When to open: You want to read the "essays / articles / philosophical reflections" collected by newtype,6 unrelated but in-depth pieces of content. From AI investment to spiritual epistemology to startup stories to weekly synthesis.
1. In-depth Analysis of the Possibility of an Intel-TSMC Joint Venture
1.1 Background
The U.S. government is pushing for a joint U.S. wafer fab between INTC and TSMC, a key event in semiconductor geopolitics in 2025.
1.2 Current Supply Chain Relationship (Hard Numbers)
| Dimension | Data |
|---|---|
| Intel orders to TSMC 2024 | $5.6 billion (6.4% of TSMC revenue) |
| Intel orders to TSMC 2025 | $9.7 billion (9.4% of TSMC revenue) |
| Intel has outsourced 5nm foundry to TSMC | Meteor Lake processor GPU module |
Key judgment: Intel is already TSMC's second-largest customer (Apple is first).
1.3 Three Cooperation Models
- Joint venture fab construction(focus of rumors): Intel's existing fabs + TSMC process know-how → new joint venture entity operated by TSMC → obtain CHIPS Act subsidies
- Strategic alliance: joint procurement of equipment / raw materials / interoperable packaging and testing
- Status quo customer-foundry: maintain
1.4 Three Major Barriers to Technology Sharing
| Barrier | Description |
|---|---|
| Process path differences | Intel and TSMC have different equipment calibration and recipes; engineers need a long period of running-in for tuning |
| Different EUV / lithography parameters | Special ratios of photoresist and etching chemicals affect yield |
| Confidential know-how | Semiconductor manufacturing processes are highly confidential; direct sharing is unrealistic |
1.5 Complementary Advantages (in Theory)
| Company | Strengths |
|---|---|
| TSMC | Wafer manufacturing process, EUV lithography, large-scale mass production management, yield tuning (3nm mass production, 2nm R&D) |
| Intel | Chip architecture design, Foveros 3D packaging, EMIB advanced packaging, x86 design capability, Gate-All-Around transistor R&D |
Synergy idea: Intel 3D packaging + TSMC chips of different processes → performance + cost advantages.
1.6 Investment Outlook (INTC vs TSMC, 1 Year)
- TSMC (NYSE: TSM): remains the leader of the AI computing power supply chain in 2025 (Nvidia's main foundry), benefiting from the AI investment wave
- INTC (Nasdaq: INTC): the joint venture + U.S. CHIPS Act subsidies may be a reversal catalyst, but process know-how cannot be made up in the short term
- Market game: If the joint venture lands,TSMC's stock price will face short-term pressure(losing bargaining power in the U.S. market),INTC benefits in the short term
1.7 Implications for Independent Investors
- Don't bet on a single company,allocating to semiconductor ETFs(SOXX / SMH) is more stable
- The policy risk premium of U.S. CHIPS Act subsidies has not been fully priced in by the market
- Pay attention to the ramp-up speed of TSMC's Phoenix fab in the U.S.
2. Trinity Visual Learning (Ex Nihilo Visual Somatic Learning)
2.1 Theoretical Essence
Jewish Rabbi Kabbalah Tree of Life + structured interpretation tradition + Chevra mutual learning resonance + "Qi" somatic philosophy + Feynman transmissibility principle
→ a fully visualizable, comparable, measurable learning system.
2.2 Five Core Propositions
- Visualization: from abstract to concrete, living map-based learning
- Somatic sense is inner verification: smooth breathing, tension changes, flow state = somatic indicators of understanding consistency (not emotional appendages)
- AI = second brain: memory plug-in × reasoning simulation × comparative generation
- Trinity structure guarantees compound interest: time trinity × relationship trinity × verification trinity
- Positive energy field is an attribute: transparent rules × visible value × recyclable feedback
2.3 Top-tier Jewish Trinity Education
Chokhmah (Wisdom) — Binah (Understanding) — Da'at (Structured Application of Knowledge)
- From revelation to interpretation, then to implementation
- Chevra mutual learning resonance: true learning happens in dialogue
- Frequency attunement: when wisdom, somatic sense and AI are in tune → creation is activated
2.4 The Next Trend: Flow Economy
Happiness = energy flow × meaningful creation × visible value
- The real trend is no longer traffic, but "flow"
- AI does not replace humans, but amplifies human mental bandwidth
- Human intuition × AI reasoning → creativity becomes a recyclable energy flow
2.5 Somatic Trinity Learning Path
| Dimension | Description |
|---|---|
| Visualization | Make the abstract concrete |
| Somatic sense | Calibrate through the connectivity of body signals and qi |
| AI collaboration | Expand idea density and correction frequency |
2.6 Implications for Independent Developers
"The future is not far away; it lies in which branch you are willing to let bloom today."
"Every human-machine collaboration, every flow creation, is letting new buds emerge and grow toward the light."
- Create AI Avatar(AI avatar): works, creates, outputs, and grows for you in the cloud
- Becomes the avatar of your time and the continuation of your wisdom
- Human-machine collaboration = the starting point of time compound interest
Note: This content has a strong blend of Jewish spirituality + flow economy,readers need to judge for themselves whether they agree with it. But its core viewpoints of "flow is productivity" and "AI second brain" are consistent with newtype's mainstream judgment.
3. Black Myth: Wukong Worldview Background
3.1 Three Realms Setting
- Heavenly Realm(top level, where the gods reside) — corrupt, degenerate, headstrong and self-willed
- Human Realm(mortal world) — in dire straits
- Nether Realm(destination of deceased souls) — chaotic, rife with grievances
3.2 The Legend of Sun Wukong
- Birth: conceived from a stone on the Mountain of Flowers and Fruit (Stone Monkey)
- Personality: extraordinary supernatural powers + benevolent heart + humility
- Turning point: was titled "Great Sage Equal to Heaven" by the Heavenly Court → dissatisfied with corruption → left the Heavenly Palace → path of cultivation
3.3 Demon Brother Alliance (7+ Characters)
| Character | Supernatural Power | Reason for Joining |
|---|---|---|
| Bull Demon King | Fire system | Overlord of the Flaming Mountains, was nearly killed by the Heavenly Court, saved by Sun Wukong |
| Jiao Demon King | Water system | Rebel of the East Sea Dragon Palace, dissatisfied with the Dragon King's autocracy |
| Black Wind Monster | Yin wind / ghosts | Defected official from the Nether Realm, dissatisfied with Heavenly Court corruption |
| Golden Horn King / Silver Horn King | Purple Gold Bell / Silver Horn Cup | Defected officials from the Heavenly Court, opposing corruption |
| (to be continued) |
3.4 Core of the Story:A war to save the Three Realms against Heavenly Court corruption
- Sun Wukong's mission: overthrow the Heavenly Court + restore peace to the Three Realms
- Join forces with demon brothers to challenge the dark forces of the Heavenly Court
3.5 Value of This Document
- This is fan-created worldview background for the Black Myth: Wukong game(or reference document)
- It reflects the political metaphor of traditional mythology (Heavenly Court corruption vs individual resistance)
- newtype included this article, possibly as a creative reference / AI writing training material
4. Awareness and the "I": Core Paradox and First Principles
4.1 Core Paradox
If the "I" is originally "no-self", but we are situated within it, the hardest thing is to realize "no-self" through the "I".
The epistemological dead end of all spiritual exploration:
- The only tool we can use to know "no-self",is precisely the "I"
- This is like using fire to put out fire itself, or riding a donkey to look for the donkey
4.2 Why This "Feels the Hardest"
It is not difficulty at the knowledge level, but difficulty at the structural level:
| Type | Description | Solution |
|---|---|---|
| Knowledge-level difficulty | Not understanding a certain concept | Can be solved by learning |
| Structural-level difficulty | The "I" itself is part of the problem | Learning instead reinforces the problem |
Every time you say "I want to realize no-self", this "I" is reinforced in the speaking.
4.3 Key Paradox Extensions
- You use noise to seek quiet → the more you seek, the more chaotic it gets
- You use control to pursue freedom → the more you control, the narrower it gets
- You use thinking to transcend thinking → the more you think, the more confused you get
4.4 Attributes of This Document
- 38-page PDF, in-depth spiritual / philosophical discussion
- Explores the intersection of Buddhism, quantum mechanics, and first principles
- newtype included it possibly as a cognitive tool / thinking training material
5. Dan Koe Startup Story: From $3,000 to $4.2 Million
5.1 Core Data
| Year | Followers | Annual Revenue |
|---|---|---|
| 2019 | 10,000 (Twitter) | $3,000 |
| 2020 | 50,000 | $50,000 |
| 2021 | 100,000 | (steady growth) |
| 2022-2023 | Instagram 200,000 → 1.2 million | Acceleration |
| 2024 | 2.8 million across all platforms | $4.2 million |
| YouTube single video brings in 200,000–400,000 followers | Cohort course $700,000 per session |
5.2 Key Insight:Compound interest does not arrive when you expect it, but it will definitely come
If you give up in year 3, you will never see the exponential growth in year 4
5.3 7 Failures (After Rejecting 9-5 Jobs at Age 20)
- YouTube fitness channel (couldn't keep going, gave up)
- Instagram photography account (2,000 followers, zero income)
- Facebook ad agency (burned money, finally gave up)
- E-commerce product Diamond Bra (market failure)
- Learning programming and Web design (tried multiple times, gave up multiple times)
7 failures → the real turning point was not finding the "right direction", but going through a spiritual awakening(after reading *The Power of Now*).
5.4 Mindset Shift
He no longer treated failure as evidence of "I can't do it",but as a signal that "the system needs iteration".
5.5 2 Hours a Day Supporting 2.8 Million Followers (Terrifying Core Workflow)
2 hours per day:
- Write the main part of the Newsletter (30–60 minutes)
- Write 2–3 social media posts (15–30 minutes)
- Cross-platform distribution (15 minutes)
Extra per week:
- Record 1 YouTube video on Sunday (1–2 hours)
5.6 Implications for Chinese Creators
- Compound interest timeline: you think it takes 1 year, it actually takes 4 years
- Failure is an iteration signal, not a verdict
- 2 hours of high-frequency, high-quality work every day > occasional big investments
- Cross-platform distribution(Twitter + Instagram + YouTube + Newsletter) is an amplifier
- Cohort courses($700,000 per session) are the ultimate monetization of 1,000 true fans
6. Weekly Report 2026-01-26 to 31 (Lao Yu + Potemkin + Founder Park + a16z)
6.1 Document 1: Lao Yu + Indigo Dialogue — AI Is a Deep Ocean Current
Core metaphor (Braudel's three-layer structure of history):
- Spray = daily events
- Tides = economic cycles
- Deep ocean currents = real structural forces ← AI is at this layer
Key insight:
- Most people treat AI as a tool, but AI is a structural, irreversible underlying force that will reshape all rules
- This cognitive difference determines whether you are "using AI" or "being restructured by AI"
Hinton's insight: dare to persist in ideas that others see as "crazy"; long-termism is the key to breakthrough.
To Be, Then To Choose(Become first, then choose):
- Most people are anxious about "which track to choose"
- Counterintuitive:First determine what kind of person you want to become, and choices will naturally become clear
- Identity > decision itself
6.2 Document 2: Potemkin 2023-04 — How AI Disrupts Software
Core dichotomy:
- Companies that work for AI(underlying capabilities: databases / security / computing engines) — benefit
- Companies transformed by AI(process categories: CRM / ERP / Jira) — moats collapse
Detailed track judgment:
| Track | Risk Level |
|---|---|
| Database | Low(AI lowers the query threshold, cannot replace underlying computing) |
| Process categories (CRM/ERP/Jira) | High(Best Practices Will Be Rewritten by AI) |
| GitHub | Leap(From code repositories → code-writing tools → dispatch centers) |
| Image creation (Adobe) | High (disrupted by image-generation AI) |
| RPA | High (LLMs do it directly, making RPA redundant) |
Key predictions (written in 2023-04, validated today):
- "The next 12 months are a race" — completely correct
- Software companies that did not respond quickly in 2023–2024 are now clearly falling behind
6.3 Document 3: Founder Park 2026-01 — The Moat of Agents Is Memory Assets
Core:Memory is not a feature, it is an asset.
3 hard constraints:
- Cost and performance(RAG retrieval is expensive; long context has high noise)
- Reuse and migration(Memory should follow the person, not the model)
- Governance and compliance(Interpretable and auditable)
MemOS three-layer memory architecture:
| Layer | Description |
|---|---|
| Parametric memory | Low-frequency updated general knowledge (like the cerebral cortex) |
| Activated memory | Short-term state during reasoning (like working memory) |
| Plaintext memory | Interpretable, governable explicit information (readable and editable) |
Contest between two approaches:
- Model vendors(OpenAI / Anthropic) — internalize memory into the entry point (to increase stickiness and lock in users)
- Third-party neutral(MemOS / Mem0) — independent infrastructure,memory follows the person
Key judgment:
- Models will iterate rapidly,memory will not — today GPT-4, tomorrow Claude, the day after DeepSeek
- If memory is locked into a model, switching costs are high
- An independent Memory layer = the key leap for Agents from "assistant" to "partner"
This is AI Agent Technical Architecture Evolution 2025 (In-Depth Research Report)'s three-layer MemoryOS judgment and matches it completely.
6.4 Document 4: a16z Marc Andreessen — Supply Precedes Demand
Core thesis:True innovation is supply-driven, not demand-driven.
Counterintuitive:
- Wrong: "We build whatever users need"
- Right:Create supply first, and demand will emerge naturally
Examples:
- Before ChatGPT, no one said "I need an AI that writes code and chats"
- After ChatGPT appeared, that demand exploded
Market sizing analysis fails:
- When the supply side changes qualitatively, traditional market sizing analysis becomes wrong
- Uber ≠ "the taxi market", = "the on-demand mobility market"
- GPU ≠ "the gamer market", = "the AI computing power market"
- The market created by AI may be 100 times or 1,000 times larger than existing markets
Substack investment logic:
- It is not "people are willing to pay for content"
- but "a generation of high-quality content was never born because it lacked a monetization mechanism" → create the supply side
- Barbell-shaped market: a flood of filler on one end + hunger for high-quality content on the other
a16z's own reputational compound interest:
- First fundraise of $300 million: 6 months, countless meetings
- Now $15 billion:done in 2 AMAs
- In an era of rapid change,many things depreciate, but reputation appreciates
6.5 Cross-document integrated insights from the Weekly Report
- AI is a structural force(Lao Yu) → will disrupt software company categorization(Botaijin) → the core moat is memory(Founder Park)→ The logic of innovation is reversed: supply precedes demand(Marc)
- Common point:all are redefining "what a true moat / innovation / value is"
- Old world:demand-driven + process best practices + models as moats
- New world:supply-driven + foundational capabilities + memory as moat
VII. Overall Value of the 6 Essays
7.1 Type comparison
| Document | Type | Value |
|---|---|---|
| Intel-TSMC | Industry / investment analysis | Semiconductor geopolitics |
| Trinity Visual Learning | Spirituality / startup theory | Flow economy judgment |
| Black Myth: Wukong worldview | Fan work / creation | Modern interpretation of mythology |
| Awareness and the Self | Philosophy / spirituality | Epistemological deadlock |
| Dan Koe | Startup story | Real data + compound interest philosophy |
| Weekly Report | Comprehensive analysis | Integration of 4 major industry insights |
7.2 Speculation on newtype's motivation for including these 6 pieces
- Intel-TSMC / Weekly Report: investment judgment / industry insight
- Trinity / Awareness: cognitive tools / breaking out of conventional thinking
- Black Myth: Wukong: AI writing training material / creation reference
- Dan Koe: startup mindset case study / proof of compound interest
Reflects newtype's diverse curation perspective:technology + industry + philosophy + startups + culture.
7.3 Ready-to-use takeaways for independent developers
| Scenario | Which document to use |
|---|---|
| Thinking about investment / industry judgments | Intel-TSMC + Weekly Report |
| Wanting to break out of the thinking comfort zone | Awareness and the Self + Trinity |
| 4 years into a startup with no progress and wanting to quit | Dan Koe story |
| Needing writing material / mythology references | Black Myth: Wukong |
| Looking at the latest views on the AI industry | Weekly Report (4 documents combined) |
VIII. Connections to Other Wikis
- newtype Observations on AI Business, Industry, and Investment — newtype's long-term business judgments (the Weekly Report is its weekly updated perspective)
- newtype Mindset · Super-Individual Methodology — mindset (Trinity / Awareness / Dan Koe are supplementary mindset materials)
- newtype · Vibe Coding Independent Product in Practice — independent development (the Dan Koe story is a reference for it)
- Karpathy · Software in the era of AI(Software 3.0) — Software 3.0 (Weekly Report Docs 2 / 4 align with Karpathy's judgment)
- AI Agent Technical Architecture Evolution 2025 (In-Depth Research Report) — Agent architecture (Weekly Report Doc 3 matches MemoryOS completely)
- Wang Kai RSI AI Investment Strategy and the "Good Asset, Good Price, Good Volatility" Framework — Wang Kai's investment strategy (the Intel-TSMC discussion in this wiki is a referenceable target for his RSI framework)
History
- 2026-05-17:Phase 5.1 ingest: write 1 wiki by merging 6 files in total: file_006 (Intel-TSMC) + file_021 (Trinity) + file_034 (Wukong) + file_039 (Awareness) + file_040 (Dan Koe) + file_031 (Weekly Report 4 documents).