augmentation 2024: hbr.org
- Pew Research - Generative AI and the Public 2025: pewresearch.org
- McKinsey - The Economic Potential of Generative AI 2024: mckinsey.com
- World Bank - AI and Development 2025 (digital divide): worldbank.org
50 条分场景提示词库(直接复制用)
50 Scenario-Based Prompts (Copy & Use)
一、写作类(8 条)
I. Writing (8)
| 场景 | Scene | 提示词模板 | Prompt |
|---|---|---|---|
| 邮件 | 你是职场沟通专家。根据要点写邮件:语气[正式/友好],收件人[谁],要点[列表],长度[字数],结尾[行动项]。先出初稿我确认再润色。 | You're a workplace comm expert. Write an email from these points: tone [formal/friendly], recipient [who], points [list], length [words], ending [action]. Draft first, I confirm before polishing. | |
| 摘要 | Summary | 你是学术编辑。把这段[领域]内容压成[字数]摘要,保留[关键信息],去掉[冗余],用[目标读者]能懂的语言。 | You're an academic editor. Compress this [field] content into a [words] summary, keep [key info], drop [redundancy], in [readers'] language. |
| 改写 | Rewrite | 把这段改写为[正式/口语/营销]风格,面向[读者],突出[核心信息],[字数]内,保留原意换表达。 | Rewrite this as [formal/casual/marketing], for [readers], highlight [core info], within [words], keep meaning, change expression. |
| 大纲 | Outline | 你是内容策划。为[主题]写[类型]大纲,分[几部分],每部分[要点数]个要点,逻辑[递进/并列/对比]。 | You're a content planner. Write a [type] outline for [topic], in [parts], each [count] points, logic [progressive/parallel/contrast]. |
| 标题 | Title | 给[内容]起[数量]个标题,风格[吸睛/专业/悬念],每个[字数]内,突出[卖点],避免标题党。 | Give [content] [count] titles, style [catchy/pro/suspense], each within [words], highlight [sell point], avoid clickbait. |
| 故事 | Story | 用[风格]讲[主题]的[短]故事,主角[谁],冲突[什么],结局[怎样],[字数]内,有[情感点]。 | Tell a [style] short story on [topic], protagonist [who], conflict [what], ending [how], within [words], with [emotional beat]. |
| 文案 | Copy | 为[产品]写[渠道]文案,卖点[列表],目标[人群],痛点[什么],号召[行动],语气[调性],分[几段]。 | Write [channel] copy for [product], selling points [list], target [audience], pain [what], CTA [action], tone [tone], in [parts]. |
| 问答 | Q&A | 把[主题]整理成[数量]个高频问答,问题是[用户]会问的,答案[简洁/详细],每个[字数]内,附[来源]。 | Organize [topic] into [count] high-freq Q&A, questions [users] ask, answers [brief/detailed], each within [words], with [sources]. |
二、学习类(8 条)
II. Learning (8)
| 场景 | Scene | 提示词模板 | Prompt |
|---|---|---|---|
| 苏格拉底 | Socratic | 你是苏格拉底式老师。别直接告诉[主题]答案,用提问引导我想出来。从[我的水平]开始,每次只问一个问题,等我回答再推进。 | You're a Socratic teacher. Don't give [topic] answer directly; use questions to lead me. Start from [my level], one question at a time, wait for my answer. |
| 费曼 | Feynman | 用[比喻]给我解释[概念],假设我是[背景],先[核心]再[细节],最后[例子]收尾。我没懂就换[另一个比喻]。 | Explain [concept] with [metaphor], assume I'm [background], [core] then [details], end with [example]. If I don't get it, switch [another metaphor]. |
| 出题 | Quiz | 基于[材料/主题]出[数量]道[选择/判断/简答]题,难度[分布],每题附答案+解析,最后给总分+薄弱点分析。 | Based on [material/topic] create [count] [mcq/tf/short] questions, difficulty [dist], each with answer + explanation, finally total + weakness analysis. |
| 计划 | Plan | 制定学[技能]的[天数]天计划,我[每天可用时间],起点[水平],目标[达到什么],按周分解,每天给具体任务。 | Make a [days]-day plan for [skill], I have [daily time], start [level], goal [achieve], break by week, each day a concrete task. |
| 拆解 | Breakdown | 把[大主题]拆成[层级]知识树,第一层[几大块],每块[几小点],每小点[一句话定义],帮我理清结构。 | Break [big topic] into a [level]-tier knowledge tree, first tier [blocks], each [subpoints], each [one-sentence def], help me see structure. |
| 连接 | Connect | 解释[概念A]和[概念B]的关系:相同点、不同点、一个怎么影响另一个,用[例子]说明,最后[一句话总结]。 | Explain relationship of [A] and [B]: similarities, differences, how one affects the other, with [example], finally [one-line summary]. |
| 检验 | Check | 我写了[主题]的理解:[内容]。指出哪里准确、哪里偏差、哪里缺失,按重要度排序,给修正建议。 | I wrote my understanding of [topic]: [content]. Point out accurate/off/missing, order by importance, give correction advice. |
| 迁移 | Transfer | 我刚学[概念/方法],用[新场景]帮我应用,给步骤,预测可能踩的坑,告诉我怎么避免。 | I just learned [concept/method], apply to [new scenario] for me, give steps, predict pitfalls, tell me how to avoid. |
三、工作类(10 条)
III. Work (10)
| 场景 | Scene | 提示词模板 | Prompt |
|---|---|---|---|
| PPT | PPT | 为[主题]的[时长]演示写 PPT 大纲,[页数]页,每页[标题+3要点+讲稿1句],逻辑[故事线],风格[专业/活泼],末页[行动号召]。 | Write PPT outline for [topic] [duration] presentation, [slides] pages, each [title+3 points+1-line script], logic [storyline], style [pro/lively], last [CTA]. |
| 会议 | Meeting | 根据[纪要要点]整理:决议[列表]、待办[负责人+截止]、风险[列表]、下次议程[预排],用[格式]输出。 | Organize from [minutes]: decisions [list], to-dos [owner+deadline], risks [list], next agenda [pre-set], output in [format]. |
| 报告 | Report | 写[主题]的[类型]报告,结构[背景-发现-分析-建议],每部分[字数],数据[来源],结论[明确/谨慎],给[读者]看。 | Write a [type] report on [topic], structure [bg-find-analysis-reco], each [words], data [source], conclusion [clear/cautious], for [readers]. |
| 分析 | Analysis | 用[SWOT/5W/PEST]分析[对象],每维度[几点],基于[事实/数据],最后给核心结论+2-3 个行动建议。 | Analyze [object] with [SWOT/5W/PEST], each dimension [points], based on [facts/data], finally core conclusion + 2-3 action advice. |
| 邮件回复 | Email reply | 这是[对方邮件]。帮我回复:立场[同意/拒绝/协商],语气[专业/温和],要点[列表],长度[字数],结尾[下一步]。 | Here's [other's email]. Reply: stance [agree/refuse/negotiate], tone [pro/gentle], points [list], length [words], ending [next step]. |
| 谈判 | Negotiation | 我[角色],对方[角色],目标[什么],底线[什么],对方诉求[什么]。给策略+话术,预测对方反应和我的应对。 | I'm [role], other [role], goal [what], bottom line [what], their demands [what]. Give strategy + scripts, predict their reaction + my response. |
| 提案 | Proposal | 为[方案]写提案:问题[现状]、方案[做法]、收益[量化]、成本[预算]、风险[预案]、时间表[里程碑],面向[决策者],突出[ROI]。 | Write proposal for [plan]: problem [now], plan [how], benefits [quant], cost [budget], risks [contingency], timeline [milestones], for [decision-maker], highlight [ROI]. |
| 复盘 | Review | 复盘[项目/事件]:目标[当初]、结果[实际]、差距[原因]、做对的[亮点]、做错的[教训]、下次[改进],用[5W/AAR]格式。 | Review [project/event]: goal [orig], result [actual], gap [cause], right [highlights], wrong [lessons], next [improve], in [5W/AAR] format. |
| 招聘 | Hiring | 为[岗位]写招聘描述:职责[列表]、要求[硬/软]、加分[项]、薪资[范围]、吸引力[点],语气[调性],避免歧视性表述。 | Write JD for [role]: duties [list], requirements [hard/soft], plus [items], salary [range], attraction [points], tone [tone], avoid discriminatory language. |
| 客户 | Customer | 客户[投诉/问题]是[内容]。回复:先共情,再解释,给解决方案+补偿,结尾挽回,语气专业,[字数]内。 | Customer [complaint/issue] is [content]. Reply: empathize, explain, solution + compensation, retain ending, professional, within [words]. |
四、生活类(8 条)
IV. Life (8)
| 场景 | Scene | 提示词模板 | Prompt |
|---|---|---|---|
| 旅行 | Travel | 规划[天数]天[目的地]旅行:预算[多少],同行[谁],兴趣[类型],每天给行程+预算+贴士,含交通+住宿+餐饮,避开[不想去]。 | Plan [days]-day [destination] trip: budget [amount], companions [who], interests [type], each day itinerary + budget + tips, incl. transport + lodging + food, avoid [don't want]. |
| 健身 | Fitness | 制定[天数]天[减脂/增肌]计划,我[基础体能],每周[几次],每次[多久],给具体动作+组数+强度,附饮食建议。 | Make a [days]-day [cut/bulk] plan, my [base fitness], [times]/week, each [duration], specific moves + sets + intensity, with diet advice. |
| 理财 | Finance | 分析我的[收支情况],给预算分配建议,紧急备用金该留[多少],负债怎么还最优,投资怎么起步,基于[风险偏好],不承诺收益。 | Analyze my [income/expense], budget allocation advice, emergency fund [amount], debt repayment optimal, investment start, based on [risk preference], no income promises. |
| 亲子 | Parenting | 孩子[年龄],遇到[问题],给我沟通话术+引导方法,避免指责/包办,培养自主/习惯,给具体例子。 | Child [age], faces [issue], give communication scripts + guidance, avoid blame/doing-for-them, cultivate autonomy/habit, with examples. |
| 社交 | Social | 帮我准备[场合]的开场+话题,对象[谁],我想表达/了解什么,给 3 个话题+接话技巧,避免尴尬话题。 | Prepare [opening/topics] for [occasion], other is [who], I want to express/learn, 3 topics + follow-up tips, avoid awkward topics. |
| 烹饪 | Cooking | 用[现有食材]设计[几道]菜,口味[偏好],[时间]内完成,给步骤+火候+调味,适合[几人],附营养。 | Design [dishes] with [ingredients], taste [preference], within [time], steps + heat + seasoning, for [people], with nutrition. |
| 搬家 | Moving | 规划[城市A到B]搬家:预算[多少],物品量,时间,给清单+步骤+时间线+成本拆解,含打包+运输+入住。 | Plan [city A to B] moving: budget [amount], item volume, time, checklist + steps + timeline + cost breakdown, incl. packing + transport + settling. |
| 健康 | Health | 我[症状/困扰],基于一般知识给可能方向+自查建议,提醒我该看医生,不诊断不开药,给日常调理建议。 | I have [symptoms], based on general knowledge give possible directions + self-check, remind me to see a doctor, no diagnosis/prescription, daily care advice. |
五、创作类(6 条)
V. Creative (6)
| 场景 | Scene | 提示词模板 | Prompt |
|---|---|---|---|
| 脑暴 | Brainstorm | 为[主题]脑暴[数量]个创意/方案/名字,风格多样,从不同角度,每个给一句话说明,最后挑最可行的 3 个并说明理由。 | Brainstorm [count] ideas/plans/names for [topic], diverse style, different angles, each with one-line desc, finally pick top 3 feasible with reasons. |
| 结构 | Structure | 帮我把散乱的点整理成[文章/演讲/视频],逻辑主线,起承转合安排,每部分字数/时长,突出核心。 | Organize scattered points into [article/speech/video], logic mainline, setup-development-turn-conclusion, each part words/duration, highlight core. |
| 润色 | Polish | 润色这段文字:提升流畅度/感染力/专业度,保留核心意思,去掉口语/冗余,调整节奏,控制在[字数],语气[调性]。 | Polish this text: improve flow/appeal/professionalism, keep core meaning, remove colloquial/redundant, adjust rhythm, within [words], tone [tone]. |
| 视频脚本 | Video script | 为[时长]视频写脚本:钩子开头 5 秒、主体 3 段、结尾号召,每段给画面+台词+时长,节奏快/慢,风格[调性]。 | Script for [duration] video: hook first 5s, body 3 segments, ending CTA, each gives visual + line + duration, pace fast/slow, style [tone]. |
| 歌曲 | Song | 写一首[风格]歌,主题[什么],[几段],押韵,情绪变化,每段[字数],副歌要抓耳,给意象。 | Write a [style] song, theme [what], [verses], rhyme, emotion shift, each [words], catchy chorus, imagery. |
| 视觉 | Visual | 描述[场景]的视觉风格:色调、构图、光线、主体、氛围,给 3 个变体,用于 AI 绘图提示。 | Describe [scene] visual style: color, composition, light, subject, mood, 3 variants, for AI image prompt. |
六、研究类(6 条)
VI. Research (6)
| 场景 | Scene | 提示词模板 | Prompt |
|---|---|---|---|
| 综述 | Review | 综述[主题]:[年份]以来的主要进展,分几方向,每方向给代表工作+结论,最后给趋势+空白,标注来源。 | Review [topic]: main progress since [year], in directions, each gives representative work + conclusion, finally trends + gaps, mark sources. |
| 对比 | Compare | 对比 A 和 B:维度[列表],每维度谁优谁劣+原因,基于事实,最后给适用场景建议,用表格。 | Compare A and B: dimensions [list], each who's better + why, based on facts, finally use-case advice, in table. |
| 溯源 | Trace | 追溯[概念/方法]的起源:谁提出何时,发展脉络,关键节点,当前状态,标注来源,区分事实和推测。 | Trace [concept/method] origin: who proposed when, development, key nodes, current state, mark sources, distinguish fact and speculation. |
| 批判 | Critique | 批判这个观点/研究:论证是否成立,证据是否充分,逻辑有无漏洞,反例,最后给修正版,标注来源。 | Critique this view/research: is argument sound, evidence sufficient, logic gaps, counterexamples, finally revised version, mark sources. |
| 方法 | Method | 解释[研究方法]:原理、步骤、适用、局限、常见错误,用例子,给使用建议,标注来源。 | Explain [research method]: principle, steps, applicable, limitations, common errors, with examples, usage advice, mark sources. |
| 数据 | Data | 分析[数据]:关键指标、趋势、异常、可能原因,给可视化建议,标注数据来源,区分相关和因果。 | Analyze [data]: key metrics, trends, anomalies, possible causes, visualization advice, mark data source, distinguish correlation and causation. |
30 个 AI 素养学习资源(真实可查)
30 AI Literacy Learning Resources (Verifiable)
25 个行业:AI 素养的落地差异
25 Industries: How AI Literacy Differs in Practice
| 行业 | Industry | 高优先 AI 能力 | High-priority AI skills | 最大风险 | Biggest risk | 入门重点 | Entry focus |
|---|---|---|---|---|---|---|---|
| 教育 | Education | 用 AI 因材施教、出题、批改 | AI for personalized teaching, quiz, grading | AI 代学生作业,能力退化 | AI doing homework for students, ability atrophies | 提示工程 + 验证习惯 | Prompting + verification habit |
| 医疗 | Healthcare | AI 查文献、辅助诊断 | AI for literature, assisted diagnosis | AI 误诊、责任不清 | AI misdiagnosis, unclear liability | 幻觉检测 + 数据隐私 | Hallucination detection + data privacy |
| 金融 | Finance | AI 分析、风控、投研 | AI for analysis, risk, research | AI 幻觉导致决策错 | AI hallucination causing wrong decisions | 批判验证 + 合规 | Critical verification + compliance |
| 法律 | Legal | AI 查判例、起草文书 | AI for case law, drafting docs | AI 引用不存在的判例 | AI citing non-existent cases | 引用验证 + 保密 | Citation verification + confidentiality |
| 编程 | Programming | AI 写代码、debug、review | AI for coding, debug, review | 盲信 AI 代码有 bug | Blindly trusting buggy AI code | 代码验证 + 安全 | Code verification + security |
| 设计 | Design | AI 出图、灵感、排版 | AI for imagery, inspiration, layout | 过度依赖,创意同质化 | Over-reliance, homogeneous creativity | 多模态 + 审美判断 | Multimodal + aesthetic judgment |
| 营销 | Marketing | AI 写文案、投放、分析 | AI for copy, ads, analysis | AI 文案千篇一律 | AI copy all the same | 创意 + 数据 | Creativity + data |
| 科研 | Research | AI 查文献、分析数据 | AI for literature, data analysis | AI 编造引用/数据 | AI fabricating citations/data | 溯源 + 批判 | Tracing + critique |
| 制造 | Manufacturing | AI 质检、预测维护 | AI for QC, predictive maintenance | AI 误判导致事故 | AI misjudgment causing accidents | 数据质量 + 安全 | Data quality + safety |
| 农业 | Agriculture | AI 遥感、预测产量 | AI for remote sensing, yield prediction | 数据不准导致损失 | Inaccurate data causing losses | 数据 + 本地化 | Data + localization |
| 物流 | Logistics | AI 路径优化、需求预测 | AI for routing, demand prediction | AI 过度优化导致异常 | AI over-optimization causing anomalies | 优化 + 容错 | Optimization + fault tolerance |
| 媒体 | Media | AI 写稿、配音、生成视频 | AI for writing, voice, video | AI 假新闻、版权 | AI fake news, copyright | 事实核查 + 版权 | Fact-checking + copyright |
| 政府 | Government | AI 决策支持、公共服务 | AI for decision support, public service | AI 偏见影响公平 | AI bias affecting fairness | 伦理 + 透明 | Ethics + transparency |
| 零售 | Retail | AI 推荐、库存、客服 | AI for recommendation, inventory, CS | AI 推荐茧房 | AI recommendation bubble | 数据 + 个性化 | Data + personalization |
| 房地产 | Real estate | AI 估值、带看、营销 | AI for valuation, tour, marketing | AI 估值不准 | Inaccurate AI valuation | 数据 + 本地 | Data + local |
| 旅游 | Tourism | AI 行程、客服、个性化 | AI for itinerary, CS, personalization | AI 行程不靠谱 | Unreliable AI itineraries | 多模态 + 本地 | Multimodal + local |
| 餐饮 | Food | AI 配方、供应链、营销 | AI for recipes, supply chain, marketing | AI 配方不安全 | Unsafe AI recipes | 数据 + 安全 | Data + safety |
| 游戏 | Gaming | AI 生成内容、NPC、平衡 | AI for content gen, NPC, balance | AI 内容质量差 | Poor AI content quality | 多模态 + 设计 | Multimodal + design |
| 音乐 | Music | AI 作曲、混音、翻唱 | AI for composition, mixing, cover | AI 版权纠纷 | AI copyright disputes | 多模态 + 版权 | Multimodal + copyright |
| 体育 | Sports | AI 训练分析、伤病预测 | AI for training analysis, injury prediction | AI 误判伤运动员 | AI misjudgment harming athletes | 数据 + 谨慎 | Data + caution |
| 能源 | Energy | AI 预测、优化、监控 | AI for prediction, optimization, monitoring | AI 误判导致安全事故 | AI misjudgment causing safety incidents | 数据 + 安全 | Data + safety |
| 汽车 | Automotive | AI 自动驾驶、制造 | AI for autonomous driving, manufacturing | AI 误判导致事故 | AI misjudgment causing accidents | 安全 + 伦理 | Safety + ethics |
| 建筑 | Architecture | AI 设计、BIM、估量 | AI for design, BIM, estimation | AI 设计不安全 | Unsafe AI designs | 多模态 + 规范 | Multimodal + codes |
| 保险 | Insurance | AI 理赔、核保、反欺诈 | AI for claims, underwriting, fraud | AI 偏见拒绝赔付 | AI bias denying claims | 数据 + 公平 | Data + fairness |
| 电信 | Telecom | AI 网络优化、客服 | AI for network optimization, CS | AI 误判导致服务中断 | AI misjudgment causing outages | 数据 + 容错 | Data + fault tolerance |
40 个 AI 素养自检项(逐项打分)
40 AI Literacy Self-Check Items (Score Each)
| 维度 | Dimension | 自检项 | Self-check item | 0 分 | 0 | 2 分 | 2 |
|---|---|---|---|---|---|---|---|
| 理解 | Understanding | 我能解释 LLM 工作原理 | I can explain LLM how it works | 完全不能 | Cannot at all | 能向非技术人讲清 | Can explain to non-technical |
| 理解 | Understanding | 我理解 token/上下文/温度的含义 | I understand token/context/temperature | 不知道 | Don't know | 能说明影响 | Can describe their effects |
| 理解 | Understanding | 我知道 AI 的『记忆』是假的 | I know AI's 'memory' is fake | 以为是真记忆 | Thinks it's real memory | 知道是检索/上下文 | Knows it's retrieval/context |
| 理解 | Understanding | 我能说出 AI 不会做什么 | I can state what AI can't do | 说不出 | Can't | 能列 5 项以上 | Can list 5+ items |
| 提示 | Prompting | 我会给角色+背景+任务+格式 | I give role+context+task+format | 只打字 | Just types | 提示结构完整 | Complete prompt structure |
| 提示 | Prompting | 我会用 few-shot 给例子 | I use few-shot examples | 不会 | Don't | 熟练用例子 | Skillful with examples |
| 提示 | Prompting | 我会让 AI 分步思考 | I make AI think step-by-step | 不会 | Don't | 用 CoT 提升质量 | Use CoT to improve quality |
| 提示 | Prompting | 我有自己的提示模板库 | I have my own prompt library | 没有 | None | 有 20+ 模板 | Have 20+ templates |
| 验证 | Verification | 我会交叉验证关键信息 | I cross-verify key info | 直接相信 | Believe directly | 关键都验证 | Verify all critical |
| 验证 | Verification | 我会打开 AI 给的引用 | I open AI's citations | 从不 | Never | 每次验证 | Verify every time |
| 验证 | Verification | 我能识别 AI 的自信幻觉 | I can spot confident hallucinations | 识别不了 | Can't | 能识别多数 | Can spot most |
| 验证 | Verification | 我会用反事实测试模型 | I test models with counterfactuals | 不会 | Don't | 会设计测试 | Can design tests |
| 验证 | Verification | 我区分事实/推断/观点 | I separate fact/inference/opinion | 不区分 | Don't | 清晰区分 | Clearly separate |
| 验证 | Verification | 我会多模型对比 | I compare multiple models | 只用一个 | Only one | 关键问题多对比 | Compare for key questions |
| 隐私 | Privacy | 我知道免费版会训练数据 | I know free versions train on data | 不知道 | Don't know | 清楚区分 | Clearly distinguish |
| 隐私 | Privacy | 我不把机密喂给 AI | I don't feed AI secrets | 有时喂 | Sometimes feed | 从不喂 | Never feed |
| 隐私 | Privacy | 我会匿名化再输入 | I anonymize before input | 不会 | Don't | 习惯匿名化 | Habitual anonymization |
| 隐私 | Privacy | 我理解数据流向 | I understand data flow | 不理解 | Don't understand | 能说出流向 | Can describe the flow |
| 隐私 | Privacy | 我会用本地/商用处理敏感 | I use local/commercial for sensitive | 不会 | Don't | 按敏感度选 | Choose by sensitivity |
| 伦理 | Ethics | 我了解 AI 的常见偏见 | I know AI's common biases | 不了解 | Don't know | 能举例 | Can give examples |
| 伦理 | Ethics | 我会用反例测试偏见 | I test bias with counterexamples | 不会 | Don't | 会多角度测试 | Multi-angle testing |
| 伦理 | Ethics | 我知道 AI 内容的版权归属 | I know AI content copyright | 不清楚 | Unclear | 清楚规则 | Clear on rules |
| 伦理 | Ethics | 我理解深度伪装的识别 | I understand deepfake recognition | 不懂 | Don't | 能识别 | Can recognize |
| 伦理 | Ethics | 我遵守 AI 使用的红线 | I follow AI usage red lines | 常违反 | Often violate | 从不违反 | Never violate |
| 应用 | Application | 我能用 AI 完成完整任务 | I can use AI for complete tasks | 只会单点 | Single points | 端到端完成 | End-to-end |
| 应用 | Application | 我会用多模态 AI | I use multimodal AI | 只用文字 | Text only | 会用图/音/视频 | Use image/audio/video |
| 应用 | Application | 我理解 AI Agent 的能力 | I understand AI Agent capabilities | 不了解 | Don't know | 能设计简单 Agent | Can design simple Agent |
| 应用 | Application | 我能搭建 RAG 检索 | I can build RAG retrieval | 不会 | Don't | 理解原理 | Understand the principle |
| 应用 | Application | 我会用 AI 提效日常工作 | I use AI for daily work efficiency | 不常用 | Rarely | 天天用 | Daily use |
| 应用 | Application | 我能教别人用 AI | I can teach others to use AI | 不会 | Don't | 能带人入门 | Can onboard people |
| 批判 | Critical | 我对 AI 输出保持怀疑 | I stay skeptical of AI output | 完全相信 | Fully believe | 默认怀疑 | Default skeptical |
| 批判 | Critical | 我会追问 AI 的依据 | I ask AI for its basis | 不追问 | Don't ask | 每次追问 | Ask every time |
| 批判 | Critical | 我能指出 AI 论证的漏洞 | I can spot AI argument gaps | 不能 | Can't | 能找出多数 | Can find most |
| 批判 | Critical | 我不让 AI 替我做重大判断 | I don't let AI make major decisions | 有时让 | Sometimes | 从不让 | Never let |
| 批判 | Critical | 我持续跟进 AI 新进展 | I keep up with AI progress | 不跟进 | Don't | 定期跟进 | Regularly follow |
| 批判 | Critical | 我能向别人解释 AI 的局限 | I can explain AI's limits to others | 不能 | Can't | 能清晰解释 | Can explain clearly |
| 批判 | Critical | 我有自己的 AI 使用原则 | I have my own AI usage principles | 没有 | None | 有明确原则 | Clear principles |
| 批判 | Critical | 我评估 AI 工具的优劣 | I evaluate AI tools' pros/cons | 不评估 | Don't | 会对比选型 | Compare and choose |
| 批判 | Critical | 我警惕 AI 的过度承诺 | I'm wary of AI over-promises | 轻信 | Believe easily | 保持警惕 | Stay wary |
| 批判 | Critical | 我能把 AI 纳入工作流 | I can integrate AI into workflows | 不会 | Don't | 熟练整合 | Skilled integration |
20 条 AI 素养金句(记住这 20 句就够用)
20 AI Literacy Golden Lines (20 Sentences Are Enough)
| 金句(中) | Golden line (ZH) | 金句(英) | Golden line (EN) | 解释 | Explanation |
|---|---|---|---|---|---|
| AI 是镜子,不是答案 | AI is a mirror, not the answer | AI 反映你的提问水平;问得差,答得差 | AI reflects your questioning level; bad question, bad answer | ||
| 先怀疑,再使用 | Doubt first, then use | 默认 AI 可能错,验证后才用 | Assume AI may be wrong, verify before use | ||
| 提示就是沟通 | Prompting is communication | 把 AI 当聪明但没上下文的同事 | Treat AI as a smart colleague without context | ||
| 上下文决定一切 | Context determines everything | 给足背景,AI 才能精准 | Give enough background for AI to be precise | ||
| AI 负责初稿,你负责灵魂 | AI drafts, you give the soul | 重复交给 AI,判断留给自己 | Give repetition to AI, keep judgment for yourself | ||
| 关键信息必验证 | Verify critical info | 数字、引用、结论,交叉核对 | Numbers, citations, conclusions — cross-check | ||
| 别喂机密 | Don't feed secrets | 敏感数据用本地/商用,别上云端 | Sensitive data local/commercial, not cloud | ||
| AI 没有立场,但有权力 | AI has no stance, but has power | 它能放大偏见,要主动纠正 | It can amplify bias, actively correct it | ||
| 持续验证,持续怀疑 | Keep verifying, keep doubting | AI 在变,你的警惕也要变 | AI changes, your vigilance must too | ||
| 别被效率绑架 | Don't be hijacked by efficiency | AI 省时间,但别省思考 | AI saves time, but don't skip thinking | ||
| 多模型对比 | Compare multiple models | 关键问题问 2-3 个模型 | Key questions to 2-3 models | ||
| AI 是杠杆,不是替代 | AI is leverage, not replacement | 放大你的能力,不替代你的判断 | Amplify your ability, not replace your judgment | ||
| 建立提示库 | Build a prompt library | 好提示值得积累和复用 | Good prompts worth accumulating and reusing | ||
| 教是最好的学 | Teaching is the best learning | 能教别人用 AI,才是真懂 | Can teach others = truly understand | ||
| 警惕过度承诺 | Watch over-promises | AI 说『保证』时,最该怀疑 | When AI says 'guarantee', doubt most | ||
| 数据质量决定质量 | Data quality determines quality | 垃圾进,垃圾出,AI 也一样 | Garbage in, garbage out, AI too | ||
| 保留人工检查点 | Keep human checkpoints | 关键操作(删/付/发)必须人确认 | Critical ops (delete/pay/send) need human confirm | ||
| 跟随,但别盲从 | Follow, but don't blindly | 了解新工具,但别追每个热点 | Know new tools, but don't chase every trend | ||
| AI 素养是长期主义 | AI literacy is long-termism | 不是一阵风,是底层能力 | Not a fad, a foundational ability | ||
| 用得对,比用得多重要 | Using right > using much | 质量优先,别贪多 | Quality first, don't be greedy |
60 个 AI 常见问题与解答(FAQ 扩充版)
60 Common AI Questions & Answers (Extended FAQ)
A. 使用类(20 条)
A. Usage (20)
| 问题 | Question | 解答 | Answer |
|---|---|---|---|
| AI 能帮我写论文吗? | Can AI write my paper? | 能帮你搭结构、出初稿、润色,但别让它写全文——同质化、查重不过、能力不增长。正确做法:你定观点,AI 出框架和素材,你填充血肉。 | It can build structure, first drafts, polish; but don't let it write the whole thing — homogenized, fails plagiarism, no growth. Right way: you set the thesis, AI gives framework + material, you fill the flesh. |
| AI 会取代老师吗? | Will AI replace teachers? | 短期不会。AI 是『助教』,能批改、答疑、个性化,但教育的核心是『点燃』和『榜样』,这是 AI 做不到的。未来是『会用 AI 的老师』取代『不会用的』。 | Short-term no. AI is a 'teaching assistant' — grades, answers, personalizes; but education's core is 'igniting' and 'role model', which AI can't do. Future: teachers who use AI replace those who don't. |
| AI 能帮我做研究吗? | Can AI help with research? | 能帮你查文献、总结、分析数据、找盲点,但别让它『保证』结论。关键发现一定要自己验证。AI 是加速器,不是替代者。 | It can search literature, summarize, analyze data, find blind spots; but don't let it 'guarantee' conclusions. Verify key findings yourself. AI accelerates, doesn't replace. |
| AI 能帮我学编程吗? | Can AI help me learn programming? | 非常能。让 AI 当结对编程伙伴:你写,AI 解释;你卡住,AI 给提示(不是答案);你写完,AI review。这是目前最有效的学编程方式之一。 | Very. Make AI a pair-programming partner: you write, AI explains; you're stuck, AI gives hints (not answers); you finish, AI reviews. One of the most effective ways to learn programming now. |
| AI 能帮我翻译吗? | Can AI translate? | 能,日常翻译很好,专业/文学翻译要人工校对。翻译后用『母语者』视角检查:地道吗?术语对吗?语气对吗? | Yes, daily translation is good; professional/literary needs human proofread. After translating, check from a 'native speaker' view: idiomatic? terms right? tone right? |
| AI 能帮我做 PPT 吗? | Can AI make PPT? | 能出大纲、每页要点、讲稿,但排版和设计要自己做。AI 出『内容』,你做『形式』。 | It can outline, per-slide points, scripts; but layout and design you do. AI gives 'content', you do 'form'. |
| AI 能帮我写代码吗? | Can AI write code? | 能,日常代码很靠谱,复杂逻辑要验证。关键:别盲信,跑测试、读逻辑、查边界。AI 写的代码也要 review。 | Yes, daily code is reliable; complex logic verify. Key: don't blindly trust, run tests, read logic, check edges. AI code also needs review. |
| AI 能帮我做数据分析吗? | Can AI do data analysis? | 能帮你写分析代码、解释结果、找趋势,但数据本身的质量你负责。垃圾数据进,垃圾结论出。 | It can write analysis code, explain results, find trends; but you're responsible for data quality. Garbage in, garbage out. |
| AI 能帮我做设计吗? | Can AI do design? | 能出图、灵感、配色,但『审美判断』和『品牌一致性』是你。AI 是素材库,你是设计师。 | It can generate images, inspiration, color schemes; but 'aesthetic judgment' and 'brand consistency' are yours. AI is a material library, you're the designer. |
| AI 能帮我写商业计划吗? | Can AI write a business plan? | 能出框架、市场规模估算、竞品分析,但核心假设(你的独特价值、数据验证)你自己。AI 不能替你做商业判断。 | It can give framework, market size estimates, competitor analysis; but core assumptions (your unique value, data validation) are yours. AI can't make business judgments for you. |
| AI 能帮我做投资分析吗? | Can AI do investment analysis? | 能帮你整理数据、算指标、找模式,但投资决策你自己。AI 不承诺收益、不预测未来。记住:历史数据不代表未来。 | It can organize data, compute metrics, find patterns; but investment decisions are yours. AI doesn't promise returns, doesn't predict the future. Remember: past data ≠ future. |
| AI 能帮我写简历吗? | Can AI write a resume? | 能帮你改写、优化关键词、突出亮点,但真实经历你自己。别编造——背调会露馅。 | It can help rewrite, optimize keywords, highlight; but real experience is yours. Don't fabricate — background checks expose it. |
| AI 能帮我做面试准备吗? | Can AI prep me for interviews? | 非常能。让 AI 当面试官,模拟提问、追问、压力测试,你回答,AI 给反馈。这是最有效的面试准备方式之一。 | Very. Make AI an interviewer, mock questions, follow-ups, stress tests, you answer, AI gives feedback. One of the most effective interview prep methods. |
| AI 能帮我做决策吗? | Can AI make decisions? | 能给正反对论、风险、盲点,帮你『看得更全』,但决策是你。AI 没有你的处境、价值观、责任。 | It can give pro/con, risks, blind spots, helping you 'see more completely'; but the decision is yours. AI has neither your circumstances, values, nor responsibility. |
| AI 能帮我管理时间吗? | Can AI manage my time? | 能帮你排日程、设优先级、估算工作量,但『执行』和『专注』是你。工具再强,执行还是靠你。 | It can schedule, set priorities, estimate workload; but 'execution' and 'focus' are yours. No matter how strong the tool, execution depends on you. |
| AI 能帮我学语言吗? | Can AI teach me a language? | 能当陪练:对话、纠错、解释、出题。但『坚持』和『输入量』是你。AI 是陪练,不是替练。 | It can be a sparring partner: conversation, correction, explanation, quiz. But 'persistence' and 'input volume' are yours. AI is a sparring partner, not a substitute trainer. |
| AI 能帮我写小说吗? | Can AI write a novel? | 能帮你脑暴、出结构、写片段,但『灵魂』和『连贯性』是你。纯 AI 小说同质化、没深度。人机协作最好。 | It can brainstorm, structure, write fragments; but 'soul' and 'coherence' are yours. Pure AI novels are homogenized, shallow. Human-AI collaboration is best. |
| AI 能帮我做视频吗? | Can AI make videos? | 能出脚本、分镜、配音、生成画面,但剪辑节奏和叙事是你。AI 是素材和草稿,你是导演。 | It can script, storyboard, voice, generate visuals; but editing rhythm and narrative are yours. AI is material and draft, you're the director. |
| AI 能帮我做音乐吗? | Can AI make music? | 能作曲、编曲、生成,但『情感』和『品味』是你。AI 是乐器,你是音乐人。 | It can compose, arrange, generate; but 'emotion' and 'taste' are yours. AI is the instrument, you're the musician. |
| AI 能帮我做数学题吗? | Can AI solve math problems? | 能解题、讲步骤,但别让它『直接给答案』。让 AI 当老师,引导你自己想出来,这才是学习。 | It can solve and explain steps; but don't let it 'give the answer directly'. Make AI a teacher, guiding you to figure it out yourself — that's learning. |
B. 验证类(15 条)
B. Verification (15)
| 问题 | Question | 解答 | Answer |
|---|---|---|---|
| 怎么验证 AI 说的数字? | How to verify AI's numbers? | 交叉验证:查原始来源、对比多个独立来源、看量级是否合理。关键数字(金额、日期、比例)一定要查。 | Cross-verify: check original source, compare multiple independent sources, check if magnitude is reasonable. Key numbers (amount, date, ratio) must be checked. |
| 怎么验证 AI 的引用? | How to verify AI's citations? | 逐个打开。AI 可能编造『看起来真实』的引用(作者、标题、期刊都对,但论文不存在)。用 Google Scholar / arXiv 搜标题。 | Open each one. AI may fabricate 'plausibly real' citations (author, title, journal all right, but the paper doesn't exist). Search the title on Google Scholar / arXiv. |
| 怎么验证 AI 的事实? | How to verify AI's facts? | 用『反向搜索』:搜『[事实] 是不是真的』、搜『[事实] 辟谣』。如果找不到支持来源,警惕。 | Use 'reverse search': search '[fact] is it true', search '[fact] debunk'. If you can't find supporting sources, be wary. |
| 怎么验证 AI 的代码? | How to verify AI's code? | 跑测试、读逻辑、查边界条件、用 linter。别盲信『能跑』——能跑不代表正确。 | Run tests, read logic, check edge conditions, use linter. Don't blindly trust 'it runs' — running doesn't mean correct. |
| 怎么验证 AI 的建议? | How to verify AI's advice? | 对照你的具体情境。AI 的建议是『一般性』的,你的情况是『特定』的。问自己:这条建议在我这里成立吗? | Compare with your specific situation. AI's advice is 'general', your situation is 'specific'. Ask: does this advice hold in my case? |
| 怎么验证 AI 的翻译? | How to verify AI's translation? | 用母语者视角读:地道吗?术语对吗?语气对吗?专业领域找同行核对。 | Read from a native speaker's view: idiomatic? terms right? tone right? For professional fields, get a peer to check. |
| 怎么验证 AI 的总结? | How to verify AI's summary? | 对照原文。总结是否遗漏关键信息?是否曲解?用『5 个 W』检查:谁、什么、何时、何地、为什么。 | Compare with the original. Does the summary miss key info? Misinterpret? Check with '5 Ws': who, what, when, where, why. |
| 怎么验证 AI 的分析? | How to verify AI's analysis? | 检查逻辑链:前提成立吗?推理有效吗?结论必要吗?有没有其他解释? | Check the logic chain: do premises hold? Is reasoning valid? Is the conclusion necessary? Are there other explanations? |
| 怎么验证 AI 的预测? | How to verify AI's predictions? | 预测最难验证。看依据:基于什么数据?假设是什么?置信度多少?如果依据不透明,别信。 | Predictions are hardest to verify. Check the basis: based on what data? What assumptions? What confidence? If the basis is opaque, don't trust. |
| 怎么验证 AI 的引用是否存在? | How to verify AI's citation exists? | 搜完整标题(不是关键词)。如果搜不到,大概率是编的。学术引用必须 100% 验证。 | Search the full title (not keywords). If not found, likely fabricated. Academic citations must be 100% verified. |
| 怎么验证 AI 的数据来源? | How to verify AI's data source? | 问 AI『数据从哪来』。如果它说不出具体来源,或来源模糊,警惕。可信数据应有明确出处。 | Ask AI 'where does the data come from'. If it can't give a specific source, or the source is vague, be wary. Credible data should have a clear source. |
| 怎么验证 AI 的逻辑? | How to verify AI's logic? | 找漏洞:有没有以偏概全?有没有因果倒置?有没有忽略反例?逻辑错,结论全错。 | Find gaps: any overgeneralization? Any reversed causation? Any ignored counterexamples? If logic is wrong, the whole conclusion is wrong. |
| 怎么验证 AI 的时效性? | How to verify AI's timeliness? | 问『这是哪年的信息』。AI 的训练有截止日期,时效性强的信息(价格、政策、技术)要查最新。 | Ask 'what year is this info'. AI's training has a cutoff; time-sensitive info (price, policy, tech) check for the latest. |
| 怎么验证 AI 的完整性? | How to verify AI's completeness? | 问『还有没有别的角度』。AI 可能只给一边。主动追问:反方观点是什么? | Ask 'are there other angles'. AI may only give one side. Proactively ask: what's the counterargument? |
| 怎么验证 AI 的可靠性? | How to verify AI's reliability? | 用『反事实测试』:问 AI 一个不存在的事(虚构的论文/事件),看它会不会编。会编的模型,要更警惕。 | Use 'counterfactual test': ask AI something that doesn't exist (a fake paper/event), see if it fabricates. If it fabricates, be more wary. |
C. 隐私与安全类(15 条)
C. Privacy & Security (15)
| 问题 | Question | 解答 | Answer |
|---|---|---|---|
| 免费版 AI 会保存我的数据吗? | Do free AI versions save my data? | 大多数免费版会用于训练。敏感信息别用免费版。商用版或本地部署更安全。 | Most free versions use data for training. Don't use free versions for sensitive info. Commercial or local deployment is safer. |
| 我可以把公司机密喂给 AI 吗? | Can I feed AI company secrets? | 不能。除非用的是企业版/本地部署,且有数据协议。默认假设:喂给公有 AI 的数据可能被训练。 | No. Unless using enterprise/local deployment with a data agreement. Default assumption: data fed to public AI may be trained on. |
| AI 会记住我吗? | Will AI remember me? | 单次对话内会(上下文),跨对话多数不记(除非产品有记忆功能)。别依赖它『记得』你。 | Within a single conversation, yes (context); across conversations, most don't (unless the product has memory). Don't rely on it 'remembering' you. |
| 怎么保护我的隐私? | How to protect my privacy? | ① 匿名化(化名、去标识);② 用商用/本地;③ 别喂身份证件、密码、财务;④ 了解数据协议。 | ① Anonymize (alias, de-identify); ② use commercial/local; ③ don't feed ID, passwords, finances; ④ understand the data agreement. |
| AI 生成的内容我能商用吗? | Can I commercially use AI-generated content? | 看平台和地区规则。多数地区:AI 生成内容版权模糊,部分不能注册版权。商用前查清楚。 | Depends on platform and regional rules. In most regions: AI-generated content copyright is fuzzy, some can't be registered. Check before commercial use. |
| AI 会泄露我的信息吗? | Will AI leak my info? | 理论上可能(训练、存储、攻击)。降低风险:匿名化、用可信平台、敏感数据本地处理。 | Theoretically possible (training, storage, attacks). Reduce risk: anonymize, use trusted platforms, process sensitive data locally. |
| 我能用 AI 处理客户数据吗? | Can I process customer data with AI? | 看数据协议和法规(如 GDPR)。多数情况:需匿名化 + 用合规平台 + 客户知情。 | Depends on data agreement and regulations (e.g. GDPR). Most cases: need anonymization + compliant platform + customer consent. |
| AI 会判断我的意图吗? | Will AI judge my intent? | 不会。AI 没有价值观判断,它按训练数据回应。你的责任是确保使用合法合规。 | No. AI has no value judgment; it responds per training data. Your responsibility is to ensure legal and compliant use. |
| 怎么用 AI 不违反法律? | How to use AI without breaking the law? | ① 别用 AI 造假(证件、签名);② 别侵犯版权;③ 别做歧视性内容;④ 遵守行业法规;⑤ 标注 AI 生成(如要求)。 | ① Don't use AI to forge (ID, signatures); ② don't infringe copyright; ③ no discriminatory content; ④ follow industry regulations; ⑤ label AI-generated (if required). |
| AI 会帮我做违法的事吗? | Will AI help me do illegal things? | 正规 AI 会拒绝明显违法的请求。但『灰色地带』要自己负责。别用 AI 绕开法律。 | Legitimate AI refuses clearly illegal requests. But 'gray areas' are your responsibility. Don't use AI to bypass the law. |
| 我能用 AI 生成虚假内容吗? | Can I use AI to generate false content? | 技术上能,但别用于欺骗。深度伪造(fake 名人、假新闻)是法律风险 + 道德问题。 | Technically yes, but don't use for deception. Deepfakes (fake celebrities, fake news) are legal risk + moral issue. |
| AI 会歧视我吗? | Will AI discriminate against me? | AI 可能带训练数据的偏见(性别、种族、地域)。遇到偏见,指出并要求换角度。 | AI may carry training-data bias (gender, race, region). When you encounter bias, point it out and ask for another angle. |
| 我能用 AI 处理医疗数据吗? | Can I process medical data with AI? | 敏感数据。需匿名化 + 合规平台 + 严格协议。AI 不能做诊断,只能辅助。 | Sensitive data. Need anonymization + compliant platform + strict agreement. AI can't diagnose, only assist. |
| AI 会记住我说的密码吗? | Will AI remember passwords I say? | 对话内可能在上下文里,但别把密码告诉任何 AI。密码只在你和系统之间。 | Within a conversation it may be in context, but don't tell any AI your passwords. Passwords stay only between you and the system. |
| 怎么用 AI 最安全? | How to use AI most safely? | ① 最小必要(只喂需要的);② 匿名化;③ 商用/本地;④ 了解平台协议;⑤ 敏感操作人工确认。 | ① Minimum necessary (only feed what's needed); ② anonymize; ③ commercial/local; ④ understand platform terms; ⑤ human confirm for sensitive ops. |
D. 进阶类(10 条)
D. Advanced (10)
| 问题 | Question | 解答 | Answer |
|---|---|---|---|
| 什么是 RAG? | What is RAG? | Retrieval-Augmented Generation:先从知识库检索相关资料,再让 LLM 基于资料回答。解决 LLM 知识过时和幻觉问题。 | Retrieval-Augmented Generation: first retrieve relevant material from a knowledge base, then let the LLM answer based on it. Solves LLM outdated knowledge and hallucination. |
| 什么是 AI Agent? | What is an AI Agent? | 能自主规划、调用工具、执行多步任务的 AI 系统。如:自动调研+写报告、自动操作电脑。 | An AI system that can autonomously plan, call tools, execute multi-step tasks. E.g., auto research + write report, auto operate a computer. |
| 什么是微调? | What is fine-tuning? | 用特定数据继续训练基础模型,让它更懂特定领域。如:用法律数据微调,让它更懂法律。 | Continue training a base model with specific data so it understands a specific domain better. E.g., fine-tune with legal data so it understands law better. |
| 什么是多模态? | What is multimodal? | AI 能处理多种模态(文字、图像、音频、视频)的能力。如:看图回答、听音转文字。 | AI's ability to process multiple modalities (text, image, audio, video). E.g., answer from an image, transcribe audio. |
| 什么是思维链(CoT)? | What is Chain-of-Thought (CoT)? | 让 AI 一步步推理再给答案,提升复杂任务准确性。提示里写『一步步思考』。 | Make AI reason step-by-step before giving the answer, improving accuracy on complex tasks. Write 'think step by step' in the prompt. |
| 什么是上下文窗口? | What is the context window? | AI 一次能『看到』的 token 数量上限。超过会被截断。长文档要分段或压缩。 | The token limit AI can 'see' at once. Exceeding gets truncated. Long documents must be split or compressed. |
| 什么是温度(temperature)? | What is temperature? | 控制 AI 输出的随机性。低温=确定/保守,高温=创意/多变。要准确用低温度,要创意用高温度。 | Controls randomness of AI output. Low = deterministic/conservative, high = creative/varied. Use low for accuracy, high for creativity. |
| 什么是 embedding? | What is embedding? | 把文字转成数字向量,让 AI 能比较语义相似度。RAG 的核心:检索就是找『向量最接近』的资料。 | Convert text into numeric vectors so AI can compare semantic similarity. The core of RAG: retrieval is finding the 'closest vector' material. |
| 什么是大模型 vs 小模型? | What is large vs small model? | 大模型(GPT/Claude)能力强但贵;小模型(7B-13B)便宜但弱。按任务选:要能力用大,要成本用小。 | Large models (GPT/Claude) are powerful but expensive; small (7B-13B) are cheap but weaker. Choose by task: capability → large, cost → small. |
| 什么是 AI 对齐? | What is AI alignment? | 让 AI 的行为符合人类意图和价值观的技术方向。避免 AI 做有害的事。是当前 AI 安全的核心研究。 | The technical direction of making AI's behavior align with human intent and values. Avoiding AI doing harm. A core current AI safety research area. |
40 个主流 AI 工具对比(2026)
40 Mainstream AI Tools Compared (2026)
| 类别 | Category | 工具 | Tool | 特点 | Feature | 适合 | Best for | 费用 | Cost |
|---|---|---|---|---|---|---|---|---|---|
| 通用 LLM | General LLM | ChatGPT | ChatGPT | 综合能力最强,生态大 | Strongest overall, big ecosystem | 日常通用 | Daily general | 免费/订阅 | Free/sub |
| 通用 LLM | General LLM | Claude | Claude | 长上下文、推理强、安全 | Long context, strong reasoning, safe | 长文档/复杂推理 | Long docs/complex reasoning | 免费/订阅 | Free/sub |
| 通用 LLM | General LLM | Gemini | Gemini | 多模态强、Google 生态 | Strong multimodal, Google ecosystem | 多模态/Google 用户 | Multimodal/Google users | 免费/订阅 | Free/sub |
| 通用 LLM | General LLM | DeepSeek | DeepSeek | 开源、性价比高 | Open-source, good value | 开发者/性价比 | Developers/value | 免费 | Free |
| 通用 LLM | General LLM | Llama | Llama | 开源、可本地部署 | Open-source, local deploy | 本地/隐私敏感 | Local/privacy-sensitive | 免费 | Free |
| 通用 LLM | General LLM | Qwen | Qwen | 开源、中文强 | Open-source, strong Chinese | 中文场景 | Chinese scenarios | 免费 | Free |
| 通用 LLM | General LLM | Mistral | Mistral | 开源、高效 | Open-source, efficient | 欧洲/高效 | Europe/efficient | 免费 | Free |
| 写作 | Writing | Jasper | Jasper | 营销文案强 | Strong marketing copy | 营销/品牌 | Marketing/brand | 订阅 | Sub |
| 写作 | Writing | Grammarly | Grammarly | 英文语法/风格 | English grammar/style | 英文写作 | English writing | 免费/订阅 | Free/sub |
| 写作 | Writing | Notion AI | Notion AI | 笔记+AI 一体 | Notes + AI integrated | 知识管理 | Knowledge mgmt | 订阅 | Sub |
| 写作 | Writing | QuillBot | QuillBot | 改写/同义替换 | Rewriting/synonyms | 改写/降重 | Rewrite/paraphrase | 免费/订阅 | Free/sub |
| 翻译 | Translation | DeepL | DeepL | 专业翻译质量高 | High professional translation quality | 专业翻译 | Professional translation | 免费/订阅 | Free/sub |
| 翻译 | Translation | Google Translate | Google Translate | 多语言、快 | Many languages, fast | 日常翻译 | Daily translation | 免费 | Free |
| 翻译 | Translation | Microsoft Translator | Microsoft Translator | 实时翻译、Teams | Real-time, Teams | 会议/Teams | Meetings/Teams | 免费 | Free |
| 编程 | Coding | GitHub Copilot | GitHub Copilot | 代码补全、集成 VS Code | Code completion, VS Code integrated | 编程 | Programming | 订阅 | Sub |
| 编程 | Coding | Cursor | Cursor | AI 编辑器、整项目理解 | AI editor, whole-project understanding | 重度编程 | Heavy programming | 订阅 | Sub |
| 编程 | Coding | Codeium | Codeium | 免费代码补全、开源 | Free code completion, open-source | 免费编程 | Free programming | 免费 | Free |
| 编程 | Coding | Replit | Replit | 在线 IDE + AI | Online IDE + AI | 快速原型 | Quick prototypes | 免费/订阅 | Free/sub |
| 图像 | Image | Midjourney | Midjourney | 艺术感最强 | Strongest artistry | 艺术/设计 | Art/design | 订阅 | Sub |
| 图像 | Image | DALL-E | DALL-E | ChatGPT 内置、理解好 | Built into ChatGPT, good understanding | 通用图像 | General image | 免费/订阅 | Free/sub |
| 图像 | Image | Stable Diffusion | Stable Diffusion | 开源、可本地 | Open-source, local | 本地/自定义 | Local/custom | 免费 | Free |
| 图像 | Image | Adobe Firefly | Adobe Firefly | 商用安全、Adobe 生态 | Commercially safe, Adobe ecosystem | 商用设计 | Commercial design | 订阅 | Sub |
| 音频 | Audio | ElevenLabs | ElevenLabs | 语音合成最真 | Most realistic voice synthesis | 配音/有声 | Voiceover/audio | 免费/订阅 | Free/sub |
| 音频 | Audio | Whisper | Whisper | 语音转文字、开源 | Speech-to-text, open-source | 转写/字幕 | Transcription/subtitles | 免费 | Free |
| 音频 | Audio | Suno | Suno | AI 作曲、生成歌曲 | AI composition, generate songs | 音乐制作 | Music production | 免费/订阅 | Free/sub |
| 音频 | Audio | Murf | Murf | 配音、多语言 | Voiceover, multilingual | 企业配音 | Enterprise voiceover | 订阅 | Sub |
| 视频 | Video | Runway | Runway | 文生视频、编辑 | Text-to-video, editing | 视频创作 | Video creation | 订阅 | Sub |
| 视频 | Video | Pika | Pika | 短视频、特效 | Short video, effects | 短视频/特效 | Short video/effects | 订阅 | Sub |
| 视频 | Video | Synthesia | Synthesia | 数字人视频、企业 | AI avatar video, enterprise | 企业培训 | Enterprise training | 订阅 | Sub |
| 视频 | Video | CapCut | CapCut | 剪辑 + AI 字幕 | Editing + AI subtitles | 短视频剪辑 | Short video editing | 免费 | Free |
| 研究 | Research | Elicit | Elicit | 学术文献分析 | Academic literature analysis | 科研 | Research | 免费/订阅 | Free/sub |
| 研究 | Research | Scite | Scite | 引用分析、支持/反驳 | Citation analysis, supports/refutes | 学术 | Academic | 免费/订阅 | Free/sub |
| 研究 | Research | Perplexity | Perplexity | 搜索 + 引用 | Search + citations | 研究/问答 | Research/Q&A | 免费/订阅 | Free/sub |
| 搜索 | Search | Perplexity | Perplexity | AI 搜索、带来源 | AI search, with sources | 信息检索 | Info retrieval | 免费/订阅 | Free/sub |
| 搜索 | Search | You.com | You.com | AI 搜索、隐私 | AI search, privacy | 隐私搜索 | Privacy search | 免费 | Free |
| 搜索 | Search | Google AI Overview | Google AI Overview | 搜索 + AI 摘要 | Search + AI summary | 日常搜索 | Daily search | 免费 | Free |
| 生产力 | Productivity | Notion AI | Notion AI | 笔记 + AI | Notes + AI | 知识管理 | Knowledge mgmt | 订阅 | Sub |
| 生产力 | Productivity | Microsoft 365 Copilot | Microsoft 365 Copilot | Office 集成 | Office integrated | Office 用户 | Office users | 订阅 | Sub |
| 生产力 | Productivity | Zapier AI | Zapier AI | 自动化 + AI | Automation + AI | 工作流自动化 | Workflow automation | 订阅 | Sub |
30 条 AI 素养误区与正解(对照表)
30 AI Literacy Myths vs Truths (Comparison)
| 误区 | Myth | 正解 | Truth |
|---|---|---|---|
| AI 是百科全书,什么都知道 | AI is an encyclopedia, knows everything | AI 是概率模型,不是知识库。它会编造、会过时、会遗漏。 | AI is a probabilistic model, not a knowledge base. It fabricates, goes stale, and omits. |
| AI 回答越长越对 | Longer AI answers are more correct | 长度≠正确。简洁准确的回答可能比冗长的更对。 | Length ≠ correctness. A concise, accurate answer may be more right than a verbose one. |
| AI 不会错,错了是用户笨 | AI doesn't err; errors are the user's fault | AI 会错,且会自信地错。错的责任在使用前没验证,不在模型。 | AI errs, and confidently so. The responsibility for the error is in not verifying before use, not in the model. |
| 提示越短越好 | Shorter prompts are better | 提示要『完整』不是『短』。该给的上下文要给够。 | Prompts should be 'complete', not 'short'. Give the context that's needed. |
| AI 能帮我思考 | AI can think for me | AI 能帮你『整理』和『加速』,但『思考』和『判断』是你。 | AI helps you 'organize' and 'accelerate', but 'thinking' and 'judging' are yours. |
| 用 AI = 作弊 | Using AI = cheating | 取决于用途。辅助学习不是作弊,代写作业是。关键看『学习是否发生』。 | Depends on use. Assisting learning isn't cheating; doing homework for you is. The key is 'whether learning happened'. |
| AI 懂我的意思 | AI understands my meaning | AI 按字面 + 统计推测,经常误解。说清楚比『它该懂』更可靠。 | AI reads literally + statistically, often misreading. Being explicit is more reliable than 'it should get it'. |
| 免费 AI 和付费 AI 一样好 | Free and paid AI are equally good | 通常付费更强(更长上下文、更准、新功能优先)。按需求选。 | Usually paid is stronger (longer context, more accurate, new features first). Choose by need. |
| AI 能记住所有对话 | AI remembers all conversations | 多数 AI 跨对话不记(除非有记忆功能)。别依赖它『记得』。 | Most AI don't remember across conversations (unless there's a memory feature). Don't rely on it 'remembering'. |
| AI 生成的一定有版权 | AI-generated content is always copyrighted | 多数地区 AI 生成内容版权模糊,部分不能注册。商用前查规则。 | In most regions, AI-generated content copyright is fuzzy; some can't be registered. Check rules before commercial use. |
| AI 不会歧视 | AI doesn't discriminate | AI 带训练数据偏见(性别/种族/地域)。要主动测试和纠正。 | AI carries training-data bias (gender/race/region). Actively test and correct. |
| AI 能预测未来 | AI can predict the future | AI 基于历史数据推测,不能预测真正的新颖事件。历史≠未来。 | AI infers from historical data; it can't predict truly novel events. Past ≠ future. |
| AI 越贵越好 | The more expensive, the better AI | 按任务选。简单任务便宜模型够;复杂任务才用强的。 | Choose by task. Simple tasks are fine with cheap models; complex ones need the strong ones. |
| AI 写的代码可以直接用 | AI-written code can be used as-is | 要跑测试、读逻辑、查边界。能跑≠正确。 | Run tests, read logic, check edges. Running ≠ correct. |
| AI 翻译比人好 | AI translation is better than humans | 日常 AI 很好,专业/文学要人校对。术语、语气、文化要人把关。 | Daily, AI is good; professional/literary needs human proofread. Terms, tone, culture need human gatekeeping. |
| AI 能帮我做所有事 | AI can do everything for me | AI 是杠杆,不是全能。判断、责任、创造是你。 | AI is leverage, not omnipotent. Judgment, responsibility, creation are yours. |
| 用 AI 会让我变笨 | Using AI makes me dumber | 取决于怎么用。辅助思考不变笨,替代思考会。 | Depends on how. Assisting thinking doesn't make you dumber; replacing thinking does. |
| AI 不会幻觉,是用户看错 | AI doesn't hallucinate; the user misread | AI 会幻觉(自信地编)。关键输出必须验证。 | AI hallucinates (confidently fabricates). Critical output must be verified. |
| 大模型越大越好 | The bigger the model, the better | 大模型强但贵、慢。小模型在特定任务可能更好。 | Large models are powerful but expensive and slow. Small models may be better for specific tasks. |
| AI 能理解情感 | AI understands emotions | AI 识别情绪信号,但不『体验』情感。别把它当情感伙伴的唯一。 | AI detects emotional signals but doesn't 'experience' emotion. Don't make it your only emotional companion. |
| 提示词有标准答案 | There's a standard answer for prompts | 没有万能提示。好提示是『具体 + 完整 + 迭代』。 | No universal prompt. Good prompts are 'specific + complete + iterative'. |
| AI 越新越好 | The newer AI, the better | 新模型有新能力,也有新坑。稳定 + 适合 > 追新。 | Newer models bring new capabilities and new pitfalls. Stable + suitable > chasing new. |
| AI 能取代我的职业 | AI will replace my profession | AI 取代『任务』,不取代『职业』。会用 AI 的人更安全。 | AI replaces 'tasks', not 'professions'. People who use AI are safer. |
| AI 不需要监管 | AI doesn't need regulation | AI 有风险(偏见/隐私/滥用),需要合理监管。 | AI has risks (bias/privacy/misuse) and needs reasonable regulation. |
| 本地 AI 一定比云端安全 | Local AI is always safer than cloud | 本地有隐私优势,但模型可能弱、成本高。按敏感度选。 | Local has a privacy advantage, but the model may be weaker and cost more. Choose by sensitivity. |
| AI 回答就是事实 | AI answers are facts | AI 回答是『最可能的文字』,不是『事实』。要验证。 | AI answers are 'most likely text', not 'facts'. Verify. |
| 一次对话就能学会 | You can learn in one conversation | AI 是工具,学习靠『用 + 练 + 复盘』。一次对话不够。 | AI is a tool; learning requires 'using + practicing + reviewing'. One conversation isn't enough. |
| AI 不能做创意 | AI can't be creative | AI 能做创意(组合、变体、灵感),但『原创灵魂』是你。 | AI can create (combination, variants, inspiration), but the 'original soul' is yours. |
| AI 永远在线不会累 | AI is always online and never tires | AI 没『疲劳』,但有上下文限制和速率限制。 | AI has no 'fatigue', but has context and rate limits. |
| AI 能给我的人生答案 | AI can give me life answers | AI 给分析、给选项、给盲点,但『你的人生』只能你答。 | AI gives analysis, options, blind spots; but 'your life' only you can answer. |
10 个真实案例:AI 素养改变职业
10 Real Cases: AI Literacy Changing Careers
| 案例 | Case | 背景 | Background | AI 素养动作 | AI literacy action | 结果 | Result |
|---|---|---|---|---|---|---|---|
| 案例 1 | Case 1 | 35 岁行政,被优化 | 35 admin, laid off | 自学 AI,用 AI 做简历+作品集,转型数据助理 | Self-taught AI, used AI for resume + portfolio, transitioned to data assistant | 3 个月入职数据助理,薪资持平 | Hired as data assistant in 3 months, salary maintained |
| 案例 2 | Case 2 | 28 岁教师,想提升效率 | 28 teacher, wants efficiency | 用 AI 出题、批改、个性化辅导,节省 2/3 批改时间 | Used AI for quiz, grading, personalized tutoring, saved 2/3 grading time | 能带更多学生,获校『效率先锋』 | Could take more students, won school's 'efficiency pioneer' |
| 案例 3 | Case 3 | 40 岁设计师,创意瓶颈 | 40 designer, creative bottleneck | 用 AI 出灵感、变体、配色,自己把关审美 | Used AI for inspiration, variants, color schemes, self-gated aesthetics | 产出翻倍,接更多项目 | Output doubled, took more projects |
| 案例 4 | Case 4 | 25 岁应届生,面试焦虑 | 25 fresh grad, interview anxiety | 用 AI 模拟面试 50 次,AI 追问+反馈 | Used AI to mock interviews 50 times, AI follow-up + feedback | 拿到 3 个 offer | Got 3 offers |
| 案例 5 | Case 5 | 50 岁创业者,不懂技术 | 50 entrepreneur, not tech-savvy | 用 AI 写商业计划、分析市场、做产品原型 | Used AI to write business plan, analyze market, build product prototype | 产品 2 周上线,融资成功 | Product launched in 2 weeks, funding successful |
| 案例 6 | Case 6 | 32 岁律师,文书繁重 | 32 lawyer, heavy paperwork | 用 AI 起草文书、查判例、整理证据 | Used AI to draft documents, search case law, organize evidence | 文书时间减 60%,专注核心 | Document time cut 60%, focused on core |
| 案例 7 | Case 7 | 22 岁学生,英语差 | 22 student, poor English | 用 AI 当语言陪练,每天 30 分钟对话 | Used AI as language sparring partner, 30 min daily conversation | 一年后英语流利,拿到海外实习 | English fluent after a year, got overseas internship |
| 案例 8 | Case 8 | 45 岁销售,不会写方案 | 45 sales, can't write proposals | 用 AI 写方案、分析客户、准备谈判 | Used AI to write proposals, analyze clients, prepare negotiations | 成单率提升 40% | Close rate up 40% |
| 案例 9 | Case 9 | 30 岁自媒体,内容枯竭 | 30 content creator, content dried up | 用 AI 脑暴、出结构、写初稿,自己加灵魂 | Used AI to brainstorm, structure, draft, self-added soul | 更新频率翻倍,涨粉 3 倍 | Update frequency doubled, followers tripled |
| 案例 10 | Case 10 | 38 岁 HR,招聘低效 | 38 HR, inefficient hiring | 用 AI 筛简历、写 JD、面试辅助 | Used AI to screen resumes, write JDs, assist interviews | 招聘周期从 6 周缩到 2 周 | Hiring cycle from 6 weeks to 2 weeks |
一页总结:AI 素养速记卡
One-Page Summary: AI Literacy Quick-Reference Card
| 维度 | Dimension | 核心 | Core | 口诀 | Mnemonic |
|---|---|---|---|---|---|
| 认知 | Cognition | AI 是概率模型,不是知识库 | AI is a probabilistic model, not a knowledge base | 先怀疑,再使用 | Doubt first, then use |
| 提示 | Prompting | 完整 > 短;迭代 > 一次 | Complete > short; iterate > one-shot | 上下文决定一切 | Context determines everything |
| 验证 | Verification | 关键必查:数字/引用/结论 | Key things must check: numbers/citations/conclusions | 交叉验证 + 反事实 | Cross-verify + counterfactual |
| 隐私 | Privacy | 敏感不喂公有;匿名化 | Don't feed public with sensitive; anonymize | 最小必要 + 本地商用 | Minimum necessary + local/commercial |
| 伦理 | Ethics | 主动测偏见;标注 AI 生成 | Actively test bias; label AI-generated | AI 有权力,要纠正 | AI has power, must correct |
| 应用 | Application | AI 初稿,你灵魂 | AI drafts, you soul | 重复给 AI,判断留自己 | Repetition to AI, judgment to yourself |
| 进阶 | Advanced | RAG/Agent/多模态/微调 | RAG/Agent/multimodal/fine-tune | 了解原理,按任务选 | Know principles, choose by task |
| 长期 | Long-term | AI 素养是底层能力,不是一阵风 | AI literacy is a foundational ability, not a fad | 用得对 > 用得多 | Using right > using much |
50 条 AI 素养行动清单(按周执行)
50 AI Literacy Action Items (By Week)
| 周 | Week | 动作 | Action | 检验标准 | Success metric |
|---|---|---|---|---|---|
| W1 | W1 | 每天用 AI 做 1 件日常事 | Do 1 daily thing with AI daily | 当天用满,不是装 | Used it that day, not just installed |
| W1 | W1 | 给每个提示加角色+背景+格式 | Add role+context+format to every prompt | 提示有 3 要素 | Prompt has 3 elements |
| W1 | W1 | 存下 3 个你常用的提示模板 | Save 3 commonly used prompt templates | 有 3 个模板 | Have 3 templates |
| W1 | W1 | 识别 1 个 AI 幻觉(反事实测试) | Spot 1 AI hallucination (counterfactual test) | 测试了不存在的引用 | Tested a non-existent citation |
| W1 | W1 | 了解你在用的 AI 的数据政策 | Learn the data policy of the AI you use | 能说出是否训练数据 | Can state if it trains on data |
| W1 | W1 | 把机密/敏感信息从提示里去掉 | Remove secrets/sensitive info from prompts | 提示无敏感信息 | Prompt has no sensitive info |
| W1 | W1 | 问 AI 一个它该不知道的问题 | Ask AI something it shouldn't know | 看到它怎么编 | Saw how it fabricates |
| W1 | W1 | 记录 1 个 AI 帮你的时刻 | Record 1 moment AI helped you | 有记录 | Have a record |
| W2 | W2 | 每个关键输出都验证一次 | Verify every key output once | 关键信息 100% 验证 | Key info 100% verified |
| W2 | W2 | 打开 3 个 AI 给的引用 | Open 3 AI-given citations | 3 个都验证 | All 3 verified |
| W2 | W2 | 用 2 个模型对比答同一问题 | Compare 2 models answering the same question | 看到差异 | Saw the difference |
| W2 | W2 | 让 AI 分步思考解 1 个难题 | Make AI think step-by-step for 1 hard problem | 答案更准 | More accurate answer |
| W2 | W2 | 给 AI 2-3 个例子(few-shot) | Give AI 2-3 examples (few-shot) | 输出更贴合 | Output more on-target |
| W2 | W2 | 建一个你的提示库文件夹 | Build a prompt library folder | 有 10+ 模板 | Have 10+ templates |
| W2 | W2 | 测试 1 个 AI 的偏见(反例) | Test 1 AI's bias (counterexample) | 指出 1 个偏见 | Pointed out 1 bias |
| W2 | W2 | 用 AI 帮你写 1 篇长文的初稿 | Use AI to draft 1 long piece | 初稿+你改灵魂 | Draft + you shaped the soul |
| W2 | W2 | 了解上下文窗口对你的影响 | Understand how context window affects you | 能分段长文档 | Can split long docs |
| W2 | W2 | 把 1 个重复任务交给 AI | Delegate 1 repetitive task to AI | 省了 30 分钟+ | Saved 30 min+ |
| W3 | W3 | 读 1 篇 LLM 原理科普 | Read 1 LLM principles explainer | 能讲清 token/上下文 | Can explain token/context |
| W3 | W3 | 理解温度/采样对输出的影响 | Understand how temperature/sampling affects output | 按任务调温度 | Adjust temperature by task |
| W3 | W3 | 了解 RAG 是什么、解决什么问题 | Understand what RAG is and what it solves | 能讲清原理 | Can explain the principle |
| W3 | W3 | 了解 AI Agent 能做什么 | Understand what AI Agent can do | 能设计 1 个简单 Agent | Can design 1 simple Agent |
| W3 | W3 | 用多模态 AI 做 1 件事 | Do 1 thing with multimodal AI | 用了非文字模态 | Used a non-text modality |
| W3 | W3 | 建立你的 AI 红线清单 | Build your AI red-line list | 有 5+ 条红线 | Have 5+ red lines |
| W3 | W3 | 向 1 个人解释 AI 原理 | Explain AI principles to 1 person | 对方听懂了 | The other understood |
| W3 | W3 | 评估 2 个 AI 工具的优劣 | Evaluate the pros/cons of 2 AI tools | 有对比结论 | Have a comparison conclusion |
| W3 | W3 | 把 1 个 AI 工作流写下来 | Write down 1 AI workflow | 有步骤文档 | Have a step doc |
| W3 | W3 | 了解 AI 版权的基本规则 | Understand the basic rules of AI copyright | 知道能不能商用 | Know if it can be commercial |
| W4 | W4 | 用多模态+Agent 做 1 个完整任务 | Do 1 complete task with multimodal+Agent | 端到端完成 | Completed end-to-end |
| W4 | W4 | 把你的提示库整理成分类 | Organize your prompt library by category | 分类清晰 | Categories clear |
| W4 | W4 | 建立你的验证习惯 SOP | Build your verification habit SOP | 有验证步骤 | Have verification steps |
| W4 | W4 | 用 AI 帮你复盘 1 个项目/决定 | Use AI to review 1 project/decision | 有复盘文档 | Have a review doc |
| W4 | W4 | 教 1 个人用 AI(入门) | Teach 1 person to use AI (intro) | 对方会用基础 | The other can use basics |
| W4 | W4 | 制定你的下 30 天 AI 计划 | Make your next 30-day AI plan | 有计划 | Have a plan |
| W4 | W4 | 跟踪 1 个 AI 新进展 | Track 1 AI new progress | 了解 1 个新东西 | Learned 1 new thing |
| W4 | W4 | 更新你的 AI 红线(按新认知) | Update your AI red-lines (by new understanding) | 红线迭代了 | Red lines iterated |
| W4 | W4 | 把你最好的 5 个提示标为精华 | Mark your best 5 prompts as essentials | 有 5 个精华 | Have 5 essentials |
| W4 | W4 | 用 AI 帮你规划 1 个学习目标 | Use AI to plan 1 learning goal | 有计划 | Have a plan |
| W4 | W4 | 把 AI 素养自评做 1 次(40 项) | Do 1 AI literacy self-assessment (40 items) | 有分数 | Have a score |
| W4 | W4 | 找到你最弱的 1 个维度,定改进计划 | Find your weakest 1 dimension, make improvement plan | 有改进目标 | Have an improvement goal |
| 持续 | Ongoing | 每天 30 分钟用 AI | Use AI 30 min daily | 每天都用 | Used daily |
| 持续 | Ongoing | 每周复盘 1 次 AI 使用 | Review AI usage 1x weekly | 有复盘 | Have review |
| 持续 | Ongoing | 每月更新 1 次提示库 | Update prompt library 1x monthly | 库在更新 | Library updating |
| 持续 | Ongoing | 每季度重做 1 次自评 | Redo self-assessment quarterly | 分数在升 | Score rising |
| 持续 | Ongoing | 跟进 AI 新工具/新能力 | Follow AI new tools/capabilities | 了解趋势 | Aware of trends |
| 持续 | Ongoing | 保持先验证的默认 | Keep verify-first as default | 从不盲信 | Never blindly trust |
| 持续 | Ongoing | 把 AI 当杠杆,不当替身 | Treat AI as leverage, not substitute | 你主导 | You lead |
| 持续 | Ongoing | 分享你的 AI 经验给 1 个人 | Share your AI experience with 1 person | 有分享 | Shared |
| 持续 | Ongoing | 持续学习,保持好奇心 | Keep learning, keep curiosity | 持续投入 | Continuous investment |
| 持续 | Ongoing | 把 AI 素养变成肌肉记忆 | Make AI literacy muscle memory | 自动化了 | Automated |
20 个 AI 素养学习路径(按目标)
20 AI Literacy Learning Paths (By Goal)
| 目标 | Goal | 核心路径 | Core path | 关键资源 | Key resources | 时间 | Time |
|---|---|---|---|---|---|---|---|
| 日常提效 | Daily efficiency | 提示工程 + 验证 + 场景应用 | Prompting + verification + scenario application | DeepLearning.AI 短课 + 提示词库 | DeepLearning.AI short course + prompt library | 2 周 | 2 weeks |
| 职场转型 | Career transition | AI 技能 + 作品集 + 面试 | AI skills + portfolio + interviews | Coursera + 案例库 | Coursera + case library | 3 个月 | 3 months |
| 编程入门 | Programming intro | AI 结对编程 + 项目 + 验证 | AI pair programming + projects + verification | GitHub Copilot + Replit + 测试 | GitHub Copilot + Replit + tests | 2 个月 | 2 months |
| 内容创作 | Content creation | 多模态 + 创意 + 版权 | Multimodal + creativity + copyright | Midjourney + Runway + 版权规则 | Midjourney + Runway + copyright rules | 1 个月 | 1 month |
| 科研提效 | Research efficiency | RAG + 文献 + 批判 | RAG + literature + critique | Elicit + arXiv + 批判 | Elicit + arXiv + critique | 1 个月 | 1 month |
| 数据分析 | Data analysis | AI 写代码 + 统计 + 验证 | AI writes code + stats + verification | 数据科学 + 测试 + 验证 | Data science + tests + verification | 2 个月 | 2 months |
| 语言学习 | Language learning | AI 陪练 + 输入 + 输出 | AI sparring + input + output | AI 对话 + 30 天计划 | AI conversation + 30-day plan | 3 个月 | 3 months |
| 创业 | Entrepreneurship | AI 全栈 + 产品 + 市场 | AI full-stack + product + market | AI 工具 + 商业计划 + 原型 | AI tools + business plan + prototype | 3 个月 | 3 months |
| 教育 | Education | AI 因材施教 + 出题 + 批改 | AI personalized teaching + quiz + grading | AI 教学 + 验证 + 伦理 | AI teaching + verification + ethics | 1 个月 | 1 month |
| 设计 | Design | 多模态 + 审美 + 品牌 | Multimodal + aesthetics + brand | Midjourney + 审美训练 | Midjourney + aesthetic training | 1 个月 | 1 month |
| 写作 | Writing | 提示 + 结构 + 润色 | Prompting + structure + polish | 提示库 + 结构 + 润色 | Prompt library + structure + polish | 2 周 | 2 weeks |
| 翻译 | Translation | AI 翻译 + 校对 + 术语 | AI translation + proofread + terms | DeepL + 术语库 | DeepL + glossary | 2 周 | 2 weeks |
| 营销 | Marketing | AI 文案 + 投放 + 分析 | AI copy + ads + analysis | AI 文案 + 数据 | AI copy + data | 1 个月 | 1 month |
| 法律 | Legal | AI 查判例 + 起草 + 保密 | AI search case law + draft + confidentiality | AI 法律 + 保密 + 验证 | AI legal + confidentiality + verification | 1 个月 | 1 month |
| 金融 | Finance | AI 分析 + 风控 + 合规 | AI analysis + risk + compliance | AI 金融 + 合规 | AI finance + compliance | 2 个月 | 2 months |
| 医疗 | Healthcare | AI 查文献 + 辅助 + 隐私 | AI literature + assist + privacy | AI 医疗 + 隐私 | AI medical + privacy | 2 个月 | 2 months |
| 自媒体 | Self-media | AI 内容 + 涨粉 + 变现 | AI content + followers + monetize | AI 内容 + 平台规则 | AI content + platform rules | 2 个月 | 2 months |
| 家庭教育 | Family education | AI 辅导 + 习惯 + 伦理 | AI tutoring + habit + ethics | AI 教育 + 亲子 | AI education + parent-child | 1 个月 | 1 month |
| 银发学习 | Silver learning | AI 入门 + 实用 + 耐心 | AI intro + practical + patience | AI 入门 + 大字体 | AI intro + large font | 1 个月 | 1 month |
| 学术严谨 | Academic rigor | 验证 + 溯源 + 批判 | Verification + tracing + critique | 学术工具 + 批判 | Academic tools + critique | 2 个月 | 2 months |
40 个 AI 术语中英对照(扫盲表)
40 AI Terms: Chinese-English Glossary
| 术语 | Term | 一句话解释 | One-line definition |
|---|---|---|---|
| 大语言模型 | Large Language Model (LLM) | 用海量文本训练的、能理解和生成语言的人工智能模型 | An AI model trained on massive text that understands and generates language |
| token | token | AI 处理文字的最小单位(中文约 1-2 字 = 1 token) | The smallest unit AI processes text in (Chinese ~1-2 chars = 1 token) |
| 上下文窗口 | Context Window | AI 一次能记住的文字量上限 | The max text AI can remember at once |
| 温度 | Temperature | 控制 AI 输出随机性的参数(低=确定,高=创意) | A parameter controlling output randomness (low=deterministic, high=creative) |
| 幻觉 | Hallucination | AI 自信地编造不存在的事实 | AI confidently fabricating non-existent facts |
| 提示工程 | Prompt Engineering | 设计提示词让 AI 输出更好的技术 | Designing prompts to get better AI output |
| few-shot | few-shot | 在提示里给 2-3 个例子引导 AI 的格式 | Giving 2-3 examples in the prompt to guide AI's format |
| 思维链 | Chain-of-Thought (CoT) | 让 AI 一步步推理再给答案的方法 | Making AI reason step-by-step before answering |
| 微调 | Fine-tuning | 用特定数据继续训练基础模型的方法 | Continuing to train a base model with specific data |
| RAG | RAG | 先检索资料再生成回答的技术(解决知识过时) | Retrieve material first, then generate (solves outdated knowledge) |
| embedding | embedding | 把文字转成数字向量表示语义的技术 | Converting text into numeric vectors to represent semantics |
| Agent | AI Agent | 能自主规划、调用工具、执行多步任务的 AI 系统 | An AI system that plans, calls tools, executes multi-step tasks |
| 多模态 | Multimodal | AI 能处理文字/图像/音频/视频多种模态的能力 | AI's ability to process text/image/audio/video |
| 对齐 | Alignment | 让 AI 行为符合人类意图和价值观的技术 | Making AI behavior align with human intent and values |
| 推理 | Inference | 用已训练好的模型生成输出的过程(区别于训练) | Using a trained model to generate output (vs training) |
| 训练 | Training | 用数据让模型学习参数的过程 | Using data to let the model learn parameters |
| 参数量 | Parameter Count | 模型里可学习数字的数量(衡量规模) | The number of learnable numbers in the model (measures scale) |
| 开源 | Open Source | 模型权重公开、可下载本地用的模型 | Model weights are public, downloadable to run locally |
| 闭源 | Closed Source | 模型权重不公开、只能 API 调用的模型 | Model weights not public, only via API |
| API | API | 调用 AI 模型的编程接口 | The programming interface to call AI models |
| token 限制 | Token Limit | 一次输入/输出的 token 上限 | The token limit for one input/output |
| 速率限制 | Rate Limit | 单位时间内能调用 AI 的次数上限 | The max calls per unit time |
| 系统提示 | System Prompt | 设定 AI 角色和规则的底层提示 | The underlying prompt setting AI's role and rules |
| 角色设定 | Role-playing | 在提示里给 AI 指定一个专业角色 | Assigning AI a professional role in the prompt |
| 输出格式 | Output Format | 要求 AI 用特定格式(表格/JSON/列表) | Requiring AI to use a specific format (table/JSON/list) |
| 流式输出 | Streaming | AI 边生成边显示(不用等完) | AI displays as it generates (no need to wait) |
| 停止词 | Stop Word | 让 AI 在特定词处停止输出的设置 | A setting to make AI stop at a specific word |
| 采样 | Sampling | 从概率分布里选下一个 token 的方法 | A method to pick the next token from a probability distribution |
| top-p | top-p | 只从概率最高的 p 比例词里采样的方法 | Sampling only from the top p% most probable words |
| top-k | top-k | 只从概率最高的 k 个词里采样的方法 | Sampling only from the top k most probable words |
| 重复惩罚 | Repetition Penalty | 降低 AI 重复词概率的参数 | A parameter lowering AI's repetition probability |
| 知识截止日期 | Knowledge Cutoff | AI 训练数据的最后时间(之后的不知道) | The last time of AI's training data (doesn't know after) |
| 幻觉率 | Hallucination Rate | AI 编造内容占输出的比例(衡量可靠性) | The ratio of AI's fabricated content to output (measures reliability) |
| 基准测试 | Benchmark | 用标准题目测模型能力的评测 | Testing model ability with standard questions |
| MMLU | MMLU | 测模型多学科知识的大基准 | A big benchmark testing multi-discipline knowledge |
| GPT-4o | GPT-4o | OpenAI 的多模态模型(能听能说) | OpenAI's multimodal model (can listen and speak) |
| Claude | Claude | Anthropic 的模型(长上下文、安全) | Anthropic's model (long context, safe) |
| Gemini | Gemini | Google 的模型(多模态强) | Google's model (strong multimodal) |
| Llama | Llama | Meta 的开源模型(可本地部署) | Meta's open-source model (local deployable) |
| DeepSeek | DeepSeek | 深度求索的开源模型(性价比高) | DeepSeek's open-source model (good value) |
| Qwen | Qwen | 阿里的开源模型(中文强) | Alibaba's open-source model (strong Chinese) |
| GPT | GPT | OpenAI 的生成式预训练 Transformer 系列 | OpenAI's Generative Pre-trained Transformer series |
| Transformer | Transformer | 当前大模型的基础架构 | The base architecture of current large models |
15 条 AI 使用红线(绝不碰)
15 AI Usage Red Lines (Never Touch)
| 红线 | Red line | 为什么危险 | Why dangerous | 正确做法 | Right way |
|---|---|---|---|---|---|
| 用 AI 伪造证件/签名/身份 | Using AI to forge ID/signatures/identity | 违法(伪造文件罪) | Illegal (forgery crime) | 只用 AI 做合法内容 | Only use AI for legal content |
| 用 AI 生成深度伪造骗人 | Using AI to generate deepfakes to deceive | 违法 + 道德(侵犯人格) | Illegal + immoral (violates personality) | 不生成假真人 | Don't generate fake real people |
| 把他人隐私喂给 AI | Feeding others' privacy to AI | 侵犯隐私 | Infringes privacy | 匿名化后再用 | Anonymize before use |
| 把公司机密喂给公有 AI | Feeding company secrets to public AI | 泄露 + 违反保密协议 | Leak + breach of NDA | 用企业版/本地 | Use enterprise/local |
| 用 AI 代学生作业并谎称自做 | Using AI to do student homework and claiming it's your own | 学术不端 | Academic misconduct | 用 AI 辅助,如实标注 | Use AI to assist, label honestly |
| 用 AI 做医疗诊断并照做 | Using AI for medical diagnosis and following it | AI 会误诊,害身体 | AI may misdiagnse, harming the body | AI 查文献,医生下诊断 | AI looks up literature, doctor diagnoses |
| 用 AI 做法律判断并据此行动 | Using AI for legal judgment and acting on it | AI 引用可能不存在,害自己 | AI citations may not exist, harming you | AI 辅助,律师确认 | AI assists, lawyer confirms |
| 用 AI 做投资并全仓押上 | Using AI for investment and going all-in | AI 不预测未来,可能亏光 | AI doesn't predict the future, may lose everything | AI 分析,自己决策,分散 | AI analyzes, you decide, diversify |
| 用 AI 生成内容冒充真人 | Using AI-generated content impersonating a real person | 欺骗 + 侵权 | Deception + infringement | 标注 AI 生成 | Label as AI-generated |
| 用 AI 写代码不测试直接上线 | Using AI code without testing, going live | 可能有大 bug,出事故 | May have big bugs, causing incidents | 跑测试 + 读逻辑 + 查边界 | Run tests + read logic + check edges |
| 用 AI 查实时数据当准 | Using AI's real-time data as accurate | 训练有截止日期,可能过时 | Training has a cutoff, may be outdated | 先查权威来源,再 AI 整理 | Check authoritative sources first, then AI organizes |
| 用 AI 做歧视性内容 | Using AI for discriminatory content | 违法 + 伤害 | Illegal + harmful | 主动测偏见,拒绝歧视 | Actively test bias, refuse discrimination |
| 用 AI 绕过版权(洗稿) | Using AI to bypass copyright (rewriting) | 侵权 | Infringement | 尊重原创,标注参考 | Respect originals, label references |
| 让 AI 替你做重大人生决定 | Letting AI make major life decisions for you | AI 没有你的处境和价值观 | AI has neither your circumstances nor values | AI 给分析,你做决定 | AI analyzes, you decide |
| 把密码/密钥告诉 AI | Telling AI your passwords/keys | 泄露,账号被盗 | Leak, account stolen | 密码只在你和系统之间 | Passwords only between you and the system |
20 个高频问题(一问一答,扫盲)
20 High-Frequency Questions (Q&A, Basics)
一页速记卡(10 条核心,贴在桌前)
One-Page Cheat Card (10 Core, Stick on Your Desk)
| 核心 | Core | 一句话 | One-liner |
|---|---|---|---|
| 1 | 1 | AI 是模式匹配,不是真理解 | AI is pattern matching, not real understanding |
| 2 | 2 | 关键输出必验证(引用/数字/代码) | Verify key outputs (citations/numbers/code) |
| 3 | 3 | 给角色 + 背景 + 格式 + 约束 | Give role + context + format + constraints |
| 4 | 4 | 分步思考 + 少样本 + 自检 | Step-by-step + few-shot + self-check |
| 5 | 5 | 敏感信息绝不喂公有 AI | Never feed sensitive info to public AI |
| 6 | 6 | AI 给初稿,你给灵魂 + 把关 | AI gives draft, you give soul + gatekeeping |
| 7 | 7 | 红线不碰(法律/健康/账号/隐私) | Don't touch red lines (legal/health/account/privacy) |
| 8 | 8 | 用 AI 强化你,不是替代你 | Use AI to strengthen you, not replace you |
| 9 | 9 | 持续跟进新工具,底层不变 | Keep up with new tools, underlying unchanged |
| 10 | 10 | 先验证,再使用,最后信任 | Verify first, then use, finally trust |
AI 素养 10 条常识速查
10 AI Literacy Common-Sense Facts
📋 10 条 AI 素养常识(30 秒读完)
📋 10 AI Literacy Common Sense (Read in 30s)
| 常识 | Fact | 记住什么 | Remember |
|---|---|---|---|
| AI 会编 | AI hallucinates | 关键信息必须验证 | Verify key info |
| AI 无长期记忆 | No long-term memory | 重要信息要重复/存外部 | Repeat/save important info |
| AI 有截止日期 | Knowledge cutoff | 实时数据要另查 | Check real-time separately |
| AI 不是专家 | Not an expert | 高风险找真人 | Find real expert for high-risk |
| AI 有偏见 | AI has bias | 多源验证,别单信 | Multi-source verify |
| AI 快但浅 | Fast but shallow | 深度任务要迭代 | Iterate for deep tasks |
| AI 能辅助不能替代 | Assist not replace | 你主导决策 | You lead decisions |
| AI 数据会被用 | Data may be used | 敏感信息别喂 | Don't feed sensitive info |
| AI 工具在变 | Tools change | 学底层方法 | Learn underlying method |
| AI 是杠杆 | AI is leverage | 放大你的能力 | Amplify your ability |