Lecture 01 - Introduction to AI

1. Exam Focus（真题对齐）

• High-frequency (A 档)：

  • 2022-2023 Q1(a): Relationship between AI and Machine Learning (subset / interchangeable / unrelated / AI ⊂ EC trap)

• Medium (B 档)：

  • Definition-style questions (often used as MCQ/short answer):
    • What is intelligence? (keywords: learn/adapt/problem-solve)
    • What is AI? (simulate/extend/enhance intelligence)

• Low (C 档)：

  • AI history timeline / AI winter / expert systems / deep learning milestones (background knowledge, rarely a direct scoring point)
  • AI applications examples (only for “give 1-2 examples” type questions)

• Scoring checklist：

  • [ ] Can instantly pick “ML ⊂ AI” in MCQ and reject traps (“AI=ML”, “AI ⊂ ML”, “AI ⊂ EC”, “AI and ML unrelated”)
  • [ ] Can write 3 keywords for intelligence (learn/adapt/problem-solving) in bullet form
  • [ ] Can write 1-sentence AI definition (simulate/extend/enhance intelligence)
  • [ ] Can state: EC and ML are approaches/branches within AI, not the superset

2. Key Concepts（知识点）

A 档（必会）

• AI vs ML relationship (exam MCQ trap-proof)
  • Correct: Machine Learning is a subset of AI
  • Incorrect traps:
    • AI and ML are unrelated
    • AI and ML are interchangeable
    • AI is a subset of ML
    • AI is a subset of evolutionary computation

B 档（可能考）

• What is Intelligence? (write 3 bullets)

  • Cognitive abilities (learn from experience + apply to new situations)
  • Adaptation (adjust strategies in different contexts)
  • Problem-solving (analyze → generate solutions → evaluate → implement)

• Definition of AI (1 sentence)

  • AI studies and develops theories/methods/technologies/systems for simulating, extending, and enhancing intelligence.

C 档（不太可能考）

• AI timeline / AI winter / expert systems / deep learning milestones (know only “one-line summary”)
• AI applications (be able to give 1–2 examples)

3. Must-know Formulas / Algorithms（公式/算法模板）

This lecture has no required formulas. Use these answer templates:

• Template A (MCQ / True-False: AI vs ML relationship)

  1. Identify the statement type: subset / interchangeable / unrelated
  2. Use the rule: AI is the broader field; ML is one approach inside AI
  3. Choose: “ML is a subset of AI”
  4. Reject traps: “AI=ML”, “AI ⊂ ML”, “AI ⊂ EC”, “unrelated”

• Template B (Short answer: define intelligence)

  • Write 3 bullets: learn / adapt / problem-solving
  • Each bullet: 1 short phrase (no long story)

• Template C (Short answer: define AI)

  • One sentence: simulate + extend + enhance intelligence

4. Worked Examples（例题与解答）

A1. AI vs ML relationship

Problem 1 (MCQ-style)

Which statement is correct?

A. ML is a subset of AI

B. AI and ML are unrelated

C. AI and ML are interchangeable

D. AI is a subset of ML

E. AI is a subset of evolutionary computation

Solution :

• AI is a broad field aiming to build intelligent behavior.
• ML is one approach inside AI (learning from data).
• Therefore ML ⊂ AI.
• All other statements contradict the subset relationship or invert it.

Final Answer : A

Problem 2 (True/False)

(1) AI is a subset of machine learning.

(2) Evolutionary computation is one approach within AI.

(3) AI and machine learning are interchangeable.

Solution :

(1) False. The subset direction is reversed. ML ⊂ AI.

(2) True. EC is an approach used in AI for optimization/search.

(3) False. ML is only a part of AI.

Final Answer (1) × (2) ✓ (3) ×

B1. Definitions

Problem Answer in ≤4 sentences: (1) What is intelligence? (≥3 keywords) (2) What is AI? (1 sentence)

Solution (1) Intelligence: learn from experience, adapt strategies to context, and solve problems (analyze → generate → evaluate → implement). (2) AI: studies and develops theories/methods/technologies/systems for simulating, extending, and enhancing intelligence.

Final Answer Use the two lines above.

5. Common Mistakes（易错点）

(ordered by importance)

1. Treating “AI = ML” as correct.
2. Picking “AI is a subset of ML” by reversing the subset direction.
3. Falling for the trap “AI is a subset of evolutionary computation”.
4. Writing definition answers with no keywords (e.g., “intelligence is being smart”).
5. Writing AI definition as “AI is machine learning”.
6. Giving long history stories instead of scoring points (keywords + structure).
7. Not using bullet points in definition questions (hard to award keyword marks).
8. Mixing up “approach/branch within AI” vs “superset of AI”.
9. Memorizing all AI application examples instead of knowing 1–2.
10. Answering without justification when the question asks “why” (loses explanation marks).

6. Quick Review（考前 10 分钟速记）

• One-liner: ML is a subset of AI. Reject: “AI=ML”, “AI ⊂ ML”, “AI ⊂ EC”, “AI and ML unrelated”.
• Intelligence keywords: learn / adapt / problem-solving.
• AI definition keywords: simulate / extend / enhance intelligence.
• Course focus: evolutionary computation + machine learning (EC=optimization/search, ML=prediction/classification).