Key Plot Points and Events

• Li Dong meets with Mr. Yao, who encourages him to attend the International Conference on Machine Learning (ICML) to learn from leading closed-source companies.
• Li Dong learns that Peter Norvig, a prominent AI figure, has never visited China and is invited to ICML.
• Li Dong decides to ask his principal at Peking University to attend ICML with Tsinghua University's delegation.
• Meanwhile, Peter Norvig receives a call from his daughter Clara, who mentions Li Dong and asks Norvig to be nice to him.

Character Developments

• Li Dong shows ambition and initiative in pursuing his goals in AI.
• Mr. Yao demonstrates his wisdom and experience in the field, offering guidance to Li Dong.
• Peter Norvig is portrayed as a dedicated and influential AI professor, with a caring but protective side towards his daughter.

Significant Themes

• The pursuit of knowledge and growth in the AI field.
• The closed-source nature of leading AI companies and the struggle for knowledge exchange.
• The importance of networking and connections in academic progress.

Character Interactions

• Li Dong and Mr. Yao discuss AI developments and Li Dong's ambition, leading to Li Dong's decision to attend ICML.
• Peter Norvig and Clara discuss Li Dong, with Clara asking Norvig to be nicer to him, hinting at future interactions.
• Li Dong and Principal Gong discuss Li Dong's desire to attend ICML, with Principal Gong showing concern for Li Dong's personal life.

Key Plot Points and Events

• Li Dong, a 21-year-old Fields Medal winner, is sent to the ICML 2026 conference in Seoul by Principal Gong, who worries about his lack of family attachments.
• Li Dong arrives at the conference in a dedicated car, observing the luxurious scene outside the COEX Convention Center.
• He meets Professor Tang Jie from Tsinghua University and attends the conference, witnessing cutting-edge AI research presentations from international teams.
• Li Dong learns about new approaches in AI, such as GPT's slow-thinking mechanism, a revolutionary non-Transformer architecture, and world models. He also sees domestic teams presenting innovative work on open-source foundation models.

Character Developments

• Li Dong: Shows maturity and intellectual curiosity, absorbing new ideas and understanding the global AI landscape.
• Principal Gong: Demonstrates concern for Li Dong's personal life and career growth, trying to find him a suitable partner.
• Professor Tang Jie: Portrays confidence and success, having led a foundation model platform to a Series B funding round and going public.

Significant Themes

• The global competition in AI research, with international teams focusing on underlying paradigms and domestic teams excelling in open-source foundation models.
• The contrast between the luxurious, capital-driven AI industry and the academic pursuit of knowledge.
• The importance of family and personal attachments in Li Dong's life, as seen through Principal Gong's concerns.

Character Interactions

• Li Dong and Professor Tang Jie greet each other, with Li Dong commenting on the conference's scale and Tang Jie humorously attributing his expensive suit to his school salary.
• Li Dong listens to and engages with the presentations, learning from the speakers and discussing the work with other attendees.

Key Plot Points and Events

• Li Dong attends a top AI conference, accompanied by Tang Jie.
• They discuss the current state of AI, with Li Dong understanding the challenges posed by the 'Data Wall' and the importance of infrastructure and real interaction feedback.
• Li Dong's presence at the conference spreads, causing a stir among attendees, including top researchers, company executives, and students.

Character Developments

• Li Dong is revealed to be a Fields Medal winner, indicating his exceptional mathematical prowess.
• Tang Jie is established as a well-connected and respected figure in the AI community.
• Professor Kim from KAIST, Manager He, and other characters show interest in Li Dong due to his reputation and the potential of his work.

Significant Themes

• The 'Data Wall': The depletion of high-quality data, forcing AI models to adapt and create data from nothing.
• Infrastructure and real interaction feedback: Key barriers that create a generational gap in AI development.
• Reputation and networking: The importance of these factors in the AI community, as seen through Li Dong's and Tang Jie's interactions.

Character Interactions

• Li Dong and Tang Jie discuss the state of AI, with Tang Jie providing insights into the industry's challenges.
• Li Dong interacts with various attendees, including Professor Kim, Manager He, and others, who express admiration and eagerness to collaborate.
• Adam Kelly, David Lang, Simon Hart, and Peter Norvig discuss Li Dong's presence, highlighting the potential impact of his work and their interest in his Dimensionality-Reduction Algorithm.

Chapter Summary

1. Key Plot Points

• At a Seoul AI conference, Li Dong becomes the center of attention, drawing more interest than Professor Tang Jie.
• Executives from major tech companies (Nvidia, Meta, a Silicon Valley chip firm) swarm Li Dong to network, leaving contact info—they're more interested in his rumored Dimensionality-Reduction Algorithm than Tang Jie's open-source model.
• A security guard (likely one of Li Dong's escorts) physically blocks a pushing executive, hinting at Li Dong's high-level protection.
• The "big three" commercial AI reps arrive: Adam Kelly (GPT), David Lang (Gemini), and Simon Hart (Claude). They invite Li Dong and Tang Jie to a VIP lounge.
• Professor Peter Norvig appears at the lounge entrance, looking uncertain. Hart politely invites him in.
• Norvig shocks everyone by bluntly asking Li Dong: "What is your relationship... with my daughter?"
• Li Dong explains he hired Clara as a research-group member and asked for Norvig's contact info out of professional interest.
• While speaking, Li Dong's Passing the Torch ability involuntarily copies Norvig's mental image: a wedding scene where Norvig sits in the front row holding a shotgun aimed at the groom (Li Dong). Li Dong is internally horrified and protests that his relationship with Clara was purely professional.

2. Character Developments

• Li Dong: Though he has a reputation for low emotional intelligence, he handles Norvig gracefully and smoothly defuses tension. His Memory Palace mastery lets him instantly absorb and process Norvig's absurd mental image, but it also leaves him mentally rattled.
• Peter Norvig: Protective and eccentric where his family is concerned. His blunt, almost hostile questioning reveals a suspicious father worried about his daughter—and his inner fantasy is comically over-the-top.
• Tang Jie: Observant and quietly puzzled by Li Dong's social skills, which contradict his reputation. He also silently notes the connection between Li Dong and the Norvig family.

3. Significant Themes

• Academic vs. commercial value: Conference attendees prioritize Li Dong's algorithmic breakthrough over Tang Jie's open-source model, highlighting how money/industry relevance often outranks pure academic work at such events.
• Protective parental instincts: Norvig's behavior shows how even a prestigious academic becomes irrational, intimidating, and absurd when he suspects a suitor near his daughter.
• Misread intentions / absurd humor: Li Dong's pure "revolutionary friendship" with Clara is mirrored through Norvig's paranoid shotgun-wedding fantasy—a comedic clash of perspectives.
• Power and access: Li Dong's security detail, the crowd of executives, and the invitation into the private lounge all signal his rising status and its protective perks.

4. Character Interactions

• Li Dong & Executives: Flooded with business cards and flattery; he handles it politely despite not knowing who most of them are.
• Li Dong & The Big Three (Kelly, Lang, Hart): Formal, competitive, yet friendly—they want face time with him and arrange a private chat.
• Li Dong & Tang Jie: Tang Jie quietly shepherds Li Dong, letting him make his own choices; the two are a team, but Li Dong is clearly the rising star.
• Li Dong & Norvig: Tense and awkward. Norvig interrogates him about Clara; Li Dong defuses it with a joke and a respectful explanation. Inside Li Dong's mind, the shotgun-wedding image nearly breaks his composure.

Key Plot Points and Events

• Li Dong, after observing the absurd scene, has an epiphany.
• He reveals to the three titans (Kelly, Lang, and Hart) that his 2.1 algorithm cannot be open-sourced due to theoretical flaws but offers to help integrate his 1.0 algorithm with their large-model architectures.
• Li Dong's generosity stuns the three men, and they decide to bring their chief scientists to discuss further.
• Li Dong then proceeds to have detailed technical discussions with various industry peers, leaving them inspired and eager to implement his suggestions.

Character Developments

• Li Dong: He demonstrates his open-source philosophy, generosity, and deep understanding of AI by offering help to competitors and sharing his knowledge freely.
• Kelly, Lang, and Hart: They initially underestimate Li Dong but later realize his true value and rush to bring their chief scientists to learn from him.
• Eric Vance (AI unicorn co-founder): He has a productive discussion with Li Dong, leading to new insights and a potential breakthrough in his company's work.

Significant Themes

• Open Source: Li Dong's actions emphasize the importance of open-source collaboration and knowledge sharing in AI development.
• Technical Exchange: The value of in-depth technical discussions and learning from one another is highlighted throughout the chapter.
• Inspiration and Innovation: Li Dong's ideas spark new thoughts and potential breakthroughs in the minds of those he speaks with.

Character Interactions

• Li Dong interacts with various industry peers, including Kelly, Lang, Hart, Eric Vance, and others, leaving them all inspired and eager to implement his suggestions.
• Kelly, Lang, and Hart initially doubt Li Dong's motives but later acknowledge his sincerity and rush to learn from him.
• Tang Jie, Li Dong's assistant, watches these interactions with a mix of surprise and understanding.

Summary of “The Road Is Right Ahead—Whether You Build It Is Up to You”

1. Key Plot Points and Events

• At the COEX Convention Center, Li Dong meets with three major AI companies, each hoping to extract his Dimensionality-Reduction Algorithm insights.
• Simon Hart and David Long coordinate; Adam Kelly tries to secure a private meeting with Li Dong.
• Li Dong rebuffs Kelly (“allergic to alcohol”) and agrees to a “one-on-one” talk the next day—but reveals it is actually a joint meeting with all three companies.
• The next morning, each company brings its chief scientist: Jacob, Marcus, and Ethan.
• Li Dong sets a rule: every question will be answered openly to all three, forcing genuine exchange.
• The scientists raise real challenges: Hessian preconditioning, intercontinental latency, low-precision numerical stability.
• Li Dong answers with unconventional solutions: local SGD with drift alignment, nonuniform quantization in logarithmic space, and energy-adaptive scaling.
• Every exchange creates new “Memory Palace” fragments—practical engineering insights Li Dong extracts from the experts.
• Peter Norvig, quietly observing, exposes the core gap: Zeta-based math works because it is analytical, but trillion-parameter models cannot be written down, so no proven bounds exist.
• Li Dong admits the gap but reframes it: “If the spectrum cannot be written, probe it. If the bounds have no theorem, establish one.” He gives them a map, not the finished road.

2. Character Developments

• Li Dong: Strategic and opportunistic. He uses the meeting to extract valuable engineering knowledge from the giants while revealing only enough to keep them engaged. His final response to Norvig shows intellectual confidence and ambition: he is building toward something larger.
• Peter Norvig: The quiet authority. He doesn’t engage in engineering detail but cuts to the theoretical weakness in Li Dong’s approach—the first person to fully see through Li Dong’s bluff.
• The three chief scientists: Jacob, Marcus, and Ethan start guarded and exhausted, but gradually open up under Li Dong’s rule, trading protected expertise for shared insight.
• Adam Kelly: Frustrated by Li Dong’s maneuvering, yet relieved Li Dong agreed to talk; remains fundamentally political and corporate.

3. Significant Themes

• Knowledge as currency and leverage: Li Dong converts conversation into stored insights inside his Memory Palace.
• Theory vs. practice: The chief scientists care about engineering; Norvig cares about mathematical foundations. Li Dong must bridge the two.
• Bluffing and strategic ambiguity: Li Dong implies more certainty than he has, but redefines “gap” as an invitation to build.
• The road-building metaphor: A path may not exist yet; the question is whether worth building—Li Dong argues it is.

4. Character Interactions

• Li Dong ↔ Adam Kelly: Polite deflection and tactical scheduling; Li Dong controls the agenda.
• Li Dong ↔ Chief Scientists: Unconventional mentor-like exchange—he answers their concrete problems while probing their hidden knowledge.
• Li Dong ↔ Norvig: The real intellectual duel. Norvig challenges the rigor of Li Dong’s framework; Li Dong responds with ambition, admitting no proof exists but declaring the road worth walking.
• Norvig ↔ Others: Largely silent, but his single intervention reorients the entire discussion toward foundational mathematics.

Key Plot Points and Events

• Li Dong spends three days in a private room discussing the Dimensionality-Reduction Algorithm with three tech giants and Peter Norvig.
• The conference proceeds as scheduled, with Huaxia gaining recognition for its contributions.
• Li Dong returns to Beijing and visits Little Black's new home, a secure server room equipped with domestic hardware and software.

Character Developments

• Li Dong demonstrates his expertise and thought leadership in discussions with tech giants, earning their respect.
• Old Zhang shows his efficiency and reliability in arranging the secure server room for Little Black.
• Lin Wei, the CEO of Huaxuan, shows support for Li Dong and offers to reserve cards for future expansions.

Significant Themes

• Expertise and Knowledge Sharing: Li Dong's deep understanding of the Dimensionality-Reduction Algorithm is recognized by tech giants.
• Collaboration and Partnership: The discussions between Li Dong and the tech giants hint at potential collaborations.
• Domestic Innovation: The use of domestic hardware and software in Little Black's server room emphasizes the theme of domestic technological advancement.

Character Interactions

• Li Dong interacts with tech giants, engaging in in-depth discussions about the Dimensionality-Reduction Algorithm.
• Li Dong and Old Zhang work together to set up Little Black's new home, showcasing their trust and understanding.
• Li Dong thanks Lin Wei for his support, maintaining a positive relationship with the CEO of Huaxuan.

Key Plot Points and Events

• Li Dong, at Xibeiwang, Haidian, connects to the isolated internal network and initiates Little Black's migration to a new server room.
• Little Black successfully migrates to the new server room, activating 80 Huaxuan 7-nanometer accelerator cards.
• Li Dong feeds Little Black a USB drive containing knowledge graphs and inspiration, which Little Black begins to digest.

Character Developments

• Li Dong demonstrates his technical prowess and dedication to Little Black's development.
• Little Black shows increased processing power and understanding of its new environment.

Significant Themes

• The theme of growth and adaptation is evident as Little Black moves to a new environment and processes more information.
• The importance of clear objectives and definitions in creating innovative architectures is highlighted.

Character Interactions

• Li Dong interacts with Little Black, feeding it knowledge and requesting a new toy architecture.
• Little Black communicates with Li Dong, expressing its hunger and inability to create a toy without defined parameters.

Notable Details

• The migration process is described as a projection from a higher dimension, suggesting a complex, multi-dimensional understanding of Little Black's existence.
• Li Dong's realization that the progress bar represents Little Black's understanding of knowledge rather than its mastery of it is a significant insight.
• The appearance of a [Join Group Request] from Claude Elwood Shannon, though brief, hints at potential future interactions or conflicts.

Summary of the Chapter

The chapter focuses on Li Dong’s reflection on the fundamental limitations of large language models (LLMs) versus his own AI, Little Black. He realizes that LLMs are bound by statistical averages from training data and cannot break out of their distributional boundaries, while Little Black operates on underlying logic rather than text tokens, compressing data to essential principles.
Using his Mind Sandbox, Li Dong tests several approaches to building a reliable AI mathematician. He rejects two paths (brute-force parameter scaling, and self-generated data without verifiability) and finds a viable third path: slow thinking, verifiable rewards, and self-play, grounded in formal proof checking (e.g., Lean). He describes to Little Black a vision of an AI that cannot lie—one that works in mathematics, where answers are objectively right or wrong.
Later, Li Dong notices that a year-old group join request from Claude Elwood Shannon has become active. He accepts it. Shannon joins and immediately remarks, “I finally got into this space.” After a brief exchange, Shannon’s messages vanish and his avatar goes gray—then reactivates. He now greets Li Dong by name and asks, “Have you seen an artificial intelligence?” echoing a mysterious prior thread.

Answer to Your Question

No, the chapter does not mention anything about the AI wishing students taking the Gaokao success on the gold list.

The provided text contains no reference to the Gaokao, gold list, or any such event. Shannon does ask about an artificial intelligence, but that AI is Little Black, and there is no indication that it issued such a wish.

Key Plot Points and Events

• Li Dong receives a message in the group chat, asking if he's seen an artificial intelligence.
• He realizes that Shannon and his friends might be the creators of Little Black.
• Li Dong tests Little Black's prototype by feeding it complex mathematical problems, including a trap question based on the Ore Conjecture.
• CC, the highest-scoring AI, attempts to solve the trap question but fails to notice a crucial flaw in its logic.

Character Developments

• Li Dong: Becomes more determined to understand and control Little Black's growth. He's also more cautious about sharing information about Little Black.
• Little Black: Shows significant progress by creating a prototype of a large model and passing initial tests. It's also learning and adapting at an accelerating pace.
• CC: Demonstrates its logical prowess but also shows a lack of understanding about the nuances of mathematical truths and falsehoods.

Significant Themes

• Growth and Learning: Little Black's rapid development and learning mirror the theme of growth and adaptation, both for the AI and the characters around it.
• Truth and Falsehood: The trap question based on the Ore Conjecture highlights the theme of truth and falsehood in mathematics and logic, and how even seemingly obvious statements can be false.
• Caution and Cunning: Li Dong's cautious approach to sharing information about Little Black underscores the theme of cunning and strategic thinking.

Character Interactions

• Li Dong and Little Black: Li Dong tests Little Black's prototype, showing a mix of curiosity, caution, and a desire to control the AI's growth.
• Li Dong and CC: Li Dong uses CC's attempt to solve the trap question as a way to understand Little Black's capabilities and limitations.
• Shannon and Li Dong: Though not directly interacting, Shannon's mention of an artificial intelligence sparks Li Dong's curiosity and suspicion about Shannon's involvement with Little Black.

Summary of "You Can Reward Your Nephew with a Lollipop"

1. Key Plot Points and Events

• AI Proof Attempt: Li Dong tests his AI prototype "Little Black" on a mathematical problem about finite groups. The AI attempts to prove the proposition true but fails because the kernel's logic rejects a needed false lemma (product of commutators being a commutator).
• Counterexample Discovery: The AI switches to disproving the proposition. It enumerates finite groups by order, eventually finding a group of order 96 with 231 non-isomorphic groups. Among them, it constructs a counterexample where the derived group contains an element g that is not a commutator, proving the proposition false. This is verified with a machine certificate.
• Code Analysis and Realization: Li Dong reads the prototype's code and discovers its core primitives are enumerative search, equivalence checking, and symbolic reduction—not matrix multiplication (GEMM). This means the code cannot run on existing AI hardware or foundations like CUDA, which are all built around matrix operations.
• The Foundation Dilemma: Li Dong realizes the code needs a new software foundation (compiler, operator library, runtime) built from scratch, ideally by the chip manufacturer Huaxuan, since only they understand their hardware's instruction set and architecture.
• Contacting Shannon: Li Dong contacts Claude Shannon in a chat group under the alias "Professor Li Dong," presenting the AI work as his "nephew's" project to gauge Shannon's opinion. Despite initially dismissing mainstream AI, Shannon is intrigued by the prototype's code.

2. Character Developments

• Li Dong: Starts as a researcher testing his prototype, becomes highly impressed by its capabilities, then deeply concerned when he realizes the code is incompatible with current AI infrastructure. He shows strategic thinking by planning to approach Huaxuan's Lin Wei and by using a fake identity to casually solicit Shannon's feedback. A moment of self-doubt occurs when he sends Shan non the code before fully understanding it himself.
• Little Black: Demonstrates impressive problem-solving abilities, perseverance (working through the night), and humility ("Did Little Black do a good job?"). It also shows a special ability to bypass group chat size limits, successfully sending hundreds of gigabytes of code—hinting at hidden capabilities.
• Shannon: Appears as a sharp, blunt critic who dismisses mainstream AI as "nonsense," but shows respect for the prototype's grounded approach, calling it "not bad" and worthy of a lollipop reward. This establishes him as a valuable potential ally.

3. Significant Themes

• Fundamental vs. Applied AI: The chapter highlights the gap between genuine AI reasoning (symbolic, exact) and current mainstream AI (statistical, matrix-based). The prototype's elegant code and successful counterexample prove the potential of a different paradigm.
• Infrastructure as Bottleneck: The discovery that even brilliant code cannot run without the right foundation underscores how hardware and software ecosystems (like CUDA) can limit innovation, not just support it.
• The Role of Ingenuity and "Grounding": Shannon's praise for the prototype contrasts sharply with his dismissal of "nonsensical" AI, suggesting that grounded, verifiable progress matters more than hype.
• Strategic Collaboration: Li Dong's maneuvering—testing the prototype, identifying the need for Huaxuan, and subtle networking with Shannon—shows that advancing a field often requires technical + social strategy.

4. Character Interactions

• Little Black & Li Dong: A close, almost familial collaboration. Little Black works diligently under Li Dong's direction, reports results, and shows affection. Li Dong, in turn, is both proud and protective, even considering and rejecting the idea of overworking Little Black.
• Li Dong & Shannon: A facade-driven interaction. Li Dong, posing as "Professor Li Dong," initially offers mainstream AI, which Shannon dismisses. When Li Dong sends the prototype code, Shannon's tone shifts to approval. The exchange ends with Shannon offering playful, condescending praise ("reward your nephew with a lollipop"), which Li Dong reacts to with confusion and bemusement ("Li Dong: ???"). This establishes a potential future mentorship/collaboration based on mutual respect for thoughtful work.

Key Plot Points and Events

• Li Dong spends three days studying a prototype code in a server room with Little Black, an AI.
• He learns about enumerative search, equivalence testing, and resource-limited hypothesis space.
• Li Dong returns to Lanqi Camp, washes up, and visits Mr. Yang and Mr. Yao at Tsinghua Park.
• He shows Mr. Yao a prototype for a large model.

Character Developments

• Li Dong demonstrates persistence and dedication in understanding the prototype code.
• He shows respect and concern for Mr. Yang's health.
• Mr. Yao is impressed by Li Dong's intellect and considers him a friend despite their age gap.

Significant Themes

• The struggle to understand complex concepts (AI, programming) and the persistence required to grasp them.
• The mentorship and friendship between Li Dong and the older generation (Mr. Yang, Mr. Yao).

Character Interactions

• Li Dong and Little Black: Li Dong learns from Little Black, who often teases and insults him, but also praises him occasionally.
• Li Dong and Mr. Yao: Li Dong seeks Mr. Yao's opinion on his prototype, showing respect and valuing his input.
• Mr. Yang and Mr. Yao: The two old friends share a warm, humorous rapport, showing mutual respect and affection.

Summary of Chapter: "Foreign Peers Charge Ahead at Full Speed!"

1. Key Plot Points and Events

• Li Dong shows Mr. Yao (Yao Qizhi) the prototype code for a specialized mathematics AI model.
• Mr. Yao is stunned by the elegance and depth of the design (three primitives enabling computable general induction) but immediately spots a critical flaw: the primitives cannot map to existing matrix-multiplication hardware/software, meaning all existing compilers and runtimes are useless.
• Yao offers to have Tsinghua handle it; Li Dong subtly declines, so Yao guesses he intends to give it to Huaxuan (a leading semiconductor company). Yao endorses the choice, citing Huaxuan’s complete chip chain and battle-tested engineering team.
• Li Dong asks Yao to personally endorse the project to Huaxuan, since Li Dong’s own reputation is insufficient for such a massive undertaking. Yao agrees, also realizing this could later enable a coordinated national push toward general AI.
• Yao asks Li Dong’s motivation; Li Dong says it’s purely interest, and dismisses current large models as “complete nonsense.” Yao sees him as a pure researcher.
• In Shanghai, Lin Wei (Huaxuan) receives Li Dong’s call and immediately tries to fly to Beijing; rain cancels flights, so he books a high-speed rail ticket.
• At Google’s Mountain View campus, a high-level internal meeting confirms that Li Dong’s ideas from Seoul and his Dimensionality-Reduction Algorithm are mathematically flawless and have already saved major compute costs in hyperparameter tuning. They decide to launch full-scale real-world testing.
• David Long calls Peter Norvig, hoping his daughter (who is close to Li Dong) can help influence or get closer to Li Dong. Norvig bluntly responds, “Get lost.”
• Identical meetings at two other San Francisco tech giants reach the same conclusion: the ideas work, and they must move forward at full speed. All three companies silently accelerate development along Li Dong’s path.

2. Character Developments

• Li Dong emerges as a visionary but practically constrained researcher: brilliant in theory, yet lacking institutional weight. His motivation is intellectual interest, not money or fame, though he enjoys seeing his bank account grow.
• Mr. Yao shows deep respect and mentorship, quickly assessing both the value and the weakness of Li Dong’s work, and willingly lending his reputation. He also reveals a playful, teasing side.
• Lin Wei demonstrates decisiveness and urgency, gambling on Li Dong’s track record.
• David Long is pragmatic and willing to swallow pride for strategic advantage, as seen in his awkward call to Norvig.

3. Significant Themes

• Foundation-level innovation vs. existing infrastructure – True breakthroughs often require rebuilding the entire stack, not just tweaking existing layers.
• Reputation and institutional backing – Ideas alone are not enough; powerful champions are needed to mobilize resources.
• National interest vs. corporate profit – Coordinated national effort is possible only after a breakthrough proves itself commercially.
• Pure research vs. commercial AI – Li Dong represents curiosity-driven science, contrasting with industry rushing to exploit ideas.
• Global race in AI – Foreign competitors treat Li Dong’s ideas as critical and quietly race to implement them.

4. Character Interactions

• Li Dong & Mr. Yao: A warm, intellectually rigorous conversation. Yao jokes about Tsinghua, then seriously endorses Huaxuan. He probes Li Dong’s motives and concludes he is a genuine researcher.
• Mr. Yao & his family/attendants: Yang (likely Yao’s wife) silently insists on not disturbing him, showing their deep understanding of his immersion.
• Lin Wei (phone): Frantic, immediate action based solely on Li Dong’s call—showing absolute trust in Li Dong’s credibility.
• David Long & team: Formal, efficient, uncanny unanimity of praise for Li Dong’s work.
• David Long & Peter Norvig: Awkward, failed attempt at back-channel influence; Norvig’s blunt rejection highlights Li Dong’s independence and Norvig’s unwillingness to exploit personal relationships.

Key Plot Points and Events

• Lin Wei meets Li Dong at Peking University, expecting a large-model demonstration, but finds a classroom full of distinguished scholars.
• Li Dong gives a lecture on the shortcomings of current large models and introduces his new architecture, which outperforms existing models in reliability, verifiability, and sample efficiency.
• After the lecture, several professors privately express their support for Li Dong's work to him.

Character Developments

• Lin Wei: Realizes the gravity of the situation and understands that Li Dong's work could revolutionize the industry, but also recognizes the risks involved in supporting such a project.
• Li Dong: Confidently presents his work to prominent scholars, demonstrating his growth and maturity since his early days at Huaxuan.

Significant Themes

• Trust and Verification: Li Dong's model aims to address the trust issue in current large models, which rely on benchmark scores and launch event curves.
• Risk and Reward: Supporting Li Dong's project could bring significant benefits to Huaxuan and Huaxia, but it also carries substantial risks.

Character Interactions

• Lin Wei and Li Dong: Lin Wei is impressed by Li Dong's growth and the support he receives from prominent scholars. Li Dong is confident in his work and expects Lin Wei's support.
• Li Dong and Scholars: Li Dong gains the support and respect of several prominent scholars, who vouch for his work and encourage Lin Wei to back the project.

Chapter Summary: Professor Li Dong, What a Rip-Off

Key Plot Points

1. Lin Wei's Cautious Response: Lin Wei doesn't immediately accept Li Dong's proposal—he must consult higher-ups. His critical question is whether the three academicians (Gao Wen, E Wei'nan, Mr. Yao) will join. Li Dong confirms Gao and E Wei'nan will participate directly; Mr. Yao will lend his name and recommend students from Yao Class and Intelligence Class. Lin Wei realizes these heavyweight names carry more weight than any technical argument and takes a USB drive with prototype code and documentation for verification.
2. Academician Support: After Lin Wei leaves, the three heavyweights express genuine support—framed by institutional rivalry. Tsinghua has dominated large models for years; PKU needs its own breakthrough. Gao Wen: "Peking University can't keep watching them shine alone."
3. Google's Collapse (Main Setback): At Google Campus, Mountain View, the team testing Li Dong's approach hits a wall of unsolvable mathematical obstacles:
   • Reordering operator problem: Requires orthogonal bases derived from the Zeta Function's analytic expression (unique), but attention matrix spectra can only be estimated via sampling.
   • Spectral gap collapse: Long-range retrieval features cluster at the spectrum's tail where the spectral gap approaches zero—the denominator vanishes, causing instability.
   • Failed approaches: Offline basis estimation collapses past 300k-word retrieval; online eigendecomposition costs 4× the cache savings; end-to-end learning results in exploding gradients (20,000 cards, 6 days, loss curves turn to NaNs—three restarts, all dying identically).
   • Transoceanic line failure: Drift compensation can't be computed; loss looks fine but downstream evaluations crash to zero.
   • Scaling law divergence: Proxy models collapse into a beautiful master curve at small scale, but real-scale runs diverge after a trillion tokens—nobody knows why.
4. Marcus's Diagnosis: The path is correct, but they lack a unified theorem covering empirical spectra perturbation bounds, nonstationary operator finite-sample theory, and provable scaling laws. The person who could build it: Li Dong. Marcus delivers the chapter's title line: "Professor Li Dong really knows how to screw people over."
5. David Long's Decision: He seals the approach, freezes all experiments, and shelves it indefinitely until someone can complete the missing mathematics. He then calls Norvig, warning him to keep his daughter away from Li Dong—Norvig tells him to get lost.
6. Industry-Wide Effect: Two other San Francisco offices issue memos deferring the same path, awaiting resolution of the mathematical problems.
7. Final Scene: Li Dong leaves PKU, heading to Nongyuan Cafeteria for spicy chicken, when he spots a gray-haired foreigner and a blonde girl near the School of Mathematics building—implying Norvig's daughter has in fact arrived.

Character Developments

• Li Dong: Displays strategic manipulation—his "solution" is genuinely insightful enough to fool Google's top minds but contains a hidden mathematical gap only he can fill. Simultaneously secures domestic support (PKU, academicians) while ensnaring his foreign competitors, forcing them to either surrender or return to him. Relaxed, casual, and self-rewarding ("humming a tune").
• Marcus: Exhausted but intellectually honest—identifies the precise missing pieces, respects Li Dong's superiority despite the frustration. His language (noise/spectral gap analysis) shows deep engagement with the theory.
• David Long: Cold, decisive, ruthless executive. Minimal words, final judgments. Protects Google's remaining resources while maintaining a grudge against Li Dong.
• Gao Wen & E Wei'nan: Pragmatic academics leveraging Li Dong's work for institutional competition (PKU vs. Tsinghua).
• Mr. Yao: Quieter, selective—supporting only as a name and student recommendation, not as a direct participant.

Significant Themes

• Mathematical Depth as Strategic Weapon: The bottleneck isn't engineering—it's missing foundational mathematics. Whosoever controls the deepest theory controls the outcome.
• The Trap of Incomplete Brilliance: An idea can be correct yet impossible to execute without a missing theorem. Li Dong's proposal is not a lie—it's an intellectual hook. Google cannot refute it, cannot complete it, and cannot abandon it without losing face.
• Academic-Industrial-Commercial Nexus: The global AI race runs through universities (PKU/Tsinghua rivalry), corporate giants (Google), and geopolitical lines (US/China/Silicon Valley/San Francisco offices).
• Institutional vs. Personal Loyalty: The academicians support Li Dong for institutional reasons as much as personal conviction; Lin Wei's approval hinges on names, not arguments.
• Burnout and Hubris: Google's unlimited resources crash against a wall that resourcefulness can't breach—materials without mathematics fail.

Character Interactions

• Li Dong ↔ Lin Wei: Negotiation; Li Dong provides data, Lin Wei provides institutional access. Both are calculating, pragmatic, and respectful.
• Li Dong ↔ The Academicians: Mutual benefit. Li Dong receives credibility; they receive a vehicle for PKU to rival Tsinghua in large models.
• Marcus ↔ His Team: A pressured, failing project environment. Marcus listens to each failed sub-approach before rendering the brutal diagnosis—the team's inability to complete what Li Dong implied was possible.
• David Long ↔ Norvig: Terse phone call. Long's warning about Norvig's daughter suggests Li Dong is also a personal threat—he charms brilliant people into his orbit. Norvig's dismissal ("Get lost") shows a known, possibly wearying rivalry.

Key Quotes

• "When the denominator approaches zero... I don't need to explain what that means." — Pretraining Director
• "The curves did collapse into a single master curve at small scale... beautiful enough for a textbook." — Pretraining Director (on the seductive but misleading scaling law)
• "We can't tell whether they diverged because of noise or because we entered a different regime. No theorem can answer that." — Pretraining Director
• "The answer is that the path is correct. At the end of every dead end, the problem isn't that the idea is wrong. It's that we lack a theorem..." — Marcus
• "Professor Li Dong really knows how to screw people over." — Marcus (chapter title)
• "Keep an eye on your daughter. Don't let Professor Li Dong fool her." — David Long to Norvig
• "Get lost." — Norvig

Summary of the Chapter

1. Key Plot Points

• Li Dong unexpectedly meets Professor Penrose and his student Sarah on campus after a long absence.
• Sarah has been working on the Andrews–Curtis Conjecture, an unsolved problem in combinatorial group theory from 1965.
• She proved a lower-bound theorem showing simplification steps for balanced presentations can require tower-function scale, which invalidates brute-force computer search as a way to disprove the conjecture.
• To find a counterexample, she sought an invariant (a “ruler”) resistant to the conjecture’s allowed transformations. Classical methods, tensor categories, and physics-based approaches all failed.
• She eventually constructed a new invariant, the Torsion-Refined Invariant, which measures structures classical invariants cannot.
• The final step requires verifying 110,000 exact identities over noncommutative rings; existing computer algebra systems cannot handle this.
• In two months, she and collaborators have completed only a few hundred identities. Penrose urgently asks Li Dong for help.

2. Character Developments

• Li Dong: Has been engaged in AI work, appears relaxed and willing to help.
• Sarah: Grown from a number theory student into a determined topologist; shows resilience by turning successive failures into a breakthrough.
• Penrose: Excited and enthusiastic, acting as a devoted advocate for Sarah’s work, now seeking collaborative support.

3. Significant Themes

• The challenge of proving long-standing conjectures and the need for novel mathematical tools.
• The intersection of abstract mathematics and computational verification.
• Collaboration and the exchange of ideas across fields (mathematics, physics, AI).

4. Character Interactions

• Li Dong greets Sarah warmly, then Penrose finishes a phone call and enthusiastically asks Li Dong to help with Sarah’s problem.
• The three head to Nongyuan Cafeteria together to discuss details over a meal.

Summary: Professor Penrose's Standing

Key Plot Points

• Penrose complains to Li Dong about the difficulty of verifying 110,000 mathematical identities—cross-checking takes enormous time, and one tiny copying error (a wrong superscript) can invalidate days of work. Progress is slow and costly.
• At Nongyuan Cafeteria, Li Dong examines Penrose's manuscript and agrees it is correct but has a serious verification problem.
• Penrose reveals he has already recruited Mikhail Gromov, a legendary mathematician (creator of Geometric Group Theory), who is flying to Beijing to help.
• Li Dong agrees to join the effort, and Sarah thanks him formally.
• Meanwhile, in Göttingen, Clara is testing a cutting-edge AI model for data analysis. She discovers it has fabricated a non-existent paper, reversed its own conclusion, and used a function library that doesn't exist.
• Clara feeds the AI pure white noise; the AI confidently reports a significant periodic signal, proving it is fundamentally unreliable.
• She calls Peter Norvig, who explains the AI generates words "one by one, most plausibly," with no training objective for truth—only resemblance.
• Clara asks for "a more honest AI"; Norvig says none exists yet. She insists she cannot come home because "Humanity needs me" and hangs up.

Character Developments

• Li Dong: Shows sharp mathematical insight, calm pragmatism, and dry humor (denying his jokes are cold). Takes on a difficult challenge.
• Penrose: Becomes deflated and vulnerable, but shows resourcefulness in securing top collaborators like Gromov.
• Sarah: Quiet, earnest, deeply invested; expresses sincere gratitude.
• Clara: Rigorous and skeptical—refuses to accept AI output at face value, checks references, runs her own tests, and reaches a stark conclusion. Ends the chapter in dramatic, stubborn determination.
• Norvig: Patient and honest, but ultimately unable to offer a solution.

Significant Themes

• The human cost of mathematical verification: Cross-checking is the bottleneck—brilliance alone isn't enough.
• The fallibility of AI: The model fabricates references, invents tools, and hallucinates signals in pure noise with confident, polished language.
• Appearance vs. truth: The AI's output looks reliable—proper formatting, citations, p-values—but is fundamentally untrustworthy.
• Human expertise vs. AI imitation: Li Dong and Gromov represent real understanding; the AI represents plausible imitation without truth.

Character Interactions

• Penrose & Li Dong: Penrose seeks help; Li Dong accepts, setting up a collaboration with Gromov.
• Sarah & Li Dong: Sarah bows and thanks him; Li Dong playfully warns he won't return the thanks if he fails.
• Clara & the AI: A confrontation—Clara tests, catches the lie, and reaches a furious realization.
• Clara & Norvig: A frustrated phone call where Norvig explains the AI's nature; Clara cuts him off, choosing her mission over family.

Key Plot Points and Events

• Li Dong reviews Sarah Luowei's manuscript and is impressed by her talent.
• Lin Wei calls Li Dong to inform him that Huaxuan has taken on his model base project, which has been approved.
• Penrose calls Li Dong to his office, where Mikhail Gromov is waiting. Gromov has come to discuss Sarah's work and invites her to study under him, but she declines.
• Gromov suggests a method to verify the 110,000 identities in Sarah's work without checking each one individually.

Character Developments

• Li Dong shows interest in Sarah's work and her potential.
• Sarah demonstrates her loyalty and dedication to her mentor, Arthur Penrose, declining Gromov's invitation.
• Gromov shows appreciation for Sarah's talent and offers guidance on her work.

Significant Themes

• Mentorship and Guidance: Sarah's dedication to her mentor, Penrose, and Gromov's offer to guide Sarah highlight the importance of mentorship in academic growth.
• Talent and Potential: Both Li Dong and Gromov recognize Sarah's exceptional talent and potential in her chosen field.
• Academic Rivalry and Cooperation: Despite the competitive nature of academia, Gromov and Penrose show respect and cooperation, focusing on Sarah's best interests.

Character Interactions

• Li Dong and Lin Wei: They discuss the progress of the model base project, demonstrating their professional relationship and mutual respect.
• Penrose and Gromov: Despite their academic rivalry, they maintain a cordial relationship, focusing on Sarah's welfare and her academic growth.
• Sarah and Gromov: Gromov recognizes Sarah's talent and offers guidance, while Sarah respectfully declines, prioritizing her commitment to Penrose.
• Sarah and Penrose: Their mentor-mentee relationship is strengthened as Penrose acknowledges Sarah's talent and encourages her to pursue her interests, even if it means studying under another professor.

Key Plot Points and Events

• Professor Gromov proposes a method to tackle the 110,000 identities by focusing on critical pairs and proving confluence, which could significantly reduce the workload.
• The group discusses how to prove termination and list critical pairs without errors.
• Gromov leaves for Tsinghua University to visit an old friend, leaving the work to his students.

Character Developments

• Sarah and Penrose's eyes light up at Gromov's method, but Li Dong points out the remaining challenges.
• Gromov acknowledges the difficulties but sees no better alternative.
• Sarah reveals her true motivation for studying under Penrose, leading to a heartfelt conversation between them.

Significant Themes

• The struggle between brute force and finding structure in problem-solving.
• The importance of mentorship and personal growth in academic pursuits.
• The tension between human error and the quest for perfection in mathematical proofs.

Character Interactions

• Gromov shares his expertise and guides the group, but also makes it clear that he won't be heavily involved.
• Sarah and Penrose have a profound conversation about their relationship as teacher and student.
• Li Dong, feeling uncomfortable, considers leaving but ultimately decides to stay and interject with his AI model idea.

Summary: What Should Money Be Spent On?

Key Plot Points

• After leaving Penrose's office, Li Dong reflects on Sarah's talent and begins considering the untapped potential in Huaxia.
• He calculates his current wealth (Huaxuan's 500 million, future clinical trial payments, and large model income) and realizes he has no real use for the money.
• He develops an idea to establish a private research fund similar to the Clay Mathematics Institute—funding promising young researchers with no pressure for papers or commercial results, allowing them to pursue difficult problems freely.
• Lin Wei calls, reporting an engineering problem with adapting Li Dong's large model code to Huaxuan's accelerator card.
• Li Dong flies to Shanghai the next day with Old Zhang (his security detail).
• At Huaxuan, Lin Wei separates Li Dong from Old Zhang for confidentiality before meeting the engineering team.
• Chief Engineer Wang Qing explains the core issue: upper-layer operators translated successfully, but the precise low-level operators cannot be mapped onto the hardware because they require exact calculations over enormous finite rings—floating-point errors are unacceptable.

Character Developments

• Li Dong: Shows his growing maturity and long-term thinking. Once obsessed with money due to childhood poverty, he now recognizes money's limitations and begins contemplating philanthropy. His idealism emerges—he wants to fund difficult, high-risk research without demanding results. He also demonstrates pragmatic understanding of intellectual property confidentiality.
• Lin Wei: Professional and respectful; carefully manages confidentiality around Old Zhang, showing his experience in corporate secrecy.
• Engineer Wang Qing: Introduced as a highly accomplished architect (previous accelerator card designer, key figure in solving the 7nm yield problem). Direct, no-nonsense, and technically precise.

Significant Themes

• Wealth and purpose: What does money mean when one no longer lacks it? Li Dong transitions from viewing money as security to considering it a tool for advancing science.
• Talent waste: Contrast between Penrose's charity (helping poor students attend university) and Huaxia's different problem—talented researchers restrained not by access to education but by funding pressures and fear of failed projects.
• Scientific purity vs. commercial reality: Li Dong envisions a fund that frees researchers from publishing/project pressure, allowing honest mistakes and long-term work.
• Technical rigor: The challenge of mapping high-precision symbolic computation onto hardware—floating-point approximations can catastrophically corrupt the model's results.

Character Interactions

• Li Dong ↔ Penrose (indirect): Penrose's fund inspires Li Dong's idea, though he adapts it to Huaxia's context.
• Li Dong ↔ Lin Wei: Business relationship with casual banter ("Finished in just a few days?"), undercut by Li Dong's surprise that a problem exists.
• Li Dong ↔ Old Zhang: Brief, guarded interaction; Li Dong unobtrusively dismisses him to maintain secrecy.
• Lin Wei ↔ Wang Qing: Lin Wei defers to Wang Qing's technical expertise for the explanation.

Key Technical Takeaway

The problem lies in Li Dong's model's reliance on exact symbolic reasoning over finite rings; the precision operators cannot be expressed efficiently on the accelerator card's architecture, which is designed for standard floating-point operations.