Signal to BuildSIGNAL TO BUILD中文Owner sign in ↗Explore curated knowledge, course notes and methods, with their existing content and sources preserved.
This index lists complete English editions. Additional Chinese notes and bilingual updates remain available in the Chinese index.
Ji combines chip-architecture experience with work on AI compilers at HiSilicon. He explains Nvidia through relationships among hardware, software and developers. This is a practitioner’s strategic interpretation from 2024, not independently verified company history.
A synthesis of opportunity forecasting from 3 perspectives. yitang Truman’s opportunity forecasting (Cognitive Framework chapter: 15-character core mantra + 76% probability + 7-layer spectrum + 9-character mindset) + advanced business forecasting (5-dimension industry canvas + CRn concentration + 60-75-85 tier system) / Wang Kai’s AGI perspective (2035-01-01 assumption + do not chase the main AI trend, do supporting plays) / Aschenbrenner’s Situational Awareness (AGI by 2027 / OOMs). Covers P-type vs L-type entrepreneurs, 10 classic misjudgments of business models, 4 types of market evolution, and the 4-quadrant endgame tradeoff framework.
Juntao joined OpusClip in 2023 and worked on payments, commercialization and growth products. Edited from a live presentation and questions, the episode covers acquisition, conversion, retention and insight. Its central idea is to make growth actions opportunities to understand users, beyond finding another traffic source.
Integration of 3 perspectives on overseas promotion + matrix-based operations. Wang Kai (226 posts of go-global practical experience + 6-platform SOP + multi-browser matrix v1.1 + 4-layer isolation infrastructure) / Lao Huang (newtype.pro overseas independent products + individual business experience + 3 things for 2026 + Pro User multi-tool stack) / Lao D (rap.dev domestic one-click publishing + adaptation for use in mainland China without scientific internet access). Covers the divergence between going global vs staying domestic, matrix scale of 50 channels vs intensive cultivation of 1–5 platforms, recommendation traffic vs SEO, and overseas infrastructure checklist.
Roles reverse in this episode: Mo asks and Qu mainly answers. Qu favors participating early while warning against sprinting just after the start. These positions are compatible. Time, attention and learning can receive serious commitment while money, staffing and scale require staged evidence.
An AI can retrieve information or fill out forms without knowing which result matters to you. Tao uses this distinction to explain the move from MindOS to Me.bot: first build AI that understands an individual’s needs, then let it call tools providing services. He also wants AI to keep learning from daily experience rather than depend permanently on its original workflows.
What PostHog is / core concepts / 7 chart types / key metric definition pitfalls (WAU-MAU rolling window, etc.) / one-command MCP integration / billing / ⭐ logged-in state direct investigation SOP (auditing projects built by others). A tool methodology for PMs, decision-oriented
A fusion of Prompt Engineering from 4 perspectives. newtype Lao Huang (186 posts + 4 production-grade templates + 5-element task decomposition + Thinking-Claude performative thinking + Ultra think three tiers + Karpathy Context Engineering) / Wang Kai (6000+ words on PUA-ing Claude + using AI as a ruler for quantification + risk-scoring prompts + decision interrogation SOP) / Lao D (three-layer prompt architecture: meta/system/user + Markdown imitation writing SOP + 4-step blogger imitation) / Anthropic internal (Cal Rueb CLAUDE.md anti-patterns). Covers the counter-consensus view on "whether Prompt Engineering needs to be learned", Skill-ification as the end state, Prompt scheduling Prompt, and Sequential Thinking scheduling.
Integration of RAG and knowledge base engineering from 5 perspectives. newtype Lao Huang (5 generations of architecture evolution: AnythingLLM → Cursor+Obsidian → Claude Code AI OS → newtype OS → wiki.newtype.pro / 5 key points of RAG optimization / Milvus+MCP database-model separation / end state: own content → Obsidian, others' content → NotebookLM) / LLM Interview Technical School (14 special topics: R/A/G three modules + PDF parsing + table recognition + text chunking + HYDE + hard negative samples + RAG-Fusion + GraphRAG) / Lao D's counterintuitive teaching ("RAG does not make the model stronger; it automatically strengthens the prompt") / Wang Kai (Agentic Search outperforms RAG + self-built data MCP) / Cal Rueb (Anthropic does not use RAG MCP). Covers the three-way choice among RAG vs Long Context vs NotebookLM, the economics of GraphRAG, and the philosophy of knowledge bases.
The complete curated notes for Adam Wathan & Steve Schoger's *Refactoring UI* (50+ actionable, copy-ready tips). Core stance: Good-looking UI does not rely on artistic talent, but on a set of concrete rules you can memorize — replace "going by feel" with "checking a checklist". Eight main chapters: ① Starting mindset (start from function not layout · lead with black, white and gray · limit options · define systems in advance, turn every detail you agonize over repeatedly into a rule). ② Visual Hierarchy (make important elements pop: rely not just on font size, but on 3 tiers of font weight + color; de-emphasize secondary content instead of emphasizing it; avoid labels whenever possible; classify buttons into 3 levels: primary/secondary/tertiary, dangerous actions do not need to be big and red; visual hierarchy ≠ HTML document hierarchy). ③ Whitespace and spacing (start with too much whitespace then trim back · spacing/size scales need at least a 25% difference · inter-group spacing > intra-group spacing to prevent ambiguity · grids are overrated, fixed widths are often better · do not scale proportionally · you do not need to fill the screen). ④ Color (use HSL not hex · a complete palette needs 8–10 shades of gray + 1–2 primary colors each with 9 steps + semantic auxiliary colors · predefine color steps 100–900 · grays can carry warm or cool undertones · increase saturation when lightening or darkening · shift hue when lightening to avoid washed-out gray · meet WCAG 4.5:1/3:1 contrast ratios and reverse light/dark direction to avoid oversaturation · do not rely on color alone, pair with icons/arrows). ⑤ Typography (hand-pick a type scale instead of using a formula · line length 45–75 characters · line height moves inversely with width and font size · align to baselines, do not center · left-align long text · links do not all need color · add letter spacing for all caps · use px/rem not em). ⑥ Depth (light comes from above → raised elements are bright on top with shadow below / recessed elements have shadow on top · use two layers of shadows: large soft + small sharp · 5 shadow tiers represent elevation · even flat design can create depth with light/dark contrast, solid-color shadows, and overlapping elements). ⑦ Images (use good photos, do not settle · add overlays/glows to background images to guarantee text contrast · do not enlarge small icons, do not shrink screenshots · use fixed display frames + cover for user-uploaded images). ⑧ Finishing touches (use borders sparingly, replace them with shadows/background colors/spacing · a colored top border instantly elevates polish · refine backgrounds · pay attention to empty states · break out of component conventions). Plus 3 before/after practical teardowns (content page / complex form / data dashboard). Audience: PMs building prototypes; jargon is explained inline, every tip includes Why + do/don't, all numeric values are preserved.
Integration of research methodologies from 3 perspectives. yitang Truman + Zhang Lei OSCAR five-step framework (Objective/Scope/Channel/Analyze/Report) + Research Hacker systematic approach + exploratory arsenal 5 channels + 13 IPO prospectuses + Zhang Lei nursing home 5-sheet empirical study / Wang Kai RSI using research as a ruler (competitive analysis + cross-platform benchmarking + AI interrogation for decision-making) / Wang Huiwen 60-75-85 tier system (research the present / predict the endgame / formulate strategy). Covers research tiers, 5-channel intelligence collection, expert interview SOP, user interview SOP
Lao Zhang of Ai Xuexi, 2026-05-16, single long-form deep dive. X platform For You recommendation system open-source update on May 15 (Phoenix / Grox / Hydrators / Ads Blender) → 7 actionable checklist items for creators. Answers "Why does my title look punchy but my traffic is terrible", "What scores exactly does the recommendation system assign", and "How many content goals should image-and-text creators break their work into". Technical teardown perspective, adding the 6th perspective beyond the 5 author perspectives in wiki-ai-self-media-engineering.
Robert T. Kiyosaki's 1997 classic financial literacy primer. Contrast between two fathers → The rich don't work for money / The middle class is poor because they do not understand the difference between "assets" and "liabilities" / "An asset puts money in your pocket; a liability takes money out of your pocket" / Cashflow game + Rat Race vs. Fast Track / The middle class treating their own home as an asset is a cognitive trap / Four pillars of financial intelligence (financial knowledge + investing + understanding markets + the law) + Six lessons (The rich don't work for money / Financial literacy / Mind your own business / The power of taxes and corporations / The rich invest / Don't work for money) + Five main obstacles to overcome (fear / cynicism / laziness / bad habits / arrogance)
Google Discover appearance mechanism (no special tag, based on interest + helpful content + large image 1200px+ + max-image-preview:large) + traffic is inherently unpredictable. Image SEO (use `<img src>` not CSS / image sitemap / descriptive filename + alt / og:image / primaryImageOfPage SD / SafeSearch / opt-out inline). Video SEO (must have watch page index + dedicated page for single video + Key Moments Clip/SeekToAction SD + Live Badge + allow fetching video files). Core Web Vitals 3 metrics (LCP <2.5s / INP <200ms / CLS <0.1) + Page Experience 6-dimension self-assessment. Avoid intrusive interstitials (use banner instead of modal / let Googlebot bypass mandatory interstitials). Signed Exchange (prefetch speeds up LCP + Web Packager). Site names (WebSite SD + alternateName + domain level). Web Stories (must be valid AMP + 4 required metadata + self-referential canonical + 5 content policies 280 characters / video-first). Flexible Sampling (metering 6-10 articles/month + lead-in shows first few sentences + paywall SD prevents cloaking). Package tracking (India/Japan/Brazil + 700ms API). Byline date (prominently visible + datePublished/dateModified SD). Top Places List (curated, not sponsored). Reviews system (11 languages + evaluates first-party reviews + site-level evaluation if volume is large).
Integration of ROI decision systems from 3 perspectives. yitang Truman's complete ROI system (5-lesson series covering cognition/breadth/depth/height/implementation training + the 4-character mnemonic "list, deduce, build, check" + X-type vs Y-type + ROI triangle + 13,000+ student assignments) / Wang Kai's RSI AI decision-making (viewing from the AGI perspective + using AI as a measuring stick + 5 core mindsets + absolute rational thinking as default) / Wang Kai's investment strategy (three layers of good asset + good price + good volatility + 6,000-word intrinsic value Prompt). Covers broad ROI vs narrow ROI, 3 major inertia traps in decision-making, AI interrogating decision-makers, and the mindset that ROI is a measuring stick
Read Ruan Yifeng's views on AI organized by topic: the debate over open weights vs. open source, model usage costs, career and product judgment, and verification responsibility after AI generates code. Distinguish the author's opinions from verified facts.
Find tools and reading leads from issues 394–413 of Ruan Yifeng's Science and Technology Enthusiast Weekly by task, with chapters and links to the author's original text; distinguish historical inclusion, missing issues, and actual adoption verification. This is not a latest-tools ranking.
Distilled notes from Sajid's UI video tutorial (from the Will's Design Notes series), covering 7 topics across 8 sections: ① Only newbies obsess over building beautiful websites (design = solving problems, not art; check top sites on Semrush, they are all simple or even boring; five pillars = layout/typography/color/images/content) ② Originality is overrated (flat design/skeuomorphism/hamburger menu/pull-to-refresh/like button/Google/Facebook/Uber/Airbnb are all remixes; innovation lies in execution, not first invention) ③ Typography accounts for 80% of UI (remove typography styling and the interface falls apart; you only need 3 font sizes + font weights + 3 levels of lightness; use em for line-height to get natural breathing room) ④ Five rules for designing top-tier websites (design less/law of similarity and law of proximity/add more spacing then remove it/design systems/hierarchy is everything) ⑤ Fix boring UI (layer same-hue colors with different lightness + shadows; this gives the highest return from ordinary to good) ⑥ Spacing selection method (anchor font size with rem, tight inside groups and loose outside groups, button vertical padding < horizontal padding, reduce from large values downward) ⑦ HSL color palette (only adjust L lightness to create harmonious color scales; light comes from above so top is bright and bottom is dark; dark mode uses 3 lightness layers at 0/5/10%) ⑧ Motion Contract for web animation (translate perceptual motion effects into 6-dimensional parameterized constraints to feed to AI/developers). Sajid's unique arguments that distinguish him from Refactoring UI: three core mantras of 'function > beauty', 'copy mature work > originality', 'typography accounts for 80%' + treating UI as a minimal variable system (4 typography variables + 3 gray scales are enough) + the Motion Contract motion delivery method. Target readers: PMs making prototypes; jargon is explained in-industry, each section comes with Why + numerical values + do/don't. Division of labor with refactoring-ui: that article is a full handbook of 50+ tips, this article highlights Sajid's core mantras and positions plus his two exclusive tricks of motion/lightness layering; overlapping details on spacing/color/typography point to refactoring-ui and are not rewritten.
Integration of Scaling Law + AGI + Software 3.0 from 4 perspectives. Karpathy's June 2025 talk (Software 1.0/2.0/3.0 triad + LLM = OS/Utility/Fab three analogies + LLM psychology + Iron Man Suit 4 principles + Agentic Decade 2025–2035) / Aschenbrenner's 165-page Situational Awareness (AGI by 2027 + OOMs counting + Trillion-dollar cluster + Intelligence Explosion) / Three-school debate (Zhu Xiaohu application layer / Cohere Aidan pragmatic / Kai-Fu Lee ceiling / Jensen Huang synthesis) / Wang Kai "Viewing the present from an AGI perspective" + Embiricos AGI bottleneck = human typing speed. Covers AGI timeline disagreements, Software 3.0 adoption forecasts, Vibe Coding natural language programming
Full overview of Safe Browsing + Security Issues report. Strict definitions of Malware vs Unwanted Software + 5 major behavioral prohibitions. 4-step fix for Social Engineering (phishing/deceptive ads/unclearly labeled third parties) + screenshots of 3 typical violations. Full-role checklist for preventing malware. Repeat Offender 30-day no-appeal policy. 7 major methods to prevent UGC spam (reputation / noindex for new users / nofollow+ugc / manual review / blocklist / reCAPTCHA / monitoring). Usage of 4 major search operators (site/filetype/imagesize/src) + limitations (inaccurate; prioritize URL Inspection for debugging). Removals tool: temporary removals for ~6 months vs permanent removal via page deletion/password protection/noindex. 5 major sources of redacted information leaks (cropping illusions/OCR/document history/metadata/URL filenames) + 5-step remediation after improper redaction.