Skip to main content
  • Home
  • Library
  • Read
    • Popular
    • Charts
    • Browse
    • Categories
    • Wikis
    • Glossary
  • Request
  • Activity
  • Leaderboard
  • Help
Our Apps
  • DaoSearch
  • Narratorr
Social
  • Discord
  • Patreon

Recaps

The AI Wave Tycoon's Rise
VolumesArcsSummaries

Vol. 3Tangyuan’s Rise

Han Luyi brought Tangyuan online, secured domestic computing power through the Ximing acquisition, and advanced Source Intelligence’s AI products and infrastructure. The volume leaves the company leading a national domestic AI accelerator project as it works to scale its L100 cluster.

  • Jiang Yixin launched Riding the Wind, while Han Luyi advanced Glutinous Rice Ball’s training and used Sight to counter interference.
  • Han Luyi proved Dingsheng had trained Kun Yuan on private data, secured its removal and two thousand graphics cards, and deployed Tangyuan as the model behind Source Code Technology’s applications.
  • Han Luyi used a hidden watermark to prove Kun Yuan had trained on Source Intelligence’s private data, then negotiated its removal in exchange for two thousand graphics cards.
  • Han Luyi deployed Tangyuan on Source Intelligence’s own infrastructure and announced it as the model powering the company’s applications.
  • Han Luyi arranged a Beijing research center and pursued Ximing’s L100 cards to expand domestic computing power.
  • Source Intelligence acquired Ximing, found 8,360 L100 cards in its warehouse, and Han proposed a private AI tool for employee growth.
  • Jiang Yixin developed the Personal Growth System into Deep Blue, while Han Luyi adapted the L100 and tested an 8,192-card cluster.
  • Beijing Source Intelligence Technology was selected to lead a national domestic AI accelerator project, while Seek Him Technology’s computing-power proposal was rejected.
Chapters 202–305 Read from Ch. 202

Arc 9Riding the Wind Takes Flight

Source Code Technology launched Riding the Wind, weathered criticism, and faced Neus’s competing Atlas as the tool gained users and enterprise interest. Meanwhile, Han Luyi advanced Glutinous Rice Ball’s training amid interference, and Sight leveled up as training neared completion.

  • Han Luyi and Jiang Songran reviewed applicants and procurement plans for domestic AI infrastructure as Zheng Xiaobo considered bringing Han back to Dingsheng.
  • Jiang Yixin tested the coding agent, named it Riding the Wind with Su Niannian, and launched its beta, which drew users but also sharp criticism after CodeMonkey Old Zhang reported a damaging code change.
  • Ryan tested Riding the Wind on difficult coding tasks, then ordered Neus to analyze it and accelerate Atlas, while Han Luyi argued for safeguards and Su Niannian prepared a public launch.
  • Jiang Yixin demonstrated Riding the Wind at Source Code Technology’s annual party and opened its public beta, while employees celebrated and Han Luyi took Zhang Biao to Lu'an for the New Year.
  • Han Luyi corrected anomalous training data, stabilized Glutinous Rice Ball’s pretraining, and began annealing as Zhang Chi pushed Dingsheng to build an enterprise-focused competitor.
  • Jiang Yixin recognized that NeusAtlas copied Riding the Wind’s closed-beta features, and Han Luyi agreed to focus on the domestic market and let her decouple the product from specific models.
  • Han Luyi used Sight to identify and reverse computing interference against Glutinous Rice Ball, while developers debated Neus’s claim that Atlas was first and Riding the Wind registrations neared two million.
  • Jiang Yixin drafted enterprise pricing as clients asked for a paid version, Han Luyi monitored the final training and data-center deployment, and Sight leveled up.
Chapters 202–22625 summaries Read

Arc 10Tangyuan Takes the Lead

Han Luyi used proof that Dingsheng had trained Kun Yuan on Source Intelligence’s private data to secure its removal and graphics cards, then deployed Tangyuan on his own infrastructure. As NeusAI restricted Chinese access and GPT-V launched, Han introduced Tangyuan as the new model powering Source Code Technology’s applications.

  • Han Luyi used a hidden watermark to make Kun Yuan identify Source Intelligence Technology’s private dataset as unauthorized training data, and Gu Siyue helped him prepare a legal challenge.
  • After Han Luyi rejected a retroactive licensing deal, Lü Yun agreed to take Kun Yuan offline and settle the dispute by giving him two thousand graphics cards.
  • Dingsheng replaced Kun Yuan 1.0 with the weaker 1.1, while Han Luyi prepared to deploy Tangyuan on Source Intelligence’s own infrastructure.
  • Han Luyi and his team fixed a faulty network topology at Yunyin, raising the cluster’s capacity to 51 queries per second and starting Tangyuan on the system.
  • Jiang Yixin began migrating Riding the Wind and Open Things to Tangyuan after tests showed it outperforming Mercury, while Lin Yizhou quit his job after finishing freelance work with Riding the Wind.
  • NeusAI launched Venus and announced it would restrict access for Chinese users and companies, while OpenAI unveiled GPT-V and GPT-VOmni shortly afterward.
  • Han Luyi announced Tangyuan as the independently trained model behind Source Code Technology’s applications, confirmed Riding the Wind and Open Things would continue, and said API access would wait for more computing capacity.
Chapters 228–25425 summaries Read

Arc 11The L100 Acquisition

After Glutinous Rice Ball drew official and industry attention, Han Luyi negotiated a Beijing research center and pursued Ximing to secure domestic computing power. Source Intelligence acquired Ximing and its L100 cards, while Han returned to Haicheng to prepare the company’s next steps and consider an employee-focused AI tool.

  • Han Luyi demonstrated Glutinous Rice Ball to Haicheng officials and Professor Jia, then traveled to Beijing after Lu Zhengping invited him to discuss domestic AI infrastructure and investment.
  • Han Luyi proposed a focused push for domestic computing power and accepted Lu Zhengping’s offer in principle, while arranging a Beijing R&D center and a partnership with Shuimu University.
  • Zhang Wenduo tested Glutinous Rice Ball, revised his view of its abilities, and attended Han Luyi’s lecture, where Han discussed AI’s effects on work and the need for an interpretable World Model.
  • Lu Mingzhou negotiated to acquire Sili-Lumin Semiconductor, and Han Luyi convinced Xue Zhaoheng that Source Intelligence’s software could make Ximing’s L100 cards perform effectively.
  • When Ximing’s investors turned toward Penguin’s offer, Han Luyi arranged for Li Jia Tu to sell his stake to Xue Zhaoheng, giving the shareholders supporting Source Intelligence control of the sale.
  • Source Intelligence acquired Ximing for 250 million yuan, and Han Luyi found 8,360 L100 cards in its warehouse and ordered preparations to deploy 8,500 cards at Yunyin.
  • Back in Haicheng, Han Luyi planned to adapt the L100 for inference and training while Su Niannian prepared for an official inspection and took on delegated hiring decisions.
  • Han Luyi rejected employee surveillance as a use for AI and proposed a private growth tool that employees could use to track their own development.
Chapters 255–27925 summaries Read

Arc 12Deep Blue and the Beijing Institute

Jiang Yixin shaped the Personal Growth System into Deep Blue while Han Luyi advanced Source Intelligence’s L100 adaptation and opened a Beijing research institute. The company was invited to lead a national domestic AI accelerator project, while Seek Him Technology’s competing proposal was rejected.

  • Jiang Yixin tested her Personal Growth System, developed its Emergent Constellation approach with Glutinous Rice Ball, and began a prototype that generated a Pioneer character card from her work history.
  • Han Luyi hosted Zhu Jianming’s delegation, secured local support for Source Code Technology, and helped Chen Jianye pursue school admission for his daughter.
  • Jiang Yixin revised the Growth Assistant after learning from Qian Xiaole’s years of recorded reflections, and Han Luyi approved rolling it out as Deep Blue across Source Code.
  • Han Luyi used Sight to develop an L100 operator adaptation skill and optimized FusedAttentionBackward, prompting Fan Xiaotian to verify that the L100 outperformed an H100 on the benchmark.
  • Han Luyi, Zhao Wenyuan, and Jiang Songran tested the completed 8,192-card cluster at Yunyin, where scaling beyond individual clusters revealed communication congestion that needed further work.
  • Han Luyi and Zhao Wenyuan met Li Yan at an AI industry gathering, debated whether AI Agents should serve users or platforms, and discussed the practical limits of current models with attendees.
  • He Yunshen offered to connect Source Intelligence with chip-communication researchers, and Lu Zhengping told Han Luyi that the company was being invited to lead a national AI accelerator project.
  • Li Yan learned that Seek Him Technology’s Agent platform was approved but its Domestic Computing Power Cloud proposal failed, while Beijing Source Intelligence Technology was selected to lead the domestic AI accelerator project.
Chapters 280–30525 summaries Read
Volume 3 of 6
AboutStatsAPIChangelogStatusContactPrivacyTermsDMCA
© 2026 RandomTranslator·v3.26·RSS