1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
| flowchart LR
AI(["AI"]) --> Tier0["Tier 0: Energy Infrastructure"]
AI --> Tier1["Tier 1: Chips"]
AI --> Tier2["Tier 2: Data Centers"]
AI --> Tier3["Tier 3: Foundation Model Companies"]
AI --> Tier4["Tier 4: Software Infrastructure"]
AI --> Tier5["Tier 5: AI-native Applications & Services"]
Tier0 --> PowerQ["who's powering the power?"]
PowerQ --> Utilities["Utilities"]
PowerQ --> Generators["Generators"]
PowerQ --> GridPlays["Grid Plays"]
Tier0 --> Hanley["Hanley Energy"]
Hanley --> BuildDC["建造数据中心"]
Hanley --> Service24["24/7 紧急服务承包"]
Tier1 --> ArmRace["arm race"]
Tier1 --> FabQ["who's building the fabs?"]
ArmRace --> NVIDIA["NVIDIA"]
ArmRace --> GPU["GPU"]
ArmRace --> AMD["AMD"]
FabQ --> TSMC["TSMC"]
Tier2 --> Details2["details"]
Tier2 --> PlumbingQ["what about the plumbing, roofing, wiring, insulation?"]
Details2 --> Investment["2025 ¥455B dollars up 51% from Microsoft Meta Amazon"]
PlumbingQ --> NV["Northern Virginia"]
PlumbingQ --> Atlanta["Atlanta"]
PlumbingQ --> Chicago["Chicago"]
PlumbingQ --> Phoenix["Phoenix"]
PlumbingQ --> Dallas["Dallas-Ft. Worth"]
PlumbingQ --> Hillsboro["Hillsboro"]
PlumbingQ --> SV["Silicon Valley"]
PlumbingQ --> NY["New York Tri-State"]
Tier2 --> HVACQ["who's doing the HVAC and installation?"]
HVACQ --> Cert["certifications"]
HVACQ --> EngTeams["engineering teams"]
HVACQ --> Acronyms["speak in acronyms"]
Tier2 --> RoofQ["who maintaining the clean roofs?"]
RoofQ --> Detail3["detail"]
Detail3 --> HEPA["HEPA filters that clog"]
Detail3 --> Contractors["BIFG contractors"]
Detail3 --> Cost["$200->$150"]
Tier2 --> Detail4["details"]
Detail4 --> IDCLimit["机房往往是传统IDC，液冷能力有限"]
Tier3 --> BigGame["Big Game"]
BigGame --> OpenAI["OpenAI"]
BigGame --> X["X"]
BigGame --> Google["Google"]
BigGame --> Anthropic["Anthropic"]
Tier4 --> Details4["details"]
Tier4 --> Companies["Companies Like"]
Details4 --> AITools["AI tools"]
Details4 --> API["API"]
Details4 --> Deployments["Deployments"]
Details4 --> Frameworks["Frameworks"]
Companies --> Stripe["Stripe"]
Companies --> Datadog["datadog"]
Companies --> MongoDB["MongoDB"]
Tier5 --> ShinyStuff["Shiny Stuff"]
Tier5 --> Features["典型特征"]
ShinyStuff --> Apps["Apps that replaced cost centers drive real productivity"]
ShinyStuff --> BusinessModel["most people playing here building without a business model"]
ShinyStuff --> Problems["problems"]
ShinyStuff --> Risk["risk"]
Apps --> Replit["replit"]
Apps --> Cursor["cursor"]
Problems --> Backend["back-end plumbing that make real business"]
Backend --> Scale["actually scale"]
Backend --> DemoBreaks["demo breaks at 3 am"]
Backend --> Payment["connect payment processors"]
Backend --> Databases["connect databases"]
Backend --> APIs["connect APIs"]
Risk --> KeyRisk["key proper risk"]
Risk --> EntireMoat["entire moat"]
KeyRisk --> TechFounder["technical-co-founder"]
KeyRisk --> Vesting["vesting schedules"]
KeyRisk --> Retention["retention bonus"]
TechFounder --> Equity["with equity not just salary"]
EntireMoat --> Engineers["engineers"]
Features --> Revenue["收入主要来自 B 端/C 端 LLM 应用：客服、知识问答、行业 Copilot 等"]
Features --> FineTune["只做少量精调/LoRA，多数时间在跑推理。"]
Features --> Sensitive["对每次调用成本（每1M tokens成本）和回本周期（ROI）极其敏感"]
classDef tier fill:#e1f5fe,stroke:#01579b,stroke-width:3px,font-weight:bold
classDef question fill:#fff9c4,stroke:#fbc02d,stroke-width:2px
classDef company fill:#c8e6c9,stroke:#2e7d32,stroke-width:2px
classDef detail fill:#f3e5f5,stroke:#7b1fa2,stroke-width:2px
class Tier0,Tier1,Tier2,Tier3,Tier4,Tier5 tier
class PowerQ,FabQ,PlumbingQ,HVACQ,RoofQ question
class Hanley,NVIDIA,AMD,TSMC,OpenAI,Google,Anthropic,X,Stripe,Datadog,MongoDB,Replit,Cursor company
class Details2,Details4,Detail3,Detail4 detail
|