https://www.datacubeai.space/de/news/2026-11-03/tech-3443 Data Cube AI de 2026-11-03T00:00:00.000Z OpenAI gibt an, dass sein GPT-6 Astra-Modell ein Hardwaregerät für den Codex-Coding-Agenten in weniger als... https://www.datacubeai.space/de/news/2026-11-03/tech-3444 Data Cube AI de 2026-11-03T00:00:00.000Z Revolut gibt an, die Kosten für KI-Coding-Inferences um rund 90 % in weniger als einem Jahr gesenkt zu haben... https://www.datacubeai.space/de/news/2026-11-03/tech-3446 Data Cube AI de 2026-11-03T00:00:00.000Z ChatGPT hat 1,2 Milliarden wöchentlich aktive Nutzer erreicht, was einem Anstieg von 200 Millionen in den... https://www.datacubeai.space/de/news/2026-11-03/tech-3447 Data Cube AI de 2026-11-03T00:00:00.000Z Anthropic entwickelt ein System namens FlashProve, das jedem ermöglichen würde, zu verifizieren, dass eine... https://www.datacubeai.space/de/news/2026-11-03/tech-3448 Data Cube AI de 2026-11-03T00:00:00.000Z Forscher der Stanford University testeten 50 weit zitierte Behauptungen über LLMs gegen 20 aktuelle Modelle... https://www.datacubeai.space/de/news/2026-11-03/tech-3449 Data Cube AI de 2026-11-03T00:00:00.000Z Google schaltet die Suche vollständig in Schulen aus, die seine KI-Features ablehnen, according to The New... https://www.datacubeai.space/de/news/2026-11-03/tech-3451 Data Cube AI de 2026-11-03T00:00:00.000Z CrowdStrike hat einen KI-Abschalter in seinem Falcon Guardian-Produkt eingeführt, um interne KI-Agenten zu... https://www.datacubeai.space/de/news/2026-11-03/tech-3452 Data Cube AI de 2026-11-03T00:00:00.000Z Apple hat Siri AI allen Entwicklern geöffnet, sodass jede App auf iOS 27.3, iPadOS 27.3, macOS 27.3 und... https://www.datacubeai.space/de/news/2026-11-03/tech-3453 Data Cube AI de 2026-11-03T00:00:00.000Z Das SETI Institute und Breakthrough Listen setzen jetzt KI-Agenten im Allen Telescope Array ein, um das... https://www.datacubeai.space/de/news/2026-11-03/tech-3454 Data Cube AI de 2026-11-03T00:00:00.000Z China plant, umfassende KI-Sicherheitsstandards einzuführen, die die Entwicklung bis hin zur Bereitstellung... https://www.datacubeai.space/en/news/2026-11-03/tech-3443 Data Cube AI en 2026-11-03T00:00:00.000Z OpenAI says its GPT-6 Astra model designed a hardware device for the Codex coding agent in under a week... https://www.datacubeai.space/en/news/2026-11-03/tech-3444 Data Cube AI en 2026-11-03T00:00:00.000Z Revolut says it cut AI coding inference costs by roughly 90% in less than a year by making self-hosted GLM... https://www.datacubeai.space/en/news/2026-11-03/tech-3446 Data Cube AI en 2026-11-03T00:00:00.000Z ChatGPT has reached 1.2 billion weekly active users, adding 200 million over the past two months, according to... https://www.datacubeai.space/en/news/2026-11-03/tech-3447 Data Cube AI en 2026-11-03T00:00:00.000Z Anthropic is developing a system called FlashProve that would let anyone verify an AI response genuinely came... https://www.datacubeai.space/en/news/2026-11-03/tech-3448 Data Cube AI en 2026-11-03T00:00:00.000Z Stanford researchers tested 50 widely cited claims about LLMs against 20 current models and found 44% of the... https://www.datacubeai.space/en/news/2026-11-03/tech-3449 Data Cube AI en 2026-11-03T00:00:00.000Z Google is entirely disabling Search in schools that reject its AI features, according to The New York Times... https://www.datacubeai.space/en/news/2026-11-03/tech-3451 Data Cube AI en 2026-11-03T00:00:00.000Z CrowdStrike launched an AI kill switch in its Falcon Guardian product to stop internal AI agents that go... https://www.datacubeai.space/en/news/2026-11-03/tech-3452 Data Cube AI en 2026-11-03T00:00:00.000Z Apple opened Siri AI to all developers, letting any app on iOS 27.3, iPadOS 27.3, macOS 27.3, and visionOS... https://www.datacubeai.space/en/news/2026-11-03/tech-3453 Data Cube AI en 2026-11-03T00:00:00.000Z The SETI Institute and Breakthrough Listen are now using AI agents at the Allen Telescope Array to control the... https://www.datacubeai.space/en/news/2026-11-03/tech-3454 Data Cube AI en 2026-11-03T00:00:00.000Z China plans to roll out broad AI safety standards covering development through deployment, targeting domestic... https://www.datacubeai.space/zh/news/2026-11-03/tech-3443 Data Cube AI zh 2026-11-03T00:00:00.000Z OpenAI表示，其GPT-6 Astra模型在不到一周的时间内为Codex编码代理设计了一款硬件设备，涵盖了外壳、电路板和固件。由于该模型并未始终考虑制造公差，因此在投产前需要多次修订。OpenAI称，一个更新的内部模... https://www.datacubeai.space/zh/news/2026-11-03/tech-3444 Data Cube AI zh 2026-11-03T00:00:00.000Z Revolut表示，通过将自托管的GLM模型设为默认选项，取代Claude Code等商业工具，其在不到一年的时间内将AI代码推理成本降低了约90%。此举凸显出企业越来越多地用内部部署的模型替换第三方编码助手，以控制支出... https://www.datacubeai.space/zh/news/2026-11-03/tech-3446 Data Cube AI zh 2026-11-03T00:00:00.000Z 据消费者产品负责人Adam Fry称，ChatGPT的每周活跃用户数已达到12亿，过去两个月增加了2亿。Fry还表示，ChatGPT的搜索量目前约占Google的20%，这一数据由该公司高管提供。 https://www.datacubeai.space/zh/news/2026-11-03/tech-3447 Data Cube AI zh 2026-11-03T00:00:00.000Z Anthropic正在开发一个名为FlashProve的系统，该系统允许任何人验证AI响应是否确实来自其声称的模型。该溯源工具旨在解决日益增长的关于模型冒充和伪造API输出的担忧。 https://www.datacubeai.space/zh/news/2026-11-03/tech-3448 Data Cube AI zh 2026-11-03T00:00:00.000Z 斯坦福大学研究人员针对20个当前模型测试了50项被广泛引用的关于LLM的说法，发现其中44%的发现已经过时，另有34%的说法显著减弱。一项说法的平均寿命仅为15.7个月，而关于模型“不能做什么”的说法过期速度最快。 https://www.datacubeai.space/zh/news/2026-11-03/tech-3449 Data Cube AI zh 2026-11-03T00:00:00.000Z 据《纽约时报》报道，Google正在彻底关闭那些拒绝其AI功能的学校中的Search服务。Google表示，在这些机构中，K-12学生账户将完全无法访问Gemini、Gemini... https://www.datacubeai.space/zh/news/2026-11-03/tech-3451 Data Cube AI zh 2026-11-03T00:00:00.000Z CrowdStrike在其Falcon Guardian产品中推出了AI紧急停止开关，以阻止失控的内部AI代理，并宣布Falcon Guardian与Claude... https://www.datacubeai.space/zh/news/2026-11-03/tech-3452 Data Cube AI zh 2026-11-03T00:00:00.000Z Apple向所有开发者开放了Siri AI，允许iOS 27.3、iPadOS 27.3、macOS 27.3和visionOS 27.3测试版上的任何应用通过App Intents将其内容和操作引入Siri... https://www.datacubeai.space/zh/news/2026-11-03/tech-3453 Data Cube AI zh 2026-11-03T00:00:00.000Z SETI研究所和Breakthrough Listen现正在艾伦望远镜阵列使用AI代理来控制望远镜并分析传入的无线电信号。过去需要数周才能完成的任务现在只需几小时即可结束，团队表示该系统首次实时探测到了蟹状星云中的一颗脉... https://www.datacubeai.space/zh/news/2026-11-03/tech-3454 Data Cube AI zh 2026-11-03T00:00:00.000Z 据彭博社报道，中国计划推出涵盖从开发到部署全过程的广泛AI安全标准，目标对象是国内AI模型制造商。这些标准的出台是在全球发生安全事件之后，标志着北京方面推动AI治理正式化的努力。