• Home
  • Services
  • About
  • Contact
  • Blog
  • 知財活動のROICへの貢献
  • 生成AIを活用した知財戦略の策定方法
  • 生成AIとの「壁打ち」で、新たな発明を創出する方法

​
​よろず知財コンサルティングのブログ

EGGROLLでAGIは2年早まる?

2/12/2025

0 Comments

 
数十億のパラメーターを持つ大規模ニューラルネットワークに進化戦略 (ES)を適用する際の計算上およびメモリ上の障壁を克服するために設計されたEGGROLL(Evolution Guided General Optimization via Low-rank Learning)という最適化アルゴリズムが出現したことにより、AGIは2年早まるのではないかと言われ始めているそうです。
 
Evolution Strategies at the Hyperscale
https://arxiv.org/pdf/2511.16652
 
NVIDIA×オックスフォード:AGIは2年早まったか?
https://www.youtube.com/watch?v=xZr5is0MFrI
 
 
Will AGI Arrive Two Years Earlier Thanks to EGGROLL?
With the emergence of a new optimization algorithm called EGGROLL (Evolution Guided General Optimization via Low-rank Learning)—designed to overcome the computational and memory barriers involved in applying Evolution Strategies (ES) to large-scale neural networks with billions of parameters—there is growing speculation that AGI may arrive two years earlier than previously expected.

Your browser does not support viewing this document. Click here to download the document.
Your browser does not support viewing this document. Click here to download the document.
Your browser does not support viewing this document. Click here to download the document.
0 Comments

日本知財学会第 23 回年次学術研究発表会

2/12/2025

0 Comments

 
阿部12月2日 日本知財学会第 23 回年次学術研究発表会
 
2025年11月29、30日、日本知財学会第 23 回年次学術研究発表会が、「生成AIと知財戦略:企業イノベーションと知的創造サイクルの新展開」をテーマに開催されました。
シンポジウム「生成 AI と知財戦略:企業イノベーションと知的創造サイクルの新展開」のほか、「生成AIによる技術“発明”の衝撃 ~知財人財の能力開発の観点から論点を整理する~」、「生成 AI の特許実務への利用可能性と留意事項に関する検討と考察」、「AI の進化と知財実務の変化」のパネルディスカッションが企画され、一般発表でも生成AI関連の発表がありましたので、本当に久しぶりに参加しました。
旧世代の生成AIを使った発表が多かったのが残念でしたが、「AI の進化と知財実務の変化」のパネルディスカッションでの、知的財産研究教育財団 桂 正憲 理事長の発表、キヤノン(株)知的財産法務本部知的財産DX推進課 阿部 琢磨 課長の発表が印象的でした。
特に、キヤノンの阿部課長の発表は、生成AIの活用で年間数億円相当の業務効率化を達成した取組みの当初の計画から成果の現状、そして失敗した取組みまで、かなり細かいところまで発表されました。
また、「第23回年次学術研究発表会」では、生成AIを単なる「新技術」としてではなく、「知財実務を変えるツール」としてどう使いこなすか、そして「クリエイターの権利(特に声や画風)」を既存の法制度でどう守るか(あるいは法改正が必要か)という、実践的かつ法解釈の核心に迫る議論が多角的に展開されましたので、『生成AIを単なる「新技術」としてではなく、「知財実務を変えるツール」としてどう使いこなすか』という視点に絞り、「第23回年次学術研究発表会」の予稿集を生成AIに分析させました。
なお、生成AIによる調査・分析結果は、公開された情報からだけの分析であり、必ずしも実情を示したものではないこと、誤った情報も含まれていることについてはご留意されたうえで、ご参照ください。
 
第23回年次学術研究発表会
「生成AIと知財戦略:企業イノベーションと知的創造サイクルの新展開」
https://www.ipaj.org/workshop/2025/workshop_2025.html
 
 
The 23rd Annual Conference of the Japan Intellectual Property Association (JIPA)
On November 29–30, 2025, the 23rd Annual Conference of the Japan Intellectual Property Association was held under the theme “Generative AI and IP Strategy: New Developments in Corporate Innovation and the Intellectual Creation Cycle.”
In addition to the symposium “Generative AI and IP Strategy: New Developments in Corporate Innovation and the Intellectual Creation Cycle,” the program featured panel discussions such as “The Impact of Technical ‘Invention’ by Generative AI: Organizing Key Issues from the Perspective of Developing IP Talent,” “Study and Considerations on the Usability and Precautions of Generative AI in Patent Practice,” and “The Evolution of AI and Changes in IP Practice.” There were also several general presentations related to generative AI, which prompted me to attend for the first time in quite a while.
It was somewhat disappointing that many presentations still relied on older generations of generative AI. However, I was particularly impressed by the presentations in the panel discussion “The Evolution of AI and Changes in IP Practice,” especially those by Masanori Katsura, President of the Intellectual Property Research and Education Foundation, and Takuma Abe, Manager of the Intellectual Property DX Promotion Section, Intellectual Property and Legal Headquarters, Canon Inc.
In particular, Manager Abe’s presentation stood out. He disclosed, in surprising detail, Canon's initiatives to achieve work-efficiency improvements worth several hundred million yen per year through the use of generative AI—including the initial project plans, the current outcomes, and even the initiatives that did not succeed.
At the “23rd Annual Academic Research Conference,” the discussions went beyond treating generative AI as merely a “new technology.” Instead, they explored from multiple angles how to leverage it as a “tool that transforms IP practice,” and how to protect “creators’ rights (especially voices and artistic styles)” under the existing legal framework—or whether legal reform is necessary.
Focusing specifically on the perspective of “how to utilize generative AI not as a mere new technology, but as a tool that transforms intellectual property practice,” I had a generative AI analyze the conference proceedings of the 23rd Annual Academic Research Conference.
Please note that the investigation and analysis conducted by the generative AI are based solely on publicly available information, may not necessarily reflect actual circumstances, and may include inaccuracies. Please review the results with this in mind.

Your browser does not support viewing this document. Click here to download the document.
Your browser does not support viewing this document. Click here to download the document.
Your browser does not support viewing this document. Click here to download the document.
0 Comments

Gemini 4は、Gemini 3によって構築されている

1/12/2025

0 Comments

 
Googleの次世代AIモデル「Gemini 4」の開発戦略と、その基盤となる革新的な技術的特徴を生成AIに調べさせ、体系的に解説させました。
最大の特徴は、このモデルが「Gemini 3による自己構築」という再帰的な開発手法を採用している点であり、Gemini 3が合成データ生成やインフラの最適化を担うことで、開発プロセス自体がAIによって加速されています。技術面では、従来のTransformerの限界を超えることを目指し、継続的な学習と記憶の保持を可能にする「Nested Learning(入れ子構造学習)」や「HOPEアーキテクチャ」が導入されると予測されています。Gemini 4は、博士課程レベルの推論能力、Genie World Modelsの統合による物理世界の理解、および自律的なエージェント機能を大幅に強化し、AGI(汎用人工知能)への決定的な一歩となることが期待されています。
Googleは品質重視の開発哲学を掲げ、TPU v6 (Trillium)などの強力なインフラを背景に、SaaSのビジネスモデルを破壊しうる「AI認知のオペレーティングシステム」としての地位を確立しようとしています。
 
Gemini 4 Is Being Built by Gemini 3
I had generative AI investigate Google’s next-generation AI model “Gemini 4,” including its development strategy and the innovative technical foundations on which it is expected to be built, and then provide a systematic explanation.
The most notable feature is that the model adopts a recursive development method in which Gemini 3 effectively “builds” Gemini 4. By having Gemini 3 handle synthetic data generation and infrastructure optimization, the development process itself is accelerated by AI.
On the technical side, Gemini 4 is expected to go beyond the limitations of conventional Transformers, incorporating “Nested Learning”, which enables continuous learning and long-term memory retention, as well as a HOPE architecture. Gemini 4 is projected to significantly enhance PhD-level reasoning capabilities, physical-world understanding via integration with Genie World Models, and autonomous agent functions—positioning it as a decisive step toward AGI (Artificial General Intelligence).
Google upholds a quality-first development philosophy and, backed by powerful infrastructure such as TPU v6 (Trillium), aims to establish Gemini 4 as an “AI cognitive operating system” capable of disrupting the SaaS business model.

Your browser does not support viewing this document. Click here to download the document.
Your browser does not support viewing this document. Click here to download the document.
Your browser does not support viewing this document. Click here to download the document.
Your browser does not support viewing this document. Click here to download the document.
Your browser does not support viewing this document. Click here to download the document.
Your browser does not support viewing this document. Click here to download the document.
Your browser does not support viewing this document. Click here to download the document.
Your browser does not support viewing this document. Click here to download the document.
Your browser does not support viewing this document. Click here to download the document.
0 Comments

AIモデル、半導体、データセンターの三位一体戦略と宇宙

1/12/2025

0 Comments

 
2025年11月29日アップされたYouTube「【週刊AI】OpenAIの事実上「敗北宣言」の背景は、モデルx半導体xデータセンターの総合力。今はGoogle優勢かもしれないけど覇権争いは遂に宇宙へ」は、Googleの新AIモデルGemini3がOpenAIのSam Altman氏が同社社員に「大丈夫、すぐ追いつくから」というメモを送るくらい高評価を得ていること、Googleの強みはAIモデルと自社半導体のシナジー効果に加え巨大データセンターを既に持っていること、OpenAIもデータセンターを急ピッチで建設しているし、独自半導体の開発も行っているので、まだまだ勝負の行方は分からないこと、そんな中、イーロン・マスク氏とAmazon創業者のジェフ・ペゾス氏が、宇宙にデータセンターを建設する計画に乗り出したあどが紹介されています。
 
【週刊AI】OpenAIの事実上「敗北宣言」の背景は、モデルx半導体xデータセンターの総合力。今はGoogle優勢かもしれないけど覇権争いは遂に宇宙へ
https://www.youtube.com/watch?v=goFrdSA8rkE
 
The Trinity Strategy of AI Models, Semiconductors, and Data Centers, and the Move to Space
The YouTube video uploaded on November 29, 2025, titled “[Weekly AI] The Background Behind OpenAI’s De Facto ‘Concession of Defeat’: The Combined Power of Models × Semiconductors × Data Centers. Google May Be Ahead Now, but the Battle for Supremacy Is Finally Reaching Space” explains several key points.
First, Google’s new AI model Gemini 3 has received such high praise that OpenAI’s Sam Altman even sent an internal memo to employees saying, “Don’t worry, we’ll catch up soon.” Google’s strength lies not only in the synergy between its AI models and in-house semiconductors, but also in already possessing massive data centers.
At the same time, OpenAI is rapidly constructing its own data centers and developing custom chips, meaning the outcome of the race is far from decided.
Meanwhile, the video also highlights that Elon Musk and Amazon founder Jeff Bezos have embarked on plans to build data centers in space, signaling that the competition is expanding beyond Earth.

Your browser does not support viewing this document. Click here to download the document.
Your browser does not support viewing this document. Click here to download the document.
Your browser does not support viewing this document. Click here to download the document.
0 Comments

Gemini 3、Vibe Coding、Googleフルスタック戦略

1/12/2025

0 Comments

 
YouTube「Sundar Pichai: Gemini 3, Vibe Coding and Google 's Full Stack Strategy」は、2025年11月26日に公開されたアーカイブ動画で、Google DeepMindのローガン・キルパトリック氏とGoogleおよびAlphabetのCEOであるサンダー・ピチャイ氏との対談です。
Gemini 3のローンチ、Nano Banana Pro、そしてGoogleのAIの勢い、Googleのインフラへの長期的な投資、SOTAモデルの実際の出荷状況、そしてバイブコーディングの台頭、宇宙へのデータセンター設置といった将来のムーンショットへの考えについても語られています。
NotebookLMに要約レポートを作成させ、インフォグラフィック、スライド資料でまとめさせました。
 
Sundar Pichai: Gemini 3, Vibe Coding and Google 's Full Stack Strategy
https://www.youtube.com/watch?v=iFqDyWFuw1c&t=238s
 
 
Gemini 3, Vibe Coding, and Google’s Full-Stack Strategy
The YouTube video “Sundar Pichai: Gemini 3, Vibe Coding and Google’s Full Stack Strategy” is an archived interview released on November 26, 2025, featuring Logan Kilpatrick of Google DeepMind in conversation with Sundar Pichai, CEO of Google and Alphabet.
The discussion covers the launch of Gemini 3, Nano Banana Pro, the current momentum of Google’s AI initiatives, long-term investments in Google’s infrastructure, the real-world rollout of state-of-the-art models, the rise of Vibe Coding, and Google’s thinking on future moonshots such as establishing data centers in space.
I had NotebookLM generate a summary report, which I then compiled into an infographic and slide deck.
 

Your browser does not support viewing this document. Click here to download the document.
Your browser does not support viewing this document. Click here to download the document.
Your browser does not support viewing this document. Click here to download the document.
0 Comments
Forward>>

    著者

    萬秀憲

    アーカイブ

    January 2026
    December 2025
    November 2025
    October 2025
    September 2025
    August 2025
    July 2025
    June 2025
    May 2025
    April 2025
    March 2025
    February 2025
    January 2025
    December 2024
    November 2024
    October 2024
    September 2024
    August 2024
    July 2024
    June 2024
    May 2024
    April 2024
    March 2024
    February 2024
    January 2024
    December 2023
    November 2023
    October 2023
    September 2023
    August 2023
    July 2023
    June 2023
    May 2023
    April 2023
    March 2023
    February 2023
    January 2023
    December 2022
    November 2022
    October 2022
    September 2022
    August 2022
    July 2022
    June 2022
    May 2022
    April 2022
    March 2022
    February 2022
    January 2022
    December 2021
    November 2021
    October 2021
    September 2021
    August 2021
    July 2021
    June 2021
    May 2021
    April 2021
    March 2021
    February 2021
    January 2021
    December 2020
    November 2020
    October 2020
    September 2020
    August 2020
    July 2020
    June 2020

    カテゴリー

    All

    RSS Feed

Copyright © よろず知財戦略コンサルティング All Rights Reserved.
サイトはWeeblyにより提供され、お名前.comにより管理されています
  • Home
  • Services
  • About
  • Contact
  • Blog
  • 知財活動のROICへの貢献
  • 生成AIを活用した知財戦略の策定方法
  • 生成AIとの「壁打ち」で、新たな発明を創出する方法