The Rise of DeepSeek How a One Year Old Chinese Company Disrupted the AI Landscape (Steve Russell)（Z-Library）

【One-Line Pitch】 A brisk, narrative account of how Liang Wenfeng's Chengdu startup turned a cheap, open-source model into a global shockwave—best for readers who want the business and geopolitical story of DeepSeek rather than a technical manual. 【Book Arc】 - **Opening (~0%–6%)**: Sets up the premise—DeepSeek as the "unexpected guest" that rewrote AI's rules—and previews the five-chapter structure from founding vision to regulation. - **Early (~6%–25%)**: Covers Liang Wenfeng's quant-fund background, his frustration with incumbent AI economics, the May 2023 founding, the deliberately small Chengdu team, and the philosophy of democratizing AI. - **Early–Middle (~25%–44%)**: The R1 unveiling in Chengdu: the live demo, the $0.55-per-million-token pricing against OpenAI's $15, the open-source release, and the immediate market and media reaction. - **Middle (~44%–63%)**: The "DeepSeek Effect" on markets and business—the so-called DeepSeek Dip, sell-offs in Nvidia and ASML, and how cost efficiency pulled small firms, universities, and independent developers into AI adoption. - **Late (~63%–69%+)**: The competitive crossroads—how OpenAI, Google, and Meta were forced to revisit pricing, timelines, and strategy—plus the geopolitical and regulatory questions the excerpts begin to raise. 【Key Takeaways】 - **The founding thesis was economic, not just technical** (Early): Liang's frustration as a quant investor with expensive, closed AI systems shaped DeepSeek's mission to make cutting-edge models broadly accessible. - **Small and efficient was a deliberate strategy** (Early): Rather than matching giants' sprawling research divisions, DeepSeek recruited a compact team and chose Chengdu partly to signal independence from established tech hubs. - **R1's headline weapon was price** (Early–Middle): At roughly $0.55 per million tokens versus OpenAI's reported $15, the model reframed affordability as a competitive weapon, not a compromise. - **Open source was a calculated gamble** (Early–Middle): Releasing R1's architecture invited copycat risk but bought global developer adoption and rapid ecosystem experimentation. - **The launch became a market event** (Middle): The "DeepSeek Dip" saw Nvidia and ASML shares fall and investors openly question U.S. technological leadership. - **Adoption spread beyond big tech** (Middle): Startups, university researchers, and solo developers—previously priced out—became the fastest-growing users of high-end AI. - **Incumbents were forced to respond** (Middle–Late): Competitors cut prices and accelerated roadmaps, validating DeepSeek's claim that it had reset industry assumptions. - **The story is as geopolitical as it is commercial** (Middle–Late): Export controls, China–U.S. rivalry, and regulatory debates frame DeepSeek as a shift in the global balance of AI power. 【Reading Tips】 - Read the opening and Chapter 1 closely for the founder's motivation; this is the book's strongest narrative thread and explains every later decision. - Skim the repeated pricing comparisons ($0.55 vs. $15) once you've absorbed them—they recur across chapters as rhetorical anchors rather than new information. - Treat the market-reaction sections as journalism, not analysis: note the dates and figures, but expect narrative color over rigorous data. - Read the late competitive and regulatory chapters as open questions rather than conclusions; the excerpts suggest the book frames them as unresolved debates. 【Coverage Limits】 The excerpts cover the book's framing, founding story, R1 launch, and market impact, but thin out in the later competitive and regulatory chapters. Technical details of R1's architecture, training methods, and benchmarks are not covered in the available material.