儲備 VS 庫存 Stockpile refers to a large, strategic reserve of goods accumulated to guard against future shortages or emergencies, while inventory is the comprehensive, active list of all materials—raw, work-in-progress, and finished goods—a company holds for day-to-day operations. Inventory is continuously moving (cycling), while a stockpile is meant to stay in reserve.

Missile displays in Tehran in 2024. 2024年德黑蘭飛彈展示。

According to classified U.S. intelligence assessments, Iran retains roughly 70 percent of its prewar missile stockpile.Credit...
根據美國機密情報評估，伊朗仍保留戰前飛彈庫存的約70%。2026 05


儲備是指為應對未來短缺或緊急情況而累積的大量策略性物資儲備，而庫存則是公司為日常營運而持有的所有物料（包括原材料、在製品和成品）的全面、動態清單。庫存是持續流動（循環）的，而儲備則旨在保持儲備狀態。
Stockpile refers to a large, strategic reserve of goods accumulated to guard against future shortages or emergencies, while inventory is the comprehensive, active list of all materials—raw, work-in-progress, and finished goods—a company holds for day-to-day operations. Inventory is continuously moving (cycling), while a stockpile is meant to stay in reserve.

黃克孫；《 魯 拜 集 》（ the Rubaiyat）因緣 2026 感謝書林書局蘇先生Jerome Su 給我及 BEN 解些黃克孫 與 MORRIS CHANG.... /

 

黃克孫；《 魯 拜 集 》（ the Rubaiyat）因緣 2026 感謝書林書局蘇先生Jerome Su 給我及 BEN 解些黃克孫 與 MORRIS CHANG.... /

//黃來台灣四五次，第一次（1986/87）是國科會邀請，清華大學中文系透過我為他辦了座談會，有可能他們在新竹見面。他只跟我提過楊森的女兒師大英語系馬楊萬運的先生、經緯電腦的創辦人馬賓農是他的學長。當年資訊不發達，國科會邀請他，但有關黃克孫的生平、著作承辦人員都是跟我打聽。我是去美國新聞處查 Who’s Who 及Books in Print 後提供的。我猜國科會邀請他是因爲李遠哲得了諾貝爾化學獎，推測他有可能獲得物理獎。他之前沒來過台灣，但曾有中華民國國籍，因為他父親是菲律賓華僑中學的校長。// SU

//我查了一下，張忠謀是1931年次，黃克孫是1928年出生，他們只是校友或系友，不是同學。因為黃克孫是天才兒童，他在MIT讀研究所時，同學年紀都比他大，他1950年獲得博士時才22歲。// SU

《狂酒歌》晨鐘文叢 ----

Ben Chen 年過八十的黃克孫，在台大講「統計力學特論」

Kerson Huang,與張忠謀同時期在MIT唸大學。後來隔了近60年才又在台灣見面。 黃克孫，是全才型的學者，譯有《魯拜集》，也英譯過《易經》

我大一開學時晨鐘出版《狂酒歌》 《狂酒歌》台北:晨鐘文叢1971 陳次雲的譯本為第四衍譯本101首 孟祥森的根據第一衍譯本75首

1978年在英國Essex大學的宿舍某位伊朗朋友知道我喜歡他們的 the Rubaiyat 就送我一本該國出版的袖珍集-

DeepL is transforming into an "AI-native" company, shifting from a specialized translator to a broader AI platform, highlighted by a May 2026 restructuring that includes cutting 250 jobs to accelerate AI advancements.。2024 日本最熱衷應用機械翻譯DeepL， NHK 的Direct Talk 專訪DeepL 公司執行長 CEOJaroslaw Kutylowski

DeepL is transforming into an "AI-native" company, shifting from a specialized translator to a broader AI platform, highlighted by a May 2026 restructuring that includes cutting 250 jobs to accelerate AI advancements.。2024 日本最熱衷應用機械翻譯DeepL， NHK 的Direct Talk 專訪DeepL 公司執行長 CEOJaroslaw Kutylowski 


AI Overview






DeepL is transforming into an "AI-native" company, shifting from a specialized translator to a broader AI platform, highlighted by a May 2026 restructuring that includes cutting 250 jobs to accelerate AI advancements. CEO Jarek Kutylowski is focusing on integrating Language AI with new tools like "DeepL Agent" for business task automation. [1, 2, 3]

Key Aspects of DeepL's AI Transformation (As of May 2026):

• Strategic Restructuring: The company is cutting roughly 25% of its workforce (about 250 employees) to shift towards smaller, more agile teams capable of accelerating AI development in a rapidly changing landscape.
• AI-First Platform: Moving beyond simple translation, the company focuses on a unified, "AI-first" multilingual platform. This includes text, file, and real-time voice translation designed for enterprise workflows.
• DeepL Agent & Automation: The launch of DeepL Agent allows businesses to automate complex, language-focused workflows and tasks using simple instructions.
• Enterprise Focus: With over 200,000 business customers, DeepL has introduced DeepL for Enterprise to provide specialized, secure language solutions (e.g., custom glossaries, high-level data security).
• Technology & Growth: The platform, powered by the DeepL Mercury supercomputer, specializes in producing high-quality, human-like translations while expanding into writing assistance and agentic AI. [1, 2, 3, 4, 5, 6, 7, 8]

DeepL aims to maintain its position as a "global AI leader" by adapting to the rapid AI advancements, with CEO Kutylowski characterizing the shift as a necessary step for the company's future. [1]

AI Overview

An "AI-native" company is an organization built from the ground up with artificial intelligence as its foundational core, rather than adding AI as a feature to legacy systems. These companies use AI-driven, goal-oriented systems to automate operations, enable real-time decision-making, and create products that would be impossible without AI. [1, 2, 3, 4]

Key Characteristics of AI-Native Companies:

• Foundation-First Architecture: AI is ingrained in the technology stack, data structure, and user experience from the beginning.
• Continuous Feedback Loops: Systems are designed to learn, adapt, and improve automatically with every user interaction.
• AI Agent Integration: Products often rely on AI agents that actively work toward goals within defined parameters, rather than relying on manual, rule-based software.
• Data-Driven Operations: Data is not just stored; it is used in real-time to drive workflows, such as in personalized customer service or automated, adaptive workflows.
• Rapid Iteration: Because they are not hindered by legacy systems, AI-native companies operate with higher agility and faster product development cycles. [1, 2, 3, 4, 5, 6, 7]

Examples vs. Traditional Approaches:

• AI-Native: A customer service platform where AI agents autonomously resolve complex tickets from start to finish.
• AI-Enabled (Not Native): A traditional customer service platform that adds a chat bot to answer basic FAQ questions. [1, 2, 3]

AI-native companies aim for leaner operations and superior efficiency by making AI the default, not the exception. [1]

Jaroslaw Kutylowski / Founder and CEO of DeepL - Direct Talk
NHK
https://www3.nhk.or.jp › nhkworld › ondemand › video







NHK WORLD - JAPAN On Demand · AI Translation for a Better World: Jaroslaw Kutylowski / Founder and CEO of DeepL ...
NHK · Dec 8, 2022