Yukitaka Ohmura

RIGHT NOW: The Artemis II crew are attempting to resolve the toilet... difficulties... onboard Orion Integrity. Watch live on @NASASpaceflight's Artemis 24/7

【30周年記念】PC Watchは創刊30周年を迎えました。編集長よりご挨拶と今後の展望 pc.watch.impress.co.jp/docs/column/30…

🚀At re:Invent, we introduced the concept of frontier agents. Today, I'm excited to share that the first two—AWS Security Agent and AWS DevOps Agent—are now generally available: aws.amazon.com/blogs/machine-… Unlike traditional AI assistants that help with individual tasks, frontier agents work don’t need constant direction. These agents operate autonomously to solve complex problems, make decisions across multiple steps, and run persistently for hours or days until they accomplish their objectives. Here’s more on what these new class of agents can accomplish for your business: 🟠 We built AWS Security Agent with insights from operating security at Amazon scale and it’s changing how customers approach application security. Most organizations limit manual penetration testing to their most critical applications due to time and cost constraints, leaving the majority of their portfolio exposed. AWS Security Agent transforms penetration testing from a periodic event, typically covering a handful of applications on an annual or infrequent cadence, into a continuous, context-aware capability that increases the speed, frequency, and coverage of security testing across your entire application portfolio. AWS Security Agent operates like a human penetration tester. By ingesting your design documentation, source code, and architecture diagrams, it conducts context-aware analysis to uncover the critical vulnerabilities that matter to your business. In preview we helped customers like Scout24 SE, SmugMug, HENNGE, LG CNS, Bamboo Health, Wayspring, Classmethod, Inc., and SIXT gain deeper insights, faster, more frequently across more of their portfolios. 🟠 AWS DevOps Agent is your always-available operations teammate that resolves and proactively prevents incidents across AWS, multicloud, and on-premises environments. When incidents occur, it autonomously investigates root causes by correlating telemetry, code, and deployment data across your entire stack. It begins investigating the moment an alert comes in—whether at 2 AM or during peak hours—to reduce MTTR and quickly restore your application to optimal performance. We've helped customers such as United Airlines, T-Mobile, Western Governors University and Zenchef achieve lower MTTR and faster investigations, enabling 3–5x faster incident resolution. The frontier of AI agents is here. Check out our blog to start building.

【新LLMシリーズ「Namazu」の名称について】 先日、公開した Sakana Chat（chat.sakana.ai）に搭載した新LLMシリーズ「Namazu」の名称に関して、1990年代から2000年代にかけて広く親しまれた全文検索システム「Namazu」を想起される皆様から、貴重なご意見をいただきました。 全文検索システムNamazuは、日本語全文検索の礎を築いたシステムです。Sakana AIではNamazuの命名に際して商標等の法的な確認は行っていたものの、この分野の歴史的背景に対するリサーチが不足しておりました。 ご指摘を受け、全文検索システムNamazuの作者である高林様へ当社からご連絡差し上げ、経緯のご説明とともに、名称について相談いたしました。高林様からは、名称の使用をご快諾いただいたのみならず、「Namazuの発展に期待している」という旨の激励のお言葉を頂戴いたしました（私信のため大意のみ）。 偉大な先達が築き上げたその名に相応しいものとなるよう、今後の開発に取り組んでまいります。