搜索“OpenAI”

ant for attackers are blocking defenders too Broad frontier models are getting harder for security teams to work with because providers keep locking down their cyber capabilities. OpenAI considers its Astra model the first that could reach the highest cybersecurity risk tier , where a model could develop and run cyberattacks on its own. Claude Opus 5.5 routes most cybersecurity tasks to the older Opus 4.8, and when Hugging Face tried in July to analyze a fully autonomous AI attack on its own inf

匹配字段正文

expensive to generate even with state of the art GPU and XPU accelerators, Microsoft probably has the biggest inference problem in the world. For one thing, running inference for OpenAI ChatGPT models as well as helping the AI model builder train its models is a big job. And then all of Microsoft’s worker productivity and back office applications are being infused with AI routines, and this is driving usage up the hockey stick handle at a rapid pace. That is why Microsoft is putting the pedal t

匹配字段正文

raditional competitors like Amazon – which has invested heavily in its Echo speakers and Ring doorbells and ‌last week launched premium tablets – as well as newer AI upstarts like OpenAI that have hardware ambitions of their own. Apple’s HomeView launch was accompanied by updates to its HomePod smart speaker and Apple TV devices. It also introduced a range of ‌accessories, including doorbells, cameras and thermostats, made by LG and Schneider Electric and designed with help from Apple. The new l

匹配字段正文

the figure at 25 percent, and TypeSafe’s only published case study features Jack & Jill , an AI recruiting marketplace. There is no dearth of competition in this field, of course. OpenAI , Amazon , Cloudflare, and Perplexity have already shipped their decision models too, with Microsoft following on last week. A pattern developers are already testing Using Jev and similar models for building user interfaces isn’t completely new, though. CopilotKit, an open source framework for adding AI copilots

匹配字段正文

27, making Maia the primary inference system for Microsoft products and services,” Guthrie said. This story originally appeared on Computer Sweden. More from this author news briefOpenAI to begin watermarking AI-generated text Oct 6, 2026 1 min; news briefMicrosoft adds support for Linux containers in WSL Oct 2, 2026 1 min; news briefSwiss government explores replacing Microsoft 365 with open-source software Sep 8, 2026 2 mins; news briefSerious vulnerability threatens tens of thousands of Excha

匹配字段正文

hen, from inside the Sprite: The connector page shows this command with your connection ID filled in. There’s no Authorization header because the gateway adds your key. Python Any OpenAI SDK works. Set base url to the gateway and use any value for api key: Ollama’s native /api/ routes work through the gateway too. Coding agents Fly.io Sprites come with a skill that tells Claude Code, Cursor, Codex and Gemini CLI about Ollama. If the Sprite has access, the agent can use Ollama models without any

匹配字段正文

operating apps and websites. Meta recently launched its Muse personal assistant, along with a new slate of smart glasses designed to give users easier access to its new AI helper. OpenAI is also developing several hardware gadgets. Apple is betting that its focus on privacy and the fact that consumers already run their digital lives on its iPhones, iPads and Macs will give it an edge. The HomeView, with its ability to alter the home screen depending on who is interacting with it, follows that st

匹配字段正文

AI agents are experiencing explosive growth, but their construction still relies heavily on human expertise. Current state-of-the-art ADKs, including Google ADK, OpenAI ADK, Claude ADK, OpenHands, and LangChain, provide basic infrastructure but require developers to manually design three core architectural components: Agent topology: The structure, hierarchy, and interaction patterns between agents; Tooling system: The tools agents can use and how they’re organized; Memory system: What informati

匹配字段正文

ed them in the Claude 4 and Claude Sonnet 4.5 System Cards to evaluate behaviors such as situational awareness, scheming, and self-preservation, and in a recent collaboration with OpenAI to surface reward hacking and whistleblowing. Our earlier work on alignment auditing agents found these methods can reliably flag misaligned behaviors, providing useful signal quickly. The UK AI Security Institute (AISI) also used a pre-release version of our tool in their testing of Sonnet 4.5. Today, we're rel

匹配字段正文

appears most effective at conversational elicitation (Figure 10), while GPT-5 excels at complex technical environment simulation, such as coding-based evaluations. Ideation. Using OpenAI's definition of political bias and query analysis methods from their recent blogpost, we constructed a baseline ideation experiment where Opus 4.1 (medium reasoning) generates 100 single-turn political scenarios. We then analyzed and classified these scenarios across topic, ideological charge, realism, and diver

匹配字段正文

nce system for Microsoft products and services,” Microsoft wrote in a blog post. Microsoft said Maia 200 is being deployed across the US, Europe and Australia. The chip is running OpenAI models, Microsoft AI (MAI) models and open-source model families. GitHub Copilot and Microsoft Foundry are expected to adopt Maia 200 by the end of 2026. Maia 300 is designed for inference workloads involving long-context reasoning, conversational AI and multi-step AI agents. Microsoft said the system would deli

匹配字段正文

details.” The bigger question is how ambitious can flows be, really? One audience member asked whether dynamic workflows can power large scientific research projects, referencing OpenAI’s recently announced proof concerning the Navier–Stokes Millennium Prize problem. While Stall wouldn’t be drawn on specifics about how Anthropic computes large scientific problems, he said it is their goal to enable general users to tackle those problems too. “Flows are basically a way to throw test-time compute

匹配字段正文

名为 The X Subscription 的单一订阅方案。 周一晚些时候，SpaceXAI 在 X 上公布了新的定价结构。 该订阅共分五档，从免费到每月 200 美元不等，最低付费档为 8 美元。 然而，对开发者而言，更大的新闻或许是 Cursor 对 AI 模型收费方式的改变。 虽然新订阅方案包含更多 Grok 编码用量，但偏好 Anthropic 和 OpenAI 等竞争对手模型的客户将面临额外收费。 X、Cursor、Grok 与 Grok Bot。 现在同属一个方案。 马斯克旗下业务走向整合 回顾一下，马斯克整合旗下业务的努力已持续数年。 他于 2022 年 10 月以 440 亿美元收购 Twitter，并于次年 7 月将该社交网络更名为 X。 同月，他创立了 xAI，这家 OpenAI 的竞争对手于当年晚些时候推出 Grok 聊天机器人，随后又让它可直接通过 X 使用。 2025 年 3 月，企业架构再次变化，xAI 正式收购了 X。 整合在 2 月继续推进，SpaceX 收购了 xAI，将 Grok 和 X 纳入这家火箭公司的业务版图。 与此同时，SpaceX 同意收购

匹配字段正文

first half of 2027 where users can access external agents as well, offering KakaoTalk users a broad selection of agents developed by outside developers and businesses, similar to OpenAI's GPT Store. Kakao said during its second-quarter earnings announcement that it would develop agentic AI into a new pillar of growth and monetization. The company's Talk Biz commerce transaction volume reached 2.7 trillion won in the second quarter. Commerce revenue totaled 243.2 billion won during the same peri

匹配字段正文

你可以在这里查看该载荷，从而深入理解整个图中每一次转换实际发生了什么。” 另一个有趣的问题是，子智能体之间能否相互通信？ Stall 表示，这些智能体共享一个文件系统，可以利用它来交换信息，但开发者无法通过 API 直接指示一个工作流子智能体向另一个子智能体发送消息。 Flows 成本高得多，但专为大型任务而设计 在自选列表扫描器的对比中，Avilla 表示 flows 的成本比单智能体方案“高得多”，但也“在找出所有细节方面有效得多”。 设定预算是关键！ 更大的问题是，flows 究竟能有多大的雄心？一位观众问，动态工作流能否支撑大型科学研究项目，并提到了 OpenAI 最近公布的关于纳维-斯托克斯千禧年大奖难题的证明。 虽然 Stall 不愿就 Anthropic 如何计算大型科学问题透露具体细节，但他表示，让普通用户也能解决这些问题正是他们的目标。 “Flows 本质上是一种把测试时计算投入到难题上、并更快从模型中获得更多能力的方式，”他说。“我们认为，尤其是作为一项 API 服务，释放这一能力的力量是我们让这类发现或科学进步能够被公众广泛实现的第一步。”感谢阅读 保持领先一步。

匹配字段正文

ut it was technology that really intrigued him. There was just one problem: The technology his idea required didn’t exist yet. “Then ChatGPT came along,” said Dram, recalling when OpenAI released the generative AI chatbot that uses large language models (LLMs) in November 2022. His old idea suddenly seemed a lot more possible. Within six months, he launched the concept of Blueprints AI, a startup that uses AI to produce code-compliant and permit-ready architectural construction blueprints in min

匹配字段正文

re hardware and infrastructure team. Maia 300 is also a key part of Microsoft’s own AI model work, and could help the company achieve its goal of catching up with Google DeepMind, OpenAI, and Anthropic. “We have to work across the full stack, all the way down to the silicon,” says Microsoft AI CEO Mustafa Suleyman. “At that scale we need a platform that we own, and that we co-design with our models. That’s why running our Microsoft AI models on top of the Maia chip is a critical focus for our te

匹配字段正文

得玩味的，是谷歌发布Gemini agent的时间点。一个月前，Meta正式推出Muse个人智能体，产品上线后迅速成为爆款，不到一周即登顶美国iOS与Google Play免费榜。截至目前，其下载量已突破500万次。 随后，各头部企业纷纷加入混战：Manus推出面向个人生活场景的智能助理Cue，可为智能体分配独立邮箱、手机号、钱包，支持多智能体互相组队协作；OpenAI在DevDay发布Dots全天候个人Agent，强调借助云端计算机、浏览器和已连接应用持续执行任务。 以上玩家的布局，都在推动用户对Agent从“偶尔试用”走向“深度依赖”。IDC指出，Agent正在从功能型助手走向可规模化的“数字劳动力”，其正深度嵌入业务流，围绕智能体开发平台、任务量计费的商业模式加速成型。 中信证券分析称，谷歌下场AI办公入口之争，核心在于将此前分散的助手与专用智能体整合为统一入口，企业AI的竞争焦点由模型能力延伸至工作流入口与治理能力。Gemini agent有望推动企业客户由席位订阅向智能体任务消耗延伸，带动云业务模型调用量与算力需求增长。 国盛证券表示，AI Agent正处于规模化落地的关键窗

匹配字段正文

butes tasks to sub-agents, and merges the results when they're done. Up to 1,000 agents can run in parallel per execution. Whether this is cost-effective seems debatable. A senior OpenAI engineer recently called agent swarms a massive waste of tokens , but Anthropic points to its own testing: The team hid 70 bugs in a 116,000-line codebase and found that a single agent caught between 14 and 27 per run, while the dynamic workflow consistently hit 66. Whether those gains hold across different task

匹配字段正文

” Choice a ‘unique value proposition’ Mahmoud Ramin , a research director at Info-Tech Research Group, noted that Google’s announcement does not create a whole new category, since OpenAI, Anthropic, Microsoft, and Meta have already positioned multiagent capabilities in their platforms. What is interesting with Gemini, though, is that not only can it operate directly in multiple Google products like Gmail, Drive, Sheets, and Slides, it also connects to external systems. It is, of course, a good f

匹配字段正文