Skip to content

AI Industry Knowledge

Part 1 Mission

Master the entire AI industry — History / Technology / Supply Chain / Business / Applications / Geopolitics — 6 major blocks. Pure understanding before any investment intent. You can systematically discuss with anyone how each link of the AI industry formed, who occupies what position, and who is betting on what.

Part 1 Backbone — Causal Chain

History → Technological Breakthrough → Who Occupies What → Current Industry Structure → Application Realization → Geopolitical Wildcard 10 chapters unfold along this causal chain. Each chapter explicitly returns to the previous one, and the next chapter adds only 1 new dimension.

10 Chapter Map (Study in Order)

# Title Core One-Liner What You Can Do After Est. Time
C1 AI's 70-Year Brief History 4 winters, 4 revivals — "This time is different" needs proof Explain 5 eras + winters + why this time is different 45 min
C2 Transformer Revolution + Scaling Laws 1 eight-page paper changed a trillion-dollar industry Explain why 2017 is the true turning point + why capital dared to invest 45 min
C3 Why NVDA Is Not Intel History's biggest miss was a strategic error, not a technical one Explain NVDA's 20-year positioning + Intel's 4 misses 30 min
C4 Neural Networks / LLM Intuition (No Math) Can't invest in AI industry without understanding LLMs Use 3 analogies to explain LLMs, answer token / context / parameters 1 hr
C5 Why GPU/HBM/Liquid Cooling/Nuclear Power Every piece of hardware is one bottleneck of LLMs Reverse-engineer the entire hardware stack from LLMs 1 hr
C6 Supply Chain 5 Roles + 60 Tickers Mapping the supply chain = seeing how money flows Place any ticker into 1 of 5 roles, explain why 1 hr
C7 Business Models + Value Capture Profit varies per link in the supply chain — moat strength determines survival Use a 5-dimension score to judge any AI ticker's value capture 1 hr
C8 What AI Applications Look Like Today Applications determine the supply chain's next 3 years See revenue reality of 4 major application layers + ROI realization signals 1 hr
C9 US-China + Export Controls + Energy Geopolitics is the AI industry's biggest wildcard Add 3 geopolitical dimensions to your thesis 45 min
C10 5 Real Cases to Build Intuition Real cases are 10x faster than theory Use the 9-chapter framework to analyze 5 real events 30 min

Total Time: ~8-10 hours of focused reading (1-2 weekends).

Week 1

History + Technology Positioning

Build the story first — you'll know why NVDA / OpenAI each occupy their positions.

  • C1 · AI's 70-Year Brief History
  • C2 · Transformer Revolution
  • C3 · Why NVDA Is Not Intel

Week 2

Technology → Hardware → Company Map

LLM intuition → reverse-engineer hardware → supply chain down to specific companies.

  • C4 · Neural Networks / LLM Intuition
  • C5 · Hardware Stack (GPU/HBM/Liquid Cooling/Nuclear Power)
  • C6 · Supply Chain 5 Roles + 60 Tickers

Week 3

Business + Applications + Geopolitics + Cases

Who makes money + ROI realization + Geopolitical wildcard + Real-world intuition.

  • C7 · Business Models + Value Capture
  • C8 · What AI Applications Look Like Today
  • C9 · US-China + Export Controls + Energy
  • C10 · 5 Real Cases

Next

Enter Part 2

After completing Part 1's 5-item self-check → AI Investing.

→ Part 2 Journey

What You Can Do After Part 1 (Self-check)

  • Talk to a friend with zero background about the AI industry for 30 minutes, and they can follow
  • Hear "HBM shortage" / "Stargate" / "DeepSeek" and immediately know which link is affected / which tickers move
  • Place any AI ticker into 1 of 5 supply chain roles, explain why it's there, estimate value capture
  • List 5 items on the "This time is different" proof checklist, monitor if they still hold
  • Add 3 geopolitical dimensions to your thesis (US-China / Energy / Capital)

All 5 ✓ → Enter AI Investing. Any ✗ → Return to the corresponding chapter to fill the gap.

Part 1 to Part 2 Connection

Part 1 Gives You Part 2 Teaches You
Cognitive map — How each link of the AI industry formed Investment judgment — Turn cognition into view + supports + risks
Know who occupies what + who makes money + who has a moat Know how to value + multi-perspective sanity checks + write your own thesis
Know why this time is different (5-item proof) Know how to monitor if the 5 items still hold (weekly/monthly review)

Part 2 is split into two halves:

  • Universal investment model (usable for any industry): Valuation / Mental Models / Historical Comparables / Portfolio / Behavioral Finance
  • AI industry-specific analysis (unique to this industry): Thesis 4 Dimensions / KPI Predictions / Multi-PM Perspectives / Case Studies / Ongoing Review

Part 1 Is the Foundation — Part 2 Depends Entirely on It

Skipping Part 1 and going straight to Part 2 = won't understand. Part 2 assumes you know why NVDA has 75% gross margin / why OAI has a burn rate / why hyperscalers are spending $725B. All of this is in Part 1.