OpenAI

March 25, 2011 - Technology media outlet NeoWin published a blog post today (March 25) reporting that OpenAI has revamped ChatGPT's Advanced Voice Mode with a focus on optimizing conversation fluency. The new version reduces unnecessary interruptions, adds a pause feature, and strengthens voice personalization for paid subscribers, a feature that is available to all users, with limits on the free version. This update fixes the problem of "accidental interruptions" previously reported by users, and iPhone users are advised to turn on V...

3 月 25 日消息，据雷峰网今日援引多位业内人士消息，去年下半年，微软云 OpenAI 业务的全球营收中，来自中国市场的占比最高，约为 1/4。 报道提到，这些客户包括大模型公司、有大模型业务的互联网巨头以及出海中企等。 报道还称，过去两年，中国市场上，Azure OpenAI Service 等相关服务快速拉动着微软云中国的营收，目前已成为其新的增长引擎。 1AI查询获悉，微软 2025 财年…

March 24 - Frequent use of chatbots such as ChatGPT may be associated with increased loneliness and less time spent socializing, according to a new collaborative study by OpenAI and the Massachusetts Institute of Technology (MIT). The study found that those who spent more time per day communicating with ChatGPT by text or voice tended to exhibit higher levels of emotional dependence on chatbots, more frequent inappropriate use behaviors, and stronger feelings of loneliness. 1AI notes that the findings have not yet been peer-reviewed, but have raised concerns about the potential emotional impact of chatbots. ...

March 22, 2011 - OpenAI, in conjunction with the MIT Media Lab, shared the first study on how using ChatGPT affects people's emotional well-being in a blog post published yesterday (March 21), which found that factors such as how users interact with AI, their personal emotional needs, and how long they use it affect their emotional well-being. The study found that factors such as how users interact with AI, their personal emotional needs, and the length of time they use it, all affect their emotional state. Through large-scale data analysis and randomized controlled trials, the study reveals the complexity and limitations of AI emotional use, providing an important reference for the responsible development of future AI platforms. OpenAI...

March 21 news, OpenAI yesterday (March 20) released a blog post, announcing the launch of speech-to-text (speech-to-text) and text-to-speech (text-to-speech) models, to improve voice processing capabilities, support developers to build more accurate, customizable voice interaction system, and further promote the commercialization of AI voice technology applications. In terms of speech-to-text models, OpenAI has launched gpt-4o-transcribe and gpt-4o-mini-transcribe...

3 月 20 日消息，近日，人工智能公司 OpenAI 旗下的热门聊天机器人 ChatGPT 因生成虚假信息而面临又一起隐私投诉。此次事件发生在欧洲，由隐私权益倡导组织 Noyb 支持挪威一名个人发起，该个人发现 ChatGPT 生成了虚假信息，声称他因谋杀两名子女并企图杀害第三名子女而被定罪。这一事件引发了对 OpenAI 是否违反欧盟《通用数据保护条例》（GDPR）的质疑。 1AI注意到，此前…

3 月 20 日消息，OpenAI 昨日（3 月 19 日）在 X 平台发布推文，宣布通过开发者 API，正式推出 o1 系列升级版“o1-pro”，宣称其通过更高计算资源投入实现“更一致且优质的回应”。 o1-pro 的核心优势，在于大幅提升计算资源，让其更游刃有余地驾驭复杂问题，不过这款新模型仅向特定开发者开放（需在 API 服务中至少消费 5 美元），且定价远超同类产品。 根据官方公布的定价…

March 18 - Liam Fedus, OpenAI's vice president for post-training research, has announced that he's leaving the company to start an AI startup focused on materials science. 1AI noted that Fedus posted a statement on the X platform saying, "My undergraduate degree was in physics, and I've always aspired to apply AI techniques to that field." He also mentioned that since AI in science is one of the most strategic areas for OpenAI to realize general artificial intelligence (AGI), OpenAI plans to invest in his new company...

3 月 14 日消息，科技媒体 MacRumors 昨日（3 月 13 日）发布博文，报道称 OpenAI 公司已向美国政府提交一份 AI 发展建议书，呼吁放宽监管，让 AI 公司自由使用受版权保护的材料，训练 AI 模型。 OpenAI 在建议书中认为，当前美国各州实施过多的法律限制，阻碍了 AI 公司的创新能力，并削弱了美国企业家可用的训练数据质量。 OpenAI 提出“版权战略”，主张能自由…

March 12, 2011 - OpenAI released a blog post yesterday (March 11) that unveiled a new set of tools for developers to help simplify the development process and enhance the skills and capabilities of AI intelligences as they are created. This release marks a significant step towards making OpenAI's API platform more useful and relevant for real-world applications. A brief overview of the new developer tools in this release: Responses API: This API combines chat functionality with integrated tools such as web search and file search, based on...

March 11, OpenAI today announced its latest research, trying to use CoT (chain of thought) monitoring way to monitor the reasoning model's "ideas", so as to stop the AI big model nonsense, hide the real intention and other behaviors, at the same time, this is also one of the effective tools to supervise the super model. Note: AI will occasionally talk nonsense, which we generally call AI hallucinations. In addition to this, AI can be lazy or deliberately bad, such as tricking the user or simply giving up when a problem is too difficult. OpenAI uses the newly released cutting-edge model o3-mini as the monitored...

March 10, 2012 - How do you prove you're human in the age of AI? OpenAI founder Sam Altman's new startup, World Network, wants to find out. According to the Wall Street Journal, in an effort to challenge Elon Musk, Altman is developing an "all-in-one app" called World. The innovative platform aims to create a comprehensive ecosystem of mega-apps that integrate payments, chat, and microloans, all verified through a unique biometric method: users must gaze at a basketball-sized...

March 7, 2011 - At the end of March last year, OpenAI announced a "small preview" of an artificial intelligence service called Voice Engine, claiming that the technology would be able to clone a person's voice in as little as 15 seconds of speech. Nearly a year later, however, the tool has yet to be officially launched, and OpenAI has yet to say if and when it will go live. OpenAI's cautious approach to the Voice Engine could be rooted in concerns about misuse of the technology, or it could be an attempt to avoid regulatory scrutiny...

March 7, 2011 - Last September, OpenAI released the o1 family of AI models that can think deeply before responding. Last October, Microsoft announced a new Copilot feature -- Think Deeper for Pro plan users, which uses OpenAI's o1 inference model to help users solve tough problems, which has been opened to users this year for free and without restrictions. Microsoft today announced enhancements to the Copilot Think Deeper feature, now...

March 7, 2011 - OpenAI announced in a tweet on the X platform yesterday (March 6) that it has upgraded its ChatGPT app for macOS to simplify the coding process for developers by enabling them to edit code directly in integrated development environments (IDEs) such as Xcode and VS Code. The update to ChatGPT's macOS app allows users to interact and code directly with ChatGPT in Xcode and other IDEs through the "Collaborate with Apps" feature. This...

March 6, 2011 - OpenAI launched a "research preview" of its GPT-4.5 AI model on February 28th, claiming more natural interactions, a broader knowledge base, a better understanding of user intent, and a higher "emotional intelligence". OpenAI today announced that it has launched GPT-4.5 to ChatGPT Plus users earlier than expected. According to the official introduction, GPT-4.5 is first open to ChatGPT Pro users, and will be extended to Plus and Team users from now on, and then to ...

March 2, 2011 - Technology media outlet The Decoder published a blog post yesterday (March 1) reporting that OpenAI's newest model, GPT-4.5, has a significant increase in cost with limited performance improvement, prompting the industry to question its price/performance ratio. According to the official performance comparison data provided by OpenAI, the improvement of GPT-4.5 over GPT-4o is 6.8-13.2 percentage points: Professional query: GPT-4.5 outperforms GPT-4o at 63.21 TP3T. Sun...

March 1, 2011 - Tech media outlet smartprix published a blog post yesterday (February 28) reporting that OpenAI accidentally leaked the number of GPUs that may be needed for GPT-6 training in a video introducing the GPT-4.5 model, suggesting that it will be much larger than ever before. Note: At the 2:26 mark of the GPT-4.5 model introduction video, the chat transcript of OpenAI's demonstration of GPT 4.5's capabilities includes the phrase "Num GPUs for GPT 6 Training"...

Feb. 28, 2012 - OpenAI CEO Michael Altman said today that the company had to roll out its latest GPT-4.5 model in stages due to a "shortage of GPU resources. In a post on the X platform, Altman noted that GPT-4.5 is a "huge" and "expensive" model, requiring tens of thousands of additional GPUs to make it available to more ChatGPT users. GPT-4.5 will be available to ChatGPT Pro subscribers first, followed by ChatGPT Plus subscribers next week...

Feb. 28 - CNBC reported on Feb. 27 that Meta plans to launch a standalone AI app in Q2 this year as part of Zuckerberg's plan to overtake OpenAI and Google by the end of the year. Unlike rival AI tools like ChatGPT and Perplexity, Meta AI is currently limited to use through its website and its apps like Facebook and WhatsApp. While Meta's large user base can access these apps through...

Feb. 28 - OpenAI published a blog post today (Feb. 28) announcing the official launch of the GPT-4.5 AI model, claimed to be its biggest and best chat model. As a research preview, GPT-4.5 will first be available to ChatGPT Pro users. GPT-4.5 improves unsupervised learning by scaling up pre-training and post-training, making significant advances in understanding patterns, making connections, and generating ideas. Early tests show that interaction with GPT-4.5 is more natural, the knowledge base is broader, and more...

Feb. 26, 2011 - OpenAI today (Feb. 26) announced in a tweet on the X platform that it is opening up the ChatGPT Advanced Voice Mode for free, allowing more users to experience more natural voice conversations. The mode is based on the GPT-4o mini model, which is optimized for computational efficiency to match the performance of the full GPT-4o model. Previously available only to paid subscribers, this feature is now free for all users to access by clicking on the voice icon in the lower right corner of the screen...

February 26, 2011 - Earlier this year in February, OpenAI launched its Deep Research feature, promising that it would be available to Plus users in about a month. Now, the company has kept that promise. Starting today, Deep Research is now available to ChatGPT Plus, Team, Edu, and Enterprise users. Previously, users were required to subscribe to the $200 per month (note: currently around RMB 1,451) Pro plan to experience this feature. Currently...

Feb. 25, 2012 - OpenAI has updated the ChatGPT app with a new Safari extension that allows users to use ChatGPT as the default search engine in the address bar of the Safari browser, providing users with a more convenient AI search experience. After updating to the latest version of ChatGPT app, users can enable "ChatGPT Search Extension" in the Safari section of "Settings" app. Once enabled, search queries entered in the Safari address bar will be...