单词学习：![[2026052171829020.mp3]]

新闻播报：![[2026052119952602.mp3]]

Google I/O 2026 聚焦 AI 革命：Gemini 3.5 上线与多项 AI 应用及硬件齐发

简报：

• Google I/O 2026 开发者大会正式拉开序幕，聚焦 AI 技术革命，主打高速低延迟的 Gemini 3.5 Flash 模型已上线运作，进阶版 Gemini 3.5 Pro 预计下个月推出。
• 推出全天候 AI 个人助理“Gemini Spark”，具备24小时云端自主运行能力并深度整合 Google 核心生态，执行高风险操作前强制要求用户确认，下周起向美国 Google AI Ultra 订阅用户开放 Beta 测试。
• Google Docs 新增“Docs Live”语音功能，用户口述想法即可由 Gemini 自动撰写草稿并排版生成正式文件，预计今年夏天向 Pro 与 Ultra 订阅者推出。
• 推出“Ask YouTube”功能，用户可直接对影片提问，AI 能够理解问题并将时间轴自动调整至最相关片段，今夏率先对美国地区年满18岁的 YouTube Premium 用户开放。
• Google 宣布与三星、高通联手开发“Android XR 智慧眼镜”，深度结合 Gemini AI，并与 Gentle Monster 等时尚品牌跨界合作，预计于今年秋季推出。

相关链接：

• Google2026開發者大會I /O黑科技連發！Gemini Spark全天候AI助理幫工作

Google I/O 2026 Spotlights the AI Revolution: Gemini 3.5 Launches Alongside a Wave of AI Apps and Hardware 🚀🤖

Briefing:

• The Google I/O 2026 developer conference has officially begun, centering on a revolution in AI technology. The high-speed, low-latency Gemini 3.5 Flash model is now live, while the advanced Gemini 3.5 Pro is expected to launch next month.
• Google introduced “Gemini Spark,” an always-on AI personal assistant with 24/7 autonomous cloud-based operation and deep integration across Google’s core ecosystem. User confirmation is mandatory before any high-risk actions are carried out. Beta access will open next week to Google AI Ultra subscribers in the United States.
• Google Docs is adding a new voice feature called “Docs Live,” allowing users to dictate their ideas while Gemini automatically drafts, formats, and generates polished documents. The feature is expected to roll out this summer to Pro and Ultra subscribers.
• Google also unveiled “Ask YouTube,” a feature that lets users ask questions directly about videos. The AI can understand the query and automatically jump the timeline to the most relevant segment. It will debut this summer for YouTube Premium users aged 18 and older in the United States.
• Google announced a partnership with Samsung and Qualcomm to develop “Android XR smart glasses,” deeply integrated with Gemini AI, and is also collaborating with fashion brands such as Gentle Monster. The product is expected to launch this fall. 😎

Related links:

• Google 2026 Developer Conference I/O Unveils a Wave of Futuristic Tech! Gemini Spark, the Always-On AI Assistant, Helps with Work

OpenAI 推理模型推翻埃尔德什几何猜想，多位数学家证实其原创性

简报：

• OpenAI 宣布其全新通用推理模型成功推导出原创数学证明，推翻了保罗・埃尔德什于 1946 年提出的几何学著名未解猜想。
• 该模型推翻了“最优解趋近正方形网格”的传统认知，发现了一类表现更优的全新构造体系，被视为 AI 首次自主攻克数学核心领域重大难题。
• 此次发布附带了诺加・阿隆、梅兰妮・伍德及托马斯・布鲁姆等知名数学家的佐证，证实该证明并非引用学术文献中的现成解法。
• 该成果由通用推理模型而非专用数学系统得出，展示了人工智能在处理复杂逻辑推理链条及跨学科知识关联方面的显著进展。

相关链接：

• AI 推翻著名几何猜想，OpenAI 宣布攻克 80 年数学难题

OpenAI Reasoning Model Overturns Erdős Geometric Conjecture, with Multiple Mathematicians Confirming Its Originality 🚀📐

Brief:

• OpenAI announced that its new general-purpose reasoning model has successfully derived an original mathematical proof overturning a famous unsolved geometry conjecture proposed by Paul Erdős in 1946.
• The model challenges the traditional belief that the optimal solution approaches a square grid, instead discovering a new class of constructions with superior performance. This is being regarded as the first time AI has independently solved a major open problem at the core of mathematics.
• The release was accompanied by endorsements from prominent mathematicians including Noga Alon, Melanie Wood, and Thomas Bloom, confirming that the proof was not based on any existing solution from the academic literature.
• The result was produced by a general-purpose reasoning model rather than a specialized mathematics system, demonstrating significant progress in AI’s ability to handle complex chains of logical reasoning and connect knowledge across disciplines. 🧠

Related link:

• AI Overturns Famous Geometry Conjecture: OpenAI Announces Breakthrough on an 80-Year-Old Mathematical Problem

SpaceX拟定宏伟航天目标：五年内冲击年发射万次大关

简报：

• 据美国联邦航空管理局（FAA）局长透露，SpaceX正酝酿一项极具野心的计划，设定在未来五年内实现每年高达10000次的火箭发射目标，旨在大幅推动商业航天进程。
• FAA对此回应表示，在批准此类扩张计划之前，政府官员需要看到SpaceX的可靠性有所提升。
• SpaceX曾在今年1月透露另一项计划，拟发射并组建一个由100万个卫星组成的星座，让这些卫星环绕地球运行，并利用太阳能为人工智能数据中心供电。

相关链接：

• SpaceX据悉计划五年内实现每年1万次发射
• 消息称SpaceX正酝酿航天计划，设定未来五年内实现每年高达1万次火箭发射的宏伟目标

SpaceX Sets Ambitious Space Goals: Aiming for 10,000 Launches Per Year Within Five Years 🚀

Brief:

• According to the head of the U.S. Federal Aviation Administration (FAA), SpaceX is developing an extremely ambitious plan to achieve as many as 10,000 rocket launches annually within the next five years, with the goal of significantly advancing the commercial space industry.
• In response, the FAA said that before approving any such expansion plan, government officials would need to see improvements in SpaceX’s reliability.
• Earlier this January, SpaceX also revealed another plan: to launch and assemble a constellation of 1 million satellites orbiting Earth, using solar power to supply energy to artificial intelligence data centers. 🌍☀️

Related links:

• SpaceX Reportedly Plans to Reach 10,000 Launches Per Year Within Five Years
• Sources Say SpaceX Is Developing a Space Plan With the Ambitious Goal of Achieving Up to 10,000 Rocket Launches Per Year Within the Next Five Years

马斯克起诉 OpenAI 因超时效被驳回，发文抨击高层“窃取慈善”并誓言上诉

简报：

• 美国加州联邦陪审团于 5 月 18 日裁定，马斯克起诉 OpenAI 及其执行长奥特曼一案因超过 3 年法律追诉期限而败诉。
• 陪审团在不到 2 小时的讨论后达成共识，认为马斯克早知 OpenAI 转向营利模式却延迟提告，法官已正式采纳该裁决。
• 马斯克在 X 平台对判决结果表示强烈不满，公开指责奥特曼与布洛克曼“偷走慈善机构致富”，并称该判决只是“技术性问题”。
• 马斯克宣布将向美国第九巡回上诉法院提起上诉，继续要求撤换 OpenAI 领导层并追回 1800 亿美元所得。
• OpenAI 律师回应称，马斯克因在 AI 市场竞争失利才发起诉讼作为武器，并指出马斯克多年前曾提议在由其掌控的前提下实行营利化。

相关链接：

• 馬斯克告OpenAI輸了！ 怒轟奧特曼「偷慈善機構致富」喊上訴

Musk’s Lawsuit Against OpenAI Dismissed as Time-Barred, Slams Leadership for “Stealing a Charity” and Vows to Appeal ⚖️🔥

Brief:

• A federal jury in California ruled on May 18 that Elon Musk’s lawsuit against OpenAI and its CEO Sam Altman failed because it was filed after the three-year statute of limitations had expired.
• After less than two hours of deliberation, the jury reached a consensus that Musk had long known about OpenAI’s shift toward a for-profit model but waited too long to sue, and the judge has formally adopted the verdict.
• Musk expressed strong dissatisfaction with the ruling on X, publicly accusing Altman and Brockman of “stealing a charity to enrich themselves,” and said the decision was merely a “technicality.”
• Musk announced that he will appeal to the U.S. Court of Appeals for the Ninth Circuit, continuing to seek the removal of OpenAI’s leadership and the recovery of dJyH6kDkfC5RlM16.answerText80 billion in gains.
• OpenAI’s lawyers responded that Musk launched the lawsuit as a weapon after losing ground in the AI market, and noted that Musk himself had proposed years ago that OpenAI adopt a for-profit structure under his control. 🤖

Related link:

• Musk Loses OpenAI Lawsuit! Furiously Accuses Altman of “Getting Rich by Stealing a Charity” and Vows to Appeal

阿里旗舰大模型Qwen3.7-Max跻身全球前五，位列国产榜首

简报：

• 5月21日，三方机构ArtificialAnalysis公布最新全球大模型榜单，阿里新发布的旗舰模型Qwen3.7-Max得分56.6分。
• Qwen3.7-Max性能接近GPT、Claude、Gemini的最强模型，成功位列全球第五、国产大模型第一。
• 据悉，Qwen3.7-Max即将上线阿里云百炼，对外提供API服务。

相关链接：

• ArtificialAnalysis：千问3.7问鼎国产模型冠军，全球前五

Alibaba’s Flagship Large Model Qwen3.7-Max Breaks Into the Global Top Five, Ranking First Among Chinese Models 🚀

Brief:

• On May 21, third-party organization ArtificialAnalysis released its latest global large model rankings, with Alibaba’s newly launched flagship model Qwen3.7-Max scoring 56.6 points.
• Qwen3.7-Max delivers performance close to the strongest models from GPT, Claude, and Gemini, successfully ranking fifth worldwide and first among Chinese large models. 🌟
• It is reported that Qwen3.7-Max will soon be launched on Alibaba Cloud Bailian, where it will provide external API services.

Related Link:

• ArtificialAnalysis: Qwen 3.7 Crowned Champion Among Chinese Models, Global Top Five

贝索斯称AI如“推土机”将提升全人类价值，驳斥人工智能裁员焦虑论

简报：

• 亚马逊创始人贝索斯在蓝色起源基地接受采访时表示，AI将像“推土机替代铁锹”一样提升工作效率，呼吁软件工程师等从业者无需为技术变革感到焦虑。
• 贝索斯认为AI会提升所有人的价值并带动社会生产率大幅跃升，若技术普及不受限，人类将迎来食品、住房价格更亲民的“物资富足时代”。
• 针对业内关于“AI导致大规模裁员”的担忧，贝索斯持乐观态度，这与杰弗里·辛顿及Anthropic CEO达里奥·阿莫迪等人的风险警示形成鲜明对比。
• 尽管科技行业因资源向AI倾斜正经历架构调整与裁员，但相关数据显示，今年以来科技行业岗位数量已上涨三成。

相关链接：

• 亚马逊创始人贝索斯：AI 会提升所有人的价值，无需焦虑

Bezos Says AI, Like a “Bulldozer,” Will Increase the Value of All Humanity, Dismissing Fears of AI-Driven Layoffs 🚜🤖

Briefing:

• During an interview at a Blue Origin facility, Amazon founder Jeff Bezos said that AI will improve work efficiency just as “bulldozers replaced shovels,” urging software engineers and other professionals not to feel anxious about technological change.
• Bezos believes AI will increase everyone’s value and drive a major leap in social productivity. If the spread of the technology is not restricted, humanity could enter an “era of material abundance” with more affordable food and housing.
• In response to industry concerns that “AI will lead to large-scale layoffs,” Bezos remains optimistic, in sharp contrast to risk warnings from figures such as Geoffrey Hinton and Anthropic CEO Dario Amodei.
• Although the tech industry is undergoing restructuring and layoffs as resources shift toward AI, relevant data shows that the number of jobs in the sector has risen by 30% since the beginning of this year. 📈

Related link:

• Amazon founder Jeff Bezos: AI will increase everyone’s value, no need for anxiety