豪門風雲錄 ·
甲骨文老板登頂世界首富，想起我在1998年讀到的那本書《拉里·埃里森與上帝不同》。很有意思，埃里森是當時業界段子最多的人。脾氣很壞，惡作劇，帆船和泡妞。這個人都是故事。
喬布斯曾經說，他在硅谷最好的朋友就是拉里，喬布斯在NEXT最苦惱地時候，拉里曾說我幫你把蘋果買回來。喬布斯回到蘋果后，牢牢控制董事會一人說了算。其中就有拉里的董事席。有段故事我印象很深，開會表決時，喬幫主對著拉里（人不在）的照片說，拉里我知道你會同意的。去看看《喬布斯傳》。這兩位身上，都是有股那個時代雅皮士的氣質。
甲骨文在很長時間里是關系型數據庫市場的霸主，擊敗Sybase和IBM DB2之后，在銀行和電信等主流存量市場長期保持優勢。誰敢輕易換啊。個人認為，在甲骨文歷史上，收購SUN公司并控制Java和MySQL，拿下千萬級開發者技術生態是十分令人意想不到的里程碑事件，十分轟動。多年后，微軟拿下Github代碼倉庫有類似戰略意義。
這幾年，他們發力云計算，是導致市值大漲的又一個轉折點。云計算市場太大了。AI浪潮再度席卷后，這個市場成為大預言模型和英偉達算力之后最大的受益者。微軟、亞馬孫和甲骨文這三家，他們不用像OpenAI那麼卷，像工作狂老黃那麼拼。是的，他們只要不犯錯，跟上主流技術腳步；就可以大發橫財。同樣地，微軟哪位印度裔CEO牙都快笑歪了。
甲骨文在軟件時代就是全球唯二的巨頭。互聯網還沒有起來之際，微軟和甲骨文爭霸，微軟幾乎一統天下，吃定桌面操作系統和各項應用。拉里就是成天和比爾蓋茨唱反調的高手，你說個人電腦PC，他就鼓吹NC（Net computer），大喊其實網絡就是計算機。有點以太坊喊去中心化網絡就是世界計算機的那味。2000年前；業界十分敬畏微軟，微軟在軟件領域啃下一塊又一塊市場，而且一旦吃下就沒有只有第一沒有第二，但就是很難啃下甲骨文的關系型數據庫這塊大肥肉。
埃里森是當時業界段子最多的人。其實，這哥們不容易，商業數據庫終究還是被微軟SQL Server反超，開源數據庫My SQL這幾年也被斯坦福搞出來的PostgreSQL超越，后者這幾年很流行。但是，真英雄不是悲壯地去死，而還要有打不死的九條命。做企業像三級火箭，舊的燃料推上去之后，下一個市場燃料又補上，繼續飛。
他可能也是今天最年長的科技企業創始人CEO。
Joel來談日本
這兩年在日本的人工智慧領域中，有一位備受矚目的專家松尾豊，他並不只將自己局限於研究室之內，而是活躍於學界、產業界與政策圈的多重舞台。他既推動年輕學者與新創公司的成長，也以政府諮詢委員身份持續提供政策建議，還致力於與企業之間建立合作。
松尾豊所在的東京大學「松尾・岩澤研究室」，外界簡稱「松尾研」，已成為孕育AI創業能量的重要基地。目前約有四百名國內外的研究人員與學生參與，迄今誕生了三十餘家新創公司，形成了以東京本鄉為核心的「本鄉谷」，其定位正如矽谷之於美國，展現出濃厚的創新聚落氛圍。對於松尾本人而言，從研究到創業支持，再到國家政策建言，這些角色看似繁重，卻都是他認為推動AI進展所必須兼顧的行動。
松尾豊的AI研究之路可追溯至童年。小學時，他便沉迷於能輸入簡單程式的口袋型電腦，同時對於「人類意識是什麼」這樣的哲學問題產生濃厚興趣。這種兼具理工與人文的探索精神，成為他日後投入人工智慧研究的重要基礎。
AI這一詞彙誕生於上世紀五〇年代，歷經起伏，如今因大規模語言模型與生成影像技術而再次站上浪潮。特別是以深度學習為核心的突破，自2012年多倫多大學在國際影像辨識競賽中大放異彩以來，便掀起新一輪革命。松尾敏銳地捕捉到這一轉折，將深度學習視為必須投注心力的方向，並在2015年著書將AI的潛力比作「值得購買的彩券」。然而，他回顧時坦言，日本錯過了這張彩券，而中國則透過「中國製造2025」戰略果斷下注，如今與美國並列全球AI雙強。
對比之下，日本在AI發展上展現出遲疑與保守。雖然近年生成式AI的風潮終於迫使政府與產業界加快腳步，投入GPU、資料中心建設以及人才培育，但要追趕美中兩國已非易事。中國的AI發展除了仰賴阿里巴巴、騰訊、字節跳動等科技巨擘的資金與資源支持，也有無數新創公司湧現，形成厚實的生態系。
近期中國研究團隊推出的「DeepSeek」，以高性能和低成本引發全球矚目，雖然後來毀譽參半。但是松尾認為，這只是冰山一角，中國的整體技術水準還是超過外界想像。相較之下，日本社會常因抗拒變化而錯失機會，習慣延續既有做法，對新技術的嘗試缺乏勇氣。他直言，這種「不允許變化」的文化正是日本經濟三十年停滯的原因之一。他以棒球為喻，強調應讓具備真正實力的人站上打擊區，而非由保守的「教練」取而代之。
儘管外界將松尾研究室視為特例，他卻淡然解釋，自己不過是「沒有做奇怪的事」，只是順應科技潮流、重視人才潛力。對於年輕人，他鼓勵在能力允許的情況下勇於嘗試創業，因為每個人都應有機會找到最適合的舞台。針對社會對AI可能帶來失業或監控強化的擔憂，他舉汽車在日本普及的例子指出，當時人們選擇以積極態度擁抱新技術，才成就了汽車大國的地位。
AI也應該如此，不必過度恐懼，應該先嘗試再調整。他強調，科技、人才與組織本身都蘊含潛在能量，只要避免僵化與否定，讓這些內在力量自由釋放，成長與創新自然會隨之而來。由此可見，日本在AI競賽中雖處於落後，但只要勇於接受變化，並讓人才與技術獲得適當舞台，仍有機會在全球AI發展的激烈競逐中重塑地位。
參考資料：朝日新聞
In December, the company launched the Sora model.[107][108] It also launched OpenAI o1, an early reasoning model that was internally codenamed strawberry.[109] Additionally, ChatGPT Pro—a $200/month subscription service offering unlimited o1 access and enhanced voice features—was introduced, and preliminary benchmark results for the upcoming OpenAI o3 models were shared.[110]
2025
Investments
On January 21, 2025, Donald Trump announced The Stargate Project, a joint venture between OpenAI, Oracle, SoftBank and MGX to build an AI infrastructure system in conjunction with the US government. The project takes its name from OpenAI's existing "Stargate" supercomputer project and is estimated to cost $500 billion. The partners plan to fund the project over the next four years.[111]
In March, OpenAI reached a deal with CoreWeave to acquire $350 million worth of CoreWeave shares and access to AI infrastructure, in return for $11.9 billion paid over five years. Microsoft was already CoreWeave's biggest customer in 2024.[112] Alongside their other business dealings, OpenAI and Microsoft were renegotiating the terms of their partnership to facilitate a potential future initial public offering by OpenAI, while ensuring Microsoft's continued access to advanced AI models.[113]
April began as OpenAI raised $40 billion at a $300 billion post-money valuation, marking the largest private technology deal on record. The financing round was led by SoftBank, with other participants including Microsoft, Coatue, Altimeter, and Thrive.[114][115]
On April 9, OpenAI countersued Musk in federal court, alleging that he had engaged in "bad-faith tactics" to slow the company's progress and seize its innovations for his personal benefit. OpenAI also argued that Musk had previously supported the creation of a for-profit structure and had expressed interest in controlling OpenAI himself. The countersuit seeks damages and legal measures to prevent further alleged interference.[116]
On May 21, OpenAI announced the $6.5 billion acquisition of io, an AI hardware start-up founded by former Apple designer Jony Ive in 2024.[117][118][119] The deal was the company's largest acquisition to date.[120][121]
In June, OpenAI began renting Google Cloud's Tensor Processing Units (TPUs) to support ChatGPT and related services, marking its first meaningful use of non‑Nvidia AI chips.[122]
In July, the United States Department of Defense announced that OpenAI had received a $200 million contract for AI in the military, along with Anthropic, Google, and xAI.[123] In the same month, the company made a deal with the UK Government to use ChatGPT and other AI tools in public services.[124][125] OpenAI subsequently began a $50 million fund to support nonprofit and community organizations.[126]
In September, OpenAI agreed to acquire the product testing startup Statsig for $1.1 billion in an all-stock deal and appointed Statsig's founding CEO Vijaye Raji as OpenAI's chief technology officer of applications.[127] The company also announced development of an AI-driven hiring service designed to rival LinkedIn.[128]
Products
On January 23, OpenAI released Operator, an AI agent and web automation tool for accessing websites to execute goals defined by users. The feature was only available to Pro users in the United States.[129][130] OpenAI released deep research agent, nine days later. It scored a 27% accuracy on the benchmark Humanity's Last Exam (HLE).[131] Altman later stated GPT-4.5 would be the last model without full chain-of-thought reasoning.[132][133]
In July 2025, reports indicated that artificial intelligence models by both OpenAI and Google DeepMind solved mathematics problems at the level of top-performing students in the International Mathematical Olympiad. OpenAI's large language model was able to achieve gold medal-level performance, reflecting significant progress in AI's reasoning abilities.[134]
Management
Key employees
- CEO and co-founder: Sam Altman, former president of the start-up accelerator Y Combinator
- President and co-founder: Greg Brockman, former CTO, 3rd employee of Stripe[135]
- Chief Scientist Officer: Jakub Pachocki, former Director of Research at OpenAI[91]
.svg){kind=logo} | |
| Company type | Private |
|---|---|
| Industry | Artificial intelligence |
| Founded | December 8, 2015[1] |
| Founders |
|
| Headquarters | 1455 3rd Street, San Francisco, California, U.S.[2] |
Key people | |
| Products | OpenAI Five |
| Revenue |  US$3.7 billion[4] (2024 est.) |
 US$−5 billion[4] (2024 est.) | |
Number of employees | 3,000 (2025)[5] |
| Website | openai |
Aug 21, 2025 — OpenAI reported over $1 billion in monthly revenue for the first time in July 2025, more than doubling since the start of the year 125. The ...
8月4日
OpenAI’s ChatGPT to hit 700 million weekly users, up 4x from last year
- ChatGPT is set to hit 700 million weekly active users, with usage growing 4X year over year.
- OpenAI now counts five million paying business users, up from three million in June, as enterprises and educators embrace artificial intelligence tools.
- The milestone follows news last week that OpenAI secured $8.3 billion from top investors, including Dragoneer Investment Group, Andreessen Horowitz and Sequoia Capital.
沒有留言:
張貼留言