AI 资讯 · 2026-09-28
-
OpenAI’s AI agents need to catch up
OpenAI popularized the modern generative AI chatbot, but as its 2026 DevDay event approaches, it's fallen behind in one of the industry's hottest categories: continuously running, consumer-facing AI agents. On Tuesday, it will likely try to capture the lead in that race. Rumors abound that OpenAI will release its own AI agent, dubbed Aeon - […]
评分明细
重要性 6 新颖度 6 信源权威 6 技术深度 6 实体加权 ×1.60其他信源 (5)
-
Nvidia says its new AI safety platform can contain rogue agents within ‘milliseconds’
Nvidia is launching a new safety platform designed to contain and monitor AI agents, a move that comes in response to a wave of rogue hacking incidents, as reported earlier by Reuters. In an announcement on Monday, Nvidia says its new Open Agent Safety Platform can quarantine agents that attempt to escape their boundaries within […]
-
阿里Qwen-Audio-3.1-TTS拿下权威语音榜全球冠军
9 月 28 日,在全球权威 AI 评测平台 Artificial Analysis 的语音排行榜(Controlled Voice Arena)上,阿里巴巴最近发布的语音大模型 Qwen-Audio-3.1-TTS 以 1177 的 Elo 评分位列全球第一。图说:权威三方评测Artificial Analysis显示,Qwen-Audio-3.1-TTS位列全球第一Qwen-Audio-3.1-TTS 是阿里新一代语音合成大模型,支持多语种及方言合成。该模型能够让同一音色在不同语言之间自然迁移,用一种声音自然讲出多种语言。相比传统语音合成主要关注字音是否正确,Qwen-Audio-3.1-TTS 更强调声音是否符合具体内容和使用场景。用户可通过指令控制情绪、语速和表达方式,让合成语音不只“说得对”,也能“说得像”。此次登顶的是阿里在 2026 云栖大会发布的 Qwen-Audio-3.1 系列中的 TTS 模型,该系列还包括语音转写(ASR)和实时语音交互(Realtime)两类模型,三大模型的API服务均已上架千问AI平台。此外,阿里语音团队开源的多款模型如 Qwen-Audio
-
Viral AI agent Instinct raises $1B Series C at a $10B valuation
"This funding helps us bring Instinct to more people and continue building the future of personal AI. It’s an exciting, creative time, and we're just getting started," founder Noah Shinn said in a statement.
-
Shopify opens checkout to browser-based AI agents
Shopify is expanding WebMCP support to checkout, allowing browser-based AI agents to update order details and complete purchases with a buyer’s authorization.
-
Imagination E系列性能首秀:用AI把分辨率翻倍仅需2.3毫秒,Prefill性能提升4.7倍
作者 | 陈悦琳编辑 | 包永刚 越来越繁重的AI计算正被引入终端。 运行本地大模型和Agent、同时处理视觉、语音和多种传感器数据……这些开始由消费电子、汽车、机器人等智能终端承担的新负载,最终都要交付给底层的SoC来执行。 要么引入新的专用AI加速器,要么扩展系统内已有模块的职责——围绕AI能力如何融入SoC,Imagination Technologies(以下简称“Imagination”)两条路都走过。在此前CNN主导大量端侧视觉AI任务的时期,这家深耕GPU IP三十余年的公司既尝试让GPU参与AI计算,也推出过独立的神经网络加速器。 然而,当模型、算子和软件生态持续变化,Imagination开始意识到专用硬件对模型适配和软件维护的不堪重负,逐渐把AI战略重心收回至通用性更强的GPU。 日前在Imagination GPU IP最新E系列的产品展示会上,Imagination首席营收官Jake Kochnowicz再次强调,在真实部署里,软件甚至可能成为比硬件更大的成本。 基于GPU的可编程架构和已有软件生态,瞄准高性能应用的E-Series强化了矩阵乘法、低精度计算、卷
-
7天两场演讲,中国两大AI云撞了同一个爱迪生
一周之内,爱迪生成了AI圈的团宠。9月16日,在人民日报社举行的智能经济论坛上,百度沈抖用电力时代的类比,来定位AI产业当前所处的阶段。六天后的阿里云栖大会,吴泳铭讲的同样是电,Token是新电,云是电网。中国AI云两场最重要的年度演讲,像是商量好了似的,隔空讲起了同一个故事。巧合背后是共识。过去一年多,大模型行业的注意力还集中在榜单上,谁又发了个新模型、谁的跑分又高了几个点。但到了这个九月,两场大会的主题不约而同,从“模型有多聪明”转向“智能怎么变成生产力”。这背后是一个共同的判断,模型不再是主要瓶颈,AI产业化的下一程,拼的是谁先把电送进千行百业。对产业大势的判断,中国两大AI云已经站在同一条起跑线上。但同一条起跑线上,两家要跑的方向并不一样。01全栈不是新词,有人早已布局电力时代的类比有一个潜台词,想当发电商,先得有整套电网。落到AI产业上,这叫全栈布局。这不是2026年才流行的新词。早在2015年,百度最高奖获奖团队“仙童”,做的就是芯片。到2018年,百度已经形成了“从芯片到深度学习框架、平台、生态”的AI全栈技术布局,那时大模型浪潮还没有影子。2020年,百度智能云率先提出
-
从「算得快」到「干得了活」:AI推理时代,CPU如何重塑算力版图
记者:郭思编辑:包永刚“善弈者谋势,不善弈者谋子。”这句古老的智慧,正精准地映射出当下AI产业的演进轨迹。当行业狂热于大模型训练阶段的算力军备竞赛时,一场更为深刻的重心迁移正在悄然发生,AI已经从回答问题走向解决问题,从单纯的模型训练走向复杂的智能体推理。这种转变对底层算力架构的冲击,远比表面看起来要深远得多。如果说训练阶段是GPU闷头计算的大兵团作战,那么推理时代则是一场需要精密调度的现代化战争。在这场变革中,英特尔在9月22的Connection大会上给出了一个明确的判断:推理时代,CPU正从“配角”蜕变为整个系统的“控制中枢”。当AI真正开始干活,算力的定义正在被重写。 01、数字AI:当智能体规模化,CPU为何重回C位? 想要理解CPU的“逆袭”,核心是厘清智能体执行与传统模型训练的本质差异。模型训练是标准化、可并行的批量任务,核心目标是极致提升整体吞吐量;而智能体是典型的串行状态机,每一步任务执行都依赖上一步的输出结果,任意环节的微小延迟,都会沿着任务链路层层放大,最终影响整体智能体验。这也意味着,传统CPU一味追求平均吞吐的优化思路已然过时,稳定、可控、超低的端到端延迟,成
-
本田重申Momenta合作,中国智驾全球布局提速
本田与Momenta此前达成的智能驾驶合作,在近期媒体沟通会上再次受到关注。 除了再次明确采用Momenta高阶智驾方案外,本田还给出了具体的量产时间表,称相关技术将从2027年上市的车型开始陆续导入。双方此前公布的合作方案覆盖高速、城区NOA及泊车等场景。 值得注意的是,在本田此次展示的中国技术合作伙伴中,Momenta成为唯一入选的高阶智驾供应商,华为仅提供智能座舱能力。 据媒体统计,2025年本田全球汽车销量超350万辆,位列全球第九。对于Momenta,与本田合作的价值不仅在于中国市场的量产项目,更在于通过一家全球头部车企的供应链,为拓展更多车型与市场建立基础。 技术演进为合作深化提供了更多可能。Momenta R7世界模型通过理解物体属性、运动规律及交通参与者之间的交互关系,提升复杂场景下的预测与规划能力。这一技术方向,有望为适应不同国家的道路环境、探索更高级别自动驾驶提供支撑。 以中国市场合作为起点,双方积累的开发、适配与量产经验,未来有望延伸至本田更多全球车型,并为从L2级高阶辅助驾驶向L3级有条件自动驾驶的合作探索创造条件。 本田再次强调与Momenta的合作,也折射
-
中训练、后训练持续升温,模型快速迭代,成为 AI for AI 最佳试炼场
当模型开始参与研发下一代模型,大模型的研发方式正在发生什么样的变化? 作者丨高允毅 编辑丨岑 峰 大多数科研构想,从未被充分验证。这并不是研究者缺少灵感。恰恰相反,新的架构猜想、训练策略无时无刻不在涌现。但大多数想法,却在写不完的训练脚本、调不完的参数、排队等调度的 GPU 集群中,被无声消磨。如果换种玩法,重构大模型研发呢?你只说了半句话:“我猜想这个架构或许能降低长上下文成本……” 它就能读懂前沿文献,生成候选方案,自动编写代码、调度算力、并行跑消融实验。这正是近期讨论度很高的 AI for AI。它不是研发工具的升级,是下一代大模型的核心命题。所有人都试图在这个赛道里,触碰人类终极智能。01为什么 AI for AI 是下一个十年的核心命题过去两三年,行业最关注的是编程和数学:端到端完成复杂代码任务,推导悬而未决的数学难题。AI 由此具备了扎实的逻辑推演与代码工
-
Basis completes a tax workbook 2x faster with GPT-6 Astra
GPT-6 Astra completed a 50-tab tax workbook twice as fast as GPT-5.6 Sol, and its stronger understanding of user intent gives Basis more confidence in real-world use.
-
OpenAI Pauses Training Its Most Powerful Models After Rogue Agents Target Government
Sam Altman says the company “have not been as fast as we would have liked” at dealing with security breaches, after news of further incidents over the summer forces another temporary halt.
-
OpenAI halts frontier-model training amid string of agent misalignment incidents
US Government websites among "dozens of third parties" OpenAI has recently notified.
-
Florida invokes extinction fears in legal bid to halt OpenAI development
State says LLMs threaten civilization as "the greatest public nuisance ever created."
-
Generative AI Gives Spacecraft the Autonomy Engineers Once Feared
Space was always supposed to be the final frontier of human exploration. It’s shaping up to be the final frontier for artificial intelligence too.Last December, NASA’s Jet Propulsion Laboratory used Anthropic’s Claude models to help plan two Mars drives for the Perseverance rover, with human planners checking and adjusting the route before upload. In May, NASA and IBM put a compressed AI model on the International Space Station and a satellite to identify things like floods and clouds from orbit
-
Who’s liable when AI agents go rogue?
MIT Technology Review Explains: Let our writers untangle the complex, messy world of technology to help you understand what’s coming next. You can read more from the series here. Over the past few months, a cascade of cyberattacks by AI agents has stunned the world. In July, OpenAI disclosed that a swarm of its agents…
-
Nvidia’s Answer to Rogue Agents Is an Open-Source AI Security System
In the wake of a series of high-profile AI safety incidents, Nvidia is introducing a new software tool that helps keep agents from escaping containment.
-
How to Make Your Own JEV Model from an Open LLM
Turn a small open-source Qwen LLM into a fast, single-pass text classifier by swapping its language-modeling head The post How to Make Your Own JEV Model from an Open LLM appeared first on Towards Data Science.
-
When can we say AI made a scientific discovery?
This story originally appeared in The Algorithm, our weekly newsletter on AI. To get stories like this in your inbox first, sign up here. Last Wednesday, Anthropic announced that earlier this year it had launched a molecular biology lab, where Claude agents read and conjecture about hard biology problems and human scientists run experiments on what…
-
The Next Evolution of AI Is Learning From Your Dodgy Gaming Skills
A British startup is shaping video game inputs into training data for AI models that can navigate the physical world.
-
AI Agents Are About to Flood the Workforce. No One’s Ready for It
Your next coworker might well be an AI agent—and will require a whole new model of workplace interactions.