Learn how to deploy self-hosted computer vision models for privacy, low latency, and offline use. See how Roboflow Inference simplifies local deployment

Jane Street has installed Etched's first shipped AI cluster system, and was so impressed, it led another massive round, the startup says.
Uniform Manifold Approximation and Projection (UMAP) is a dimensionality reduction technique widely used for visualization and feature extraction. Applications... Uniform Manifold Approximation and Projection (UMAP) is a dimensionality reduction technique widely used for visualization and feature extraction. Applications range across exploratory data analysis, topic modeling, and single-cell analysis. Many of these workflows are iterative and exploratory, requiring UMAP to be run repeatedly as u…

Apple’s leaked camera-equipped AirPods might avoid the privacy pitfalls of other AI wearables by preventing users from recording photos and videos.
An artificial intelligence startup called Hypercubic Inc. is taking on the problem of modernizing legacy applications written in the COBOL programming language after raising $5.3 million in seed funding today. The round was led by CIV and saw participation from Y Combinator, Afore Capital, Pioneer Fund, Multimodal Investors and angels including Opendoor Chief Executive Kaz […] The post Hypercubic raises $5.3M to map out and rewrite legacy COBOL apps with AI agents appeared first on SiliconANGLE.
LMSYS 团队针对 1.6 万亿参数的 MoE 模型 DeepSeek-V4-Pro,在 H20 GPU 上通过场景化服务配置逼近 B300 性能。单节点 H20-141GB 参考实现达 271 output tokens/s,与 B300 的 383.7 tokens/s 性能差距缩小至 1.42×。 🔗 阅读原文 via AIHOT · https://aihot.virxact.com/items/cmt0e7sxo01eiro2o4l4y9jmh
Mozilla Corp. rolled out updates today to its popular Firefox browser, updating an artificial intelligence assistant called Smart Window, currently in beta, that lets users check answers, summarize open tabs and retrace search history. Smart Window is the evolution of an earlier concept introduced by the company last November, called AI Window, which explored how privacy-first AI […] The post Firefox updates Smart Window, a new in-browser AI assistant that can keep up with you appeared first on …
The financing and infrastructure role puts the AI vendor at the center of another multibillion-dollar data center project.

2026年8月18日,奕境X9预售发布会在成都盛大启幕。作为东风与华为乾崑全栈合作打造的新能源汽车品牌,奕境首款车型——第二代家庭旗舰SUV奕境X9正式开启预售,预售价29.98万起。 当下,多成员家庭出游意愿越来越强烈,家庭消费者购车需求也随之升级,希望进一步享受到最前沿的智能科技、座舱体验和安全防护。奕境X9倾注了东风的深厚造车底蕴与华为乾崑领先的汽车智能化解决方案,全面定义第二代家庭旗舰。 华为公司高级副总裁、引望公司CEO靳玉志表示,第一代家庭旗舰是“移动的家”,有冰箱、彩电和大沙发,第二代家庭旗舰不仅是它的“加强版”,还要配置两大“智慧伙伴”——全家人的专属司机和贴心助理。对此,奕境X9满配华为乾崑全栈六大汽车智能解决方案:华为乾崑智驾、鸿蒙座舱、华为乾崑赤兔平台、华为乾崑车载光、华为乾崑车云和华为鲸鳍通信,全部实现上车。 全家人的专属司机:华为乾崑智驾ADS 5全系标配,打造辅助驾驶新标杆 对于家庭出行强烈的安全性需求,靳玉志表示,奕境X9的专属司机拥有“聪明的大脑”和“灵活的小脑”,前者来自华为乾崑智驾ADS 5,负责感知路况、规划路线,自主做出辅助驾驶决策;后者来自华为…
On Tuesday, Warp introduced Warp Factories, a new infrastructure system designed to make building AI software factories as easy as possible.
ChatGPT for Teens adds age-appropriate safety measures, parental controls, and learning tools designed to steer teens away from harmful content — and from using AI to cheat on their homework.
Perplexity's India revenue rose about 60% after the Airtel offer ended for new users, even as downloads declined.
8月18日,百度发布最新季度财报,显示总营收313亿元;百度一般性业务收入252亿元,其中AI业务收入占比50%,连续两季度过半。同期,华尔街两大投资基金大幅建仓、增持百度。 据13F持仓文件,传奇投资人德鲁肯米勒旗下 Duquesne Family Office二季度建仓百度ADR 8.8万股,为2023年四季度清仓阿里巴巴以来首次重返中概股。此外,知名对冲基金经理大卫·泰珀旗下 Appaloosa LP 在清仓两支中概股的同时,增持百度超60万股。
Toronto patent research startup NLPatent rebranded as Clerq today and launched its first agentic workflows. The software runs a full patentability analysis in about 10 minutes. The company says every conclusion arrives with citations. That work normally takes days or weeks, and firms usually push it down to a junior associate or out to a search […] The post NLPatent rebrands as Clerq, launches agentic patent research workflows appeared first on SiliconANGLE.
This article is part of our exclusive IEEE Journal Watch series in partnership with IEEE Xplore. Microspines are one of many ways to enable robots to latch onto surfaces like walls and ceilings. Now roboticists in Canada are using the mini spikes to get drones to land on a more challenging, remote surface: icebergs. Like a spider, the Ice Dart can land on and latch onto steep, slippery surfaces such as icebergs and glaciers—an increasingly useful capability as activity in the Arctic increases. T…

This is today’s edition of The Download, our weekday newsletter that provides a daily dose of what’s going on in the world of technology. We still don’t know how people are really using AI AI companies like Anthropic and OpenAI regularly publish reports on how people are using their products. But they only release the…
A startup called Synthefy Inc. said today it’s going to for numbers what large language models did for words after raising $6.5 million in seed funding today. The funds will help to expand its new foundation-model platform that’s fine-tuned specifically for numerical data rather than text. The round was led by Wing Venture Capital and saw […] The post Exclusive: Synthefy raises $6.5M for its number-crunching models trained on numerical data instead of words appeared first on SiliconANGLE.
ChatGPT for Teens helps teens learn, think critically, and use AI with confidence, with stronger built-in protections, healthy-use features, and additional controls for parents.
OpenAI is strengthening monitoring, alignment, and security for frontier AI models. See how new safeguards are guiding the pace of model development.
OpenAI and CodeAI are partnering to help students build AI literacy, think critically about AI, and develop the skills to use and shape it responsibly.
OpenAI 发布 ChatGPT for Teens,为 13-17 岁用户自动启用,内置更强安全保护与家长控制。新增 Study Mode、负责任作业提醒、测验与学习可视化,以及可设定默认开启时段的 Study Hours,引导青少年分步解题而非直接给答案。OpenAI 同时宣布与 CodeAI 合作,帮助青少年理解、质疑并创造性地使用 AI。 🔗 阅读原文 via AIHOT · https://aihot.virxact.com/items/cmsykcnkt0yh1roz0a9h404r5
OpenAI 因 OpenAI-Hugging Face 事件及即将推出的 Astra 模型可能达到《预备框架》下的"关键网络安全能力"阈值,暂时放缓了模型扩展速度,包括暂停最新部署模型的强化学习训练两周,并搁置最大规模前沿 RL 运行。公司已加强研究环境安全,要求对 Astra 及网络相关负载实施最严格防护,并扩展思维链监控,采用多阶段激活分类器检测机制。 🔗 阅读原文 via AIHOT · https://aihot.virxact.com/items/cmsz0hbsg05l0ro204080cotv
经典大奖致敬AI先驱,14位青年学者站上全新焦点舞台,713篇主会论文把脉未来演进图景。 作者丨岑峰 马晓宁 当地时间 2026 年 8 月18日,第 35 届国际人工智能联合会议(IJCAI-ECAI 2026)在德国不莱梅拉开帷幕。作为历史最悠久、最具权威性的人工智能顶级学术会议之一,IJCAI 见证了 AI 半个多世纪的起伏与繁荣。今年,大会重回欧洲,不仅汇聚了全球顶尖的 AI 学者、产业领袖,更在 AGI 狂飙突进的当下,为全行业提供了一个反思基础理论、探讨技术落地的绝佳平台。 AI科技评论小分队已抵达现场,为您发回开幕式及颁奖典礼的最新报道。 01 缅怀先驱与东道主致辞:AI 发展的传承与未来 开幕式伊始,大会主席 Francesca Toni 登台致辞,正式宣布 IJCAI-ECAI 2026 开幕。 本届大会受到了德国当地政府与科研机构的高规格重视。德国联邦研究、技术与空间部的 Alexandra-Gwyn Paetz 博士,以及不莱梅市长、参议院议长 Andreas Bovenschulte 分别发表了热情洋溢的欢迎致辞。Bovenschulte 的出席不仅展现了东道主…
AI companies like Anthropic and OpenAI regularly publish reports on how people are using products like Claude and ChatGPT, but they only release the data they want us to see, AI researchers say. “There is no independent source to corroborate it,” says Anka Reuel, a computer science PhD candidate at the Stanford Trustworthy AI Research…
过去几年,大模型竞争的主线很清晰:更大的模型、更多的数据和更强的GPU集群。 但随着预训练边际收益递减,模型能力Scaling正从单纯的预训练堆叠向后训练强化学习(RL)转变,数学推理、代码生成、复杂决策、长时序Agent等高阶能力,越来越依赖后训练强化学习(RL)激发。RL通过生成、执行、反馈和奖励,让模型持续学习推理、规划与自我纠错的能力。 由此,大模型发展从“一次性训练”进入“持续训练”,模型能力Scaling也从预训练延伸至生成、反馈和迭代全过程。SemiAnalysis指出,当RL成为模型能力持续Scaling的关键,竞争的就不只是算法,还有支撑模型持续生成、训练和更新的AI基础设施系统。 问题也随之变成:当模型越依赖RL获得能力,谁来支撑这种能力持续低成本地生产。 从“模型+算力”到算法与AI基础设施共同Scaling 智谱近期的模型迭代,为观察后训练RL提供了一个窗口。 智谱公告指出,GLM-5.3与GLM-5.2在相同基座上推进强化学习,通过后训练Scaling持续提升模型智能上界。以GLM-5.2为例,其在SWE-bench Pro取得62.1分、Terminal-B…