Yu Murakami founded New York General Group, Inc. and Murakami Foundation, and leads their business operations and research and development.

 

He became interested in computers at the age of seven and has independently studied computer science and mathematics for more than 20 years since then. Among these fields, his main areas of interest are artificial intelligence and category theory, and he believes that category theory is the optimal solution for realizing artificial general intelligence (AGI). He invented a technology that incorporates category theory as the “brain” of an artificial intelligence system, while making Transformer-based large language models, Point-Voxel CNNs, and other systems function as other neural systems that exchange information between the brain and the environment. He respects the great inventor Nikola Tesla and named the collective body of these technologies the “World System.”

 

After graduating from Keio University, he began his career at PricewaterhouseCoopers (PwC). During his university years, he participated in internships at PwC, Bain & Company, and other organizations, which sparked his interest in business and finance. He later named his invention “World System on the basis of Bidirectional Encoder Representations from Transformers(BERT), Categorical Network(CN) and Point-Voxel Convolutional Neural Network(PVCNN)” and founded New York General Group, Inc. and Murakami Foundation in the United States.

 

New York General Group, Inc. and Murakami Foundation have now merged, with the former holding 100% of the latter’s shares. Thereafter, within five years of its founding, New York General Group, Inc. became one of the world’s largest asset, namely intellectual property, management companies, with consolidated group total assets exceeding USD 500 billion. The source of these assets is the vast number of intellectual properties in fields such as artificial intelligence, robotics, quantum computing, biotechnology, chemistry, pharmaceuticals/life sciences, energy, and space development, generated by Yu Murakami’s invention, “World System on the basis of Bidirectional Encoder Representations from Transformers(BERT), Categorical Network(CN) and Point-Voxel Convolutional Neural Network(PVCNN).” This intellectual property is managed by Murakami Foundation within New York General Group. Its valuation was conducted based on calculations by GPT-4 and GPT-4o, as well as opinions from qualified institutional investors and third-party experts, including Ph.D. holders in various fields. The valuation report was also submitted to U.S. regulatory authorities, and material matters concerning the series of processes were described in the Certificate of Incorporation, reviewed and approved. SEC Filings and IRS tax forms are currently being prepared.

 

New York General Group, Inc. is also registered in Japan (corporate number 4800-03-001744, stated capital of USD 10 million), and Yu Murakami currently serves as its representative.

 

Categorical AI, based on the Categorical Network he invented, outperforms large language models such as GPT-5.1 and Claude Opus 4.5 across multiple metrics, according to its technical report. In addition, the artificial intelligence technology invented by Categorical AI, “高次元微分演算子を計算するための確率的テンソル縮約ネットワーク（特願2025-046150）,” has received 特許査定 from the Japan Patent Office, among other achievements. This is the first case in Japan in which patent was granted for “artificial intelligence invented by artificial intelligence.”

 

His other major inventions include: 高次元微分演算子を計算するための確率的テンソル縮約ネットワーク（特願2025-046150・特許7734902、出願人：New York General Group, Inc.、発明者：村上 由宇、経過：特許取得）, マルチモーダル知識融合による適応的反射型神経記号システム（特願2025-077682、出願人：New York General Group, Inc.、発明者：村上 由宇）, マルチモーダル理解に基づく適応型分散圧縮システム（特願2025-079077）, 階層的適応型確率的情報顕在化機構を活用した多目的最適化部分観測マルコフ決定過程システム（特願2025-079078）, 階層的集約を用いた金融市場予測のためのマルチモーダルシステム及び方法（特願2025-046388・特許7812531、出願人：New York General Group, Inc.、発明者：村上 由宇、経過：特許取得）, 高頻度取引のための適応型インテリジェンスシステム（特願2025-042975、出願人：New York General Group, Inc.、発明者：村上 由宇）, 温度制御型エネルギー増幅機能を有する二次元/三次元ハイブリッド半導体構造及びその応用（特願2025-069364・特許7733347、出願人：New York General Group, Inc.、発明者：村上 由宇、経過：特許取得）, 動的再構成による相関復号量子コンピュータ（特願2025-057408・特許7749274、出願人：New York General Group, Inc.、発明者：村上 由宇、経過：特許取得）, 特定の疾患や細胞環境に合わせた有望な医薬品候補分子を生成するための新しい人工知能システム（特願2025-050843・特許7765795、出願人：New York General Group, Inc.、発明者：村上 由宇、経過：特許取得）, アニオン欠陥を有するバリウム-ケイ素オルトシリケート系オキシニトリド-ハイドライドの構造制御による高機能触媒及びその多元的応用方法（特願2025-080156、出願人：New York General Group, Inc.、発明者；村上 由宇）, トカマク型核融合炉用適応型エッジ局在3D磁場制御システム（特願2025-057034・特許7773161、出願人：New York General Group, Inc.、発明者：村上 由宇、経過：特許取得）, ハイブリッド変換器-マンバアーキテクチャを用いた高効率長文脈処理衛星システム（特願2025-067325・特許7866164、出願人：New York General Group, Inc.、発明者：村上 由宇）, and so on. These patents cite, as prior art, papers from top journals and conferences such as the Association for the Advancement of Artificial Intelligence, Nature, Neural Information Processing Systems, the American Chemical Society, American Physical Society Physical Review Letters, and the International Conference on Learning Representations, as well as patents from world-class universities and companies such as the Massachusetts Institute of Technology, International Business Machines Corporation, and Harvard University. They were granted patent after their novelty over these references was recognized, and they have also been cited in patents by global companies and research institutions such as China Telecom, Inspur, and the Academy of Military Science. They therefore constitute a group of technologies at the highest global level.

 

He also has experience conducting an exit in the United States through M&A, specifically the execution of a stock transfer agreement, involving a holding company with a patent portfolio related to artificial intelligence, semiconductors, and quantum computing, sold to qualified institutional investors.

 

He is also a member of organizations such as the American Mathematical Society and the American Physical Society, where he studies the mathematics and physics that form the foundation of his inventions.

 

In addition to being an inventor, he conducts fund-related operations as an Accredited Investor (17 C.F.R. § 230.501(a)(5)), Geeignete Gegenpartei (Gesetz über den Wertpapierhandel (Wertpapierhandelsgesetz – WpHG) § 67 Kunden; Verordnungsermächtigung Absatz 2 Nummer 2), Professional Investor (Securities and Futures Rules, Cap. 571D Cap. 571D §3(c)), and 適格機関投資家（金融商品取引法第二条に規定する定義に関する内閣府令第十条二十三イ、適格機関投資家の届出を金融庁長官に行った者）. He is currently participating, as a qualified institutional investor, in a fund originating from Kyushu University. He has also co-founded a pharmaceutical and food company with figures including Takehiko Fujino, Professor Emeritus at Kyushu University, and is responsible for formulating management strategy as its Chief Executive Officer.

 

At present, he is engaged not only in artificial intelligence research and development, but also in a wide range of activities, including research and development in nuclear fusion, quantum computers, rockets and satellites, semiconductors, pharmaceuticals through a partnership with the Reoken, and regional revitalization through a partnership with the Tsujita Group.

 

Papers

 

1. THE LARGEST GROTHENDIECK TOPOLOGY FOR WHICH A PRESHEAF IS A SHEAF

2. FIBERWISE COFINAL FUNCTORS AND FUBINI FORMULAE FOR POINTWISE KAN EXTENSIONS

3. A SELF-CONTAINED TÔHOKU THEOREM FOR DIAGRAM CATEGORIES

4. Finite Categorical Projection Estimators for Noninvertible Coherence Constraints

5. Boundary Spectral Process Categories: Observable Dynamics, Gauge Redundancy, and Topological Quantum Codes

6. Conservative Entropic Transport Attention: A Mass-Balanced and Stability-Certified Attention Primitive

7. Sectorwise Strictification of Projective Categorical Symmetry Actions in Anomalous Quantum Field Theory

 

Essays and Reports

 

1. The Categorical Imperative: Reshaping Reality Through the Lens of Abstract Structures

2. Future Dynamics of the Technological Landscape: Integration of the Metaverse, Quantum Computing, and Artificial General Intelligence (AGI)

3. Artificial General Intelligence (AGI): Transformative Innovation Across Biotechnology, Chemistry, Pharmaceuticals/Life Sciences, and Technology

4. Strategic Imperatives for Enterprises in the Age of Artificial General Intelligence: A Comprehensive Analysis

5. 2025 Global Economic Outlook: Navigating the Transformation Era

6. Investment Perspectives: Navigating the U.S. Policy Transformation

7. Global Megatrends Outlook 2025: Navigating the Transformative Forces Shaping Our Future

8. Is Attention All You Need? A Comprehensive Critical Examination of Transformer Architecture, Its Limitations, and the Future of Deep Learning

9. Comprehensive Analysis Report on the Global Semiconductor Industry (Japanese)

10. Innovative Outlook for Next-Generation Integrated Systems in the Robotics Industry: A Comprehensive Analytical Report Toward a Fundamental Transformation of the Industrial Structure (Japanese)

 

Patent Filing (2025, Japan）

 

Artificial Intelligence

 

　- 高次元微分演算子を計算するための確率的テンソル縮約ネットワーク （特願2025-046150・特許7734902、出願人：New York General Group, Inc.、発明者：村上 由宇、経過：特許取得）

　- 高頻度取引のための適応型インテリジェンスシステム（特願2025-042975、出願人：New York General Group, Inc.、発明者：村上 由宇）

　- 階層的集約を用いた金融市場予測のためのマルチモーダルシステム及び方法（特願2025-046388・特許7812531、出願人：New York General Group, Inc.、発明者：村上 由宇、経過：特許取得）

　- マルチモーダル知識融合による適応的反射型神経記号システム（特願2025-077682、出願人：New York General Group, Inc.、発明者：村上 由宇）

　- マルチモーダル理解に基づく適応型分散圧縮システム（特願2025-079077、出願人：New York General Group, Inc.、発明者：村上 由宇）

　- 階層的適応型確率的情報顕在化機構を活用した多目的最適化部分観測マルコフ決定過程システム（特願2025-079078、出願人：New York General Group, Inc.、発明者：村上 由宇）

　- 階層的分散マッチングによる動的環境下での最適歪み実現システム（特願2025-080158、出願人：New York General Group, Inc.、発明者：村上 由宇）

 

Robotics/Semiconductor

 

　・マルチモーダルフィードバック統合型知識検索強化ロボット制御システム（特願2025-067322、出願人：New York General Group, Inc.、発明者：村上 由宇）

　・階層的触覚記憶統合による予測的運動制御システム（特願2025-099828、出願人：New York General Group, Inc.、発明者：村上 由宇）

　・温度制御型エネルギー増幅機能を有する二次元/三次元ハイブリッド半導体構造及びその応用（特願2025-069364・特許7733347、出願人：New York General Group, Inc.、発明者：村上 由宇、経過：特許取得）

 

Bio/Chemistry/Pharmaceuticals and Life Sciences

 

　・特定の疾患や細胞環境に合わせた有望な医薬品候補分子を生成するための新しい人工知能システム（特願2025-050843・特許7765795、出願人：New York General Group, Inc.、発明者：村上 由宇、経過：特許取得）

　・組織工学のためのRNAオリガミベースのプログラム可能な細胞外マトリックス模倣材料（特願2025-064788、出願人：New York General Group, Inc.、発明者；村上 由宇）

　・pH制御型コバルト酸化触媒を用いた電気化学推進システム（特願2025-064781、出願人：New York General Group, Inc.、発明者；村上 由宇）

　・アニオン欠陥を有するバリウム-ケイ素オルトシリケート系オキシニトリド-ハイドライドの構造制御による高機能触媒及びその多元的応用方法（特願2025-080156、出願人：New York General Group, Inc.、発明者；村上 由宇）

 

Energy

 

　・トカマク型核融合炉用適応型エッジ局在3D磁場制御システム（特願2025-057034・特許7773161、出願人：New York General Group, Inc.、発明者：村上 由宇、経過：特許取得）

　・核融合炉ダイバータ用改良型ヘリカルナット冷却システム（特願2025-060194、出願人：New York General Group, Inc.、発明者：村上 由宇）

　・適応型ダイバータ-リミッタハイブリッドシステムを備えたトカマク型核融合炉（特願2025-062826、出願人：New York General Group, Inc.、発明者：村上 由宇）

　・適応型破片追跡を用いたトカマク核融合炉における放射エネルギー最適化システム（特願2025-083704、出願人：New York General Group, Inc.、発明者：村上 由宇）

 

Quantum Computing

 

　・動的再構成による相関復号量子コンピュータ（特願2025-057408・特許7749274、出願人：New York General Group, Inc.、発明者：村上 由宇、経過：特許取得）

　・量子軌道統計情報に基づく多段階最適化処理装置（特願2025-095224、出願人：New York General Group, Inc.、発明者：村上 由宇）

　・シュレーディンガー方程式の厳密解析に基づく多階層適応型動的位相補正量子誤り訂正システム（特願2025-095225、出願人：New York General Group, Inc.、発明者：村上 由宇）

 

Space

 

　・ハイブリッド変換器-マンバアーキテクチャを用いた高効率長文脈処理衛星システム（特願2025-067325、出願人：New York General Group, Inc.、発明者：村上 由宇）

　・衛星搭載型一貫性モデルによる地球システムモデル予測のリアルタイム高解像度化システム（特願2025-069190、出願人：New York General Group, Inc.、発明者：村上 由宇）

　・予測的世界モデルを用いた適応型ランデブー・ドッキング宇宙機システム（特願2025-069191、出願人：New York General Group, Inc.、発明者：村上 由宇）

　・適応型階層制御による低軌道衛星群の自律的軌道配置システム（特願2025-093732、出願人：New York General Group, Inc.、発明者：村上 由宇）

​

​

​