AI 提示詞: 技能發展提示詞

💡 AI 模型在英文提示詞下表現最佳。因此,提示詞本文以英文呈現。使用英文輸入可獲得更準確、更詳細的回應。 學新技能不必緩慢或挫折。掙扎多年與快速上手之間的差異在於結構化練習——知道先學什麼、如何有效練習、如何突破瓶頸。這些提示詞運用學習科學(刻意練習、技能遷移、間隔效應),經過四大 AI 模型實測。

最近測試日期 Mar 15, 2026 · 模型: GPT-4.1, Gemini 2.5 Pro, Claude Sonnet 4, Grok 3

技能學習路線圖

設計從新手到勝任的高效路徑——避免教學陷阱,真正建立能力

**The skill I want to learn:**
[SPECIFIC SKILL — the more precise, the better the roadmap]

**My current level:**
- Complete beginner / Some exposure / Intermediate / Advanced in related areas
- Related skills I already have: [ANYTHING TRANSFERABLE]
- What I've tried so far: [COURSES, BOOKS, PRACTICE]

**My goal with this skill:**
- What I want to be able to DO (not just know): [SPECIFIC CAPABILITY]
- Target competence level: [FUNCTIONAL / PROFICIENT / EXPERT]
- Timeline: [REALISTIC DEADLINE]

**My constraints:**
- Hours per week for practice: [HOURS]
- Budget: [AMOUNT]
- Learning style preference: [VISUAL / READING / HANDS-ON / SOCIAL]

Create a learning roadmap:

1. **Skill Decomposition**: Break this skill into 5-8 sub-skills. Which sub-skills are foundational (must learn first) vs. which can be learned in parallel?

2. **The 80/20 Analysis**: Which 20% of the sub-skills will give me 80% of the practical capability? Learn these first.

3. **Phase Plan**:
   - **Phase 1 — Foundation (first 20%):** Core concepts and basic competence
     - What to learn (minimal theory, maximum practice)
     - First project: something simple enough to finish in a week
     - How to know you're ready for Phase 2
   - **Phase 2 — Application (next 50%):** Building real capability
     - 3-5 progressively harder projects
     - Where to get feedback (communities, mentors, self-assessment)
     - Common plateaus at this stage and how to push through
   - **Phase 3 — Refinement (final 30%):** Going from good to great
     - Deliberate practice targets (specific weaknesses to drill)
     - Advanced resources worth the investment at this stage
     - How to develop your own style/approach

4. **Resource Curation**: For each phase, the ONE best resource (not 10 — one):
   - Best free resource
   - Best paid resource (if worth the investment)
   - Best community for accountability and feedback

5. **Anti-Patterns**: Common mistakes people make learning this skill and how to avoid them

進階技巧

學習最大的錯誤是花太多時間在「學習模式」(課程、書籍、教學)而不夠多在「實作模式」(專案、練習、應用)。目標是 20% 學習、80% 實作——大多數人的比例恰好相反。

已測試 Mar 15, 2026

刻意練習設計師

設計專注的練習環節,比隨意重複快十倍地建立技能

**The skill I'm practicing:**
[SKILL]

**My current level:**
[DESCRIBE HONESTLY — what you can and can't do]

**Where I'm stuck or plateaued:**
[SPECIFIC AREA WHERE PROGRESS HAS STALLED]

**How I currently practice:**
[DESCRIBE YOUR TYPICAL PRACTICE SESSION]

**Time available per session:**
[MINUTES]

**Sessions per week:**
[NUMBER]

Design a deliberate practice protocol:

1. **Weakness Diagnosis**: Based on my current level and plateau, what specific micro-skill needs the most work? (Not "get better at guitar" but "clean chord transitions between G and C at 120bpm")

2. **Session Structure** (for a [X]-minute session):
   - Warm-up (5 min): What to do to activate relevant neural pathways
   - Focused drill (60-70% of time): The specific exercise, at the edge of my ability
   - Integration (20-30% of time): Apply the drill in realistic context
   - Cool-down (5 min): Review what went well and what to focus on next session

3. **Progressive Overload**: How to make each session slightly harder:
   - Difficulty knobs to turn (speed, complexity, constraints, pressure)
   - When to increase difficulty (not when it feels easy — when accuracy hits 80%+)
   - How to scale back without feeling like failure

4. **Feedback Mechanisms**:
   - How to get immediate feedback during practice (recording, metrics, checklists)
   - Self-assessment framework for post-session review
   - When to seek external feedback vs. self-correct

5. **Practice Journal Template**: What to log after each session:
   - What I practiced
   - What was hard
   - What clicked
   - Focus for next session

6. **Weekly Practice Plan**: How to distribute [X] sessions across the week for optimal retention (spacing effect)

進階技巧

刻意練習有四個條件:每次練習有明確目標、完全專注(不分心)、即時回饋、在當前能力邊緣操作。缺少任何一個,練習就變成了重複——感覺有生產力但實際上沒有。

已測試 Mar 15, 2026

瓶頸突破器

診斷你為何卡住,並獲得突破技能瓶頸的具體策略

**The skill I'm plateaued in:**
[SKILL]

**My level:**
[WHERE I AM — be specific about what you CAN do]

**The plateau:**
- What I'm trying to get better at: [SPECIFIC CAPABILITY]
- How long I've been stuck: [DURATION]
- What my practice looks like: [TYPICAL SESSION]
- How much I practice: [FREQUENCY AND DURATION]

**What I've already tried to break through:**
[CHANGES YOU'VE MADE, NEW APPROACHES, etc.]

**My theory on why I'm stuck:**
[YOUR BEST GUESS]

Diagnose and break this plateau:

1. **Plateau Type Identification**: Which type of plateau is this?
   - **Fundamentals gap**: I skipped or didn't fully master a foundational sub-skill
   - **Practice quality**: I'm practicing in my comfort zone, not at my edge
   - **Mental model**: I'm using a framework that worked at lower levels but doesn't scale
   - **Physical/cognitive limit**: I've hit a genuine capacity wall that needs a different approach
   - **Feedback gap**: I can't tell what I'm doing wrong
   - **Motivation plateau**: Boredom or identity conflict

2. **Root Cause Investigation**: For each likely plateau type, what specific evidence points to it? What diagnostic exercise could confirm?

3. **Breakthrough Strategies** (tailored to the diagnosed type):
   - 3 specific practice modifications
   - 1 constraint challenge (practice with an artificial limitation to expose weaknesses)
   - 1 stretch challenge (attempt something well above current level to recalibrate "normal")

4. **Expert Pattern**: How do people who are one level above me practice differently? What do they focus on that I'm probably not?

5. **30-Day Breakthrough Plan**: A specific 4-week protocol designed to break this exact plateau, with weekly checkpoints to assess if it's working

進階技巧

瓶頸通常不是努力的問題——而是在錯誤的難度水平練習錯誤的內容。你要麼在練已經簡單的東西(舒適圈),要麼在跳過有隱藏缺口的基礎功。這個提示詞幫你找出是哪種情況。

已測試 Mar 15, 2026

技能遷移對照表

從現有專長中辨識可遷移的技能,加速在新領域的學習

**The new skill I want to learn:**
[NEW SKILL]

**My existing skills and expertise:**
[LIST ALL RELEVANT SKILLS, EVEN ONES THAT SEEM UNRELATED]
For each, note your level: Beginner / Intermediate / Advanced / Expert

**Why I want to learn the new skill:**
[GOAL]

Map my skill transfer advantages:

1. **Transfer Inventory**: For each existing skill, identify:
   - Direct transfers (same sub-skill, different context — e.g., writing → copywriting)
   - Analogous transfers (similar mental model, different domain — e.g., debugging code → troubleshooting hardware)
   - Meta-skill transfers (learning how to learn, practice discipline, frustration tolerance)

2. **Advantage Map**: Where will my existing expertise make me learn faster than a true beginner? Be specific about which sub-skills I can skip or accelerate through.

3. **Interference Risks**: Where might existing skills actually SLOW me down? (Expert beginners often try to apply old frameworks where they don't fit)
   - What to unlearn or set aside
   - Where "beginner's mind" is more valuable than experience

4. **Accelerated Learning Plan**: A customized roadmap that:
   - Skips content I already know from transfer
   - Focuses extra time on genuinely new sub-skills
   - Uses analogies from my expertise to build new mental models faster
   - Explicitly addresses interference risks

5. **Synergy Opportunities**: How can the new skill enhance my existing skills? Where does the combination become more valuable than either alone?

進階技巧

成年人學新技能比自己想像的更快,因為已經有可遷移的心智模型。學新樂器的音樂家不是從零開始——他們理解節奏、結構和練習紀律。明確地對照這些遷移能大幅加速學習。

已測試 Mar 15, 2026

學習自律系統

在動力消退和生活忙碌時保持技能發展的一致性

**The skill I'm developing:**
[SKILL]

**My current practice schedule:**
[WHAT I'M DOING / SUPPOSED TO BE DOING]

**My consistency track record:**
- Last month's actual practice days: [NUMBER OUT OF TARGET]
- Common reasons I skip: [LIST]
- Longest streak: [DAYS]
- What broke the streak: [WHAT HAPPENED]

**My learning style:**
- I'm motivated by: [PROGRESS TRACKING / SOCIAL PRESSURE / COMPETITION / DEADLINES / CURIOSITY]
- I tend to quit when: [BORED / FRUSTRATED / BUSY / NO VISIBLE PROGRESS]
- Best time of day for practice: [WHEN]

Design a learning accountability system:

1. **Minimum Viable Practice (MVP)**: The absolute smallest practice session that still counts — for days when everything goes wrong. Must take less than 10 minutes and still build skill.

2. **Schedule Architecture**:
   - Optimal weekly schedule given my constraints
   - How to protect practice time from encroachment
   - Backup slots for when primary time gets taken
   - Travel/disruption protocol

3. **Progress Tracking System**:
   - What to measure (leading indicators, not just outcomes)
   - Tracking method that takes <1 minute per session
   - Monthly milestone markers so I can SEE improvement
   - How to track quality, not just quantity of practice

4. **Motivation Architecture**:
   - For when I'm bored: variety injection techniques
   - For when I'm frustrated: difficulty recalibration
   - For when I'm busy: MVP protocol activation
   - For when progress feels invisible: comparison framework (me 1 month ago vs. now)

5. **Social Accountability**:
   - Finding practice partners or learning groups
   - Public commitment strategies that work for my personality
   - Teaching as accountability (explain what you learned to someone)

6. **Quarterly Review**: Assessment questions to evaluate learning ROI and adjust approach

進階技巧

技能發展最大的威脅不是缺乏天賦——而是練習環節之間的間隔。技能在習得階段長時間休息時會衰退。每天 20 分鐘的持續練習幾乎在所有技能上都勝過週末 3 小時的集中練習。

已測試 Mar 15, 2026

模型比較

基於實際測試結果 — 非假設推測。 查看測試方法

C

Claude Sonnet 4

擅長將複雜技能分解為可學習的子技能,並找出最有效的學習順序。瓶頸診斷最佳。

最佳技能拆解
G

GPT-4.1

建立最詳細、結構化的練習環節設計,附有清晰的漸進式負荷方案。

最佳練習計畫
G

Gemini 2.5 Pro

最擅長推薦具體、最新的學習資源,並將建議建立在學習科學研究之上。

最佳資源策展
G

Grok 3

對你的時間表是否切實可行以及你目前方法中哪裡在浪費時間最為直接。

最佳誠實評估

在 NailedIt 中試試

將上方的提示詞貼到 NailedIt,並排比較各模型的回應。

進階技巧

1

在購買任何課程之前先使用技能學習路線圖——大多數人在綜合課程上浪費數月,而實際上只需要 20% 的內容就能獲得 80% 的能力

2

進步停滯超過 2 週就立即使用瓶頸突破器——用錯誤方式在瓶頸期練習越久,突破就越困難

3

開始學任何新技能時都運行技能遷移對照表——你會發現自己比想像中更接近勝任,這會改變你的整個學習策略