Speech-to-Text AI Specialist Now Valued at $2B
The vendor is part of a wave of native-AI startups that are attracting strong funding.
AI BusinessOriginal ↗
Cursor today introduced Origin, a cloud service that software teams can use to store their code. The launch marks the company’s first major product update since its $60 billion sale to SpaceX Corp. in June. Cursor develops one of the industry’s most popular desktop-based vibe coding tools. The program uses artificial intelligence agents to automate […] The post Cursor launches Origin code hosting service to compete with GitHub appeared first on SiliconANGLE.
要闻提示 1.101亿元大交易！阿里巴巴正式出售灵犀互娱，买方为信宸资本 2.发回重审！大疆诉美国国防部案赢得程序性胜利 3.宇树科技：将于8月19日科创板上市 4.李书福辞任董事会主席，吉利汽车管理层大调整 5.瑞幸新联动被指抄袭，网友：羊了个羊里都能消除的程度 6.钉钉全球商业总裁杨猛将离任，下一站蚂蚁集团 7.腾讯回购67.3万股股份，耗资3.007亿港元 8.苹果最快本月底前对iPhone 17系列全球涨价：最高涨价近千元 今日头条 101亿元大交易！阿里巴巴正式出售灵犀互娱，买方为信宸资本 8月17日，灵犀互娱CEO周炳枢发布内部全员信，宣布阿里巴巴集团与信宸资本已正式达成交易协议。根据协议安排，阿里巴巴将出让所持有的灵犀互娱股份，信宸资本将成为灵犀互娱的新任股东。 有消息称，中信资本旗下信宸资本以超15亿美元（折合人民币超101 亿元）收购阿里巴巴灵犀互娱。之前的“候选者”包括三七互娱、巨人网络、中国儒意等多家上市游戏企业全程参与竞购却全部出局。（新浪科技） 国内资讯 发回重审！大疆诉美国国防部案赢得程序性胜利 8月17日消息，美国当地时间8月14日，大疆创新（DJI）诉美…
NVIDIA teams use ChatGPT Work to reduce manual tasks, connect fast-moving signals, and scale successful workflows globally.
Sentence Transformers v6.0 新增第四种模型类型 MultiVectorEncoder，可直接加载 PyLate、Stanford-NLP ColBERT 及 colpali-engine 检查点，用于 ColBERT 式晚期交互检索。 🔗 阅读原文 via AIHOT · https://aihot.virxact.com/items/cmsyqip4o14umroz0ch5iv0iv
StartupBench 是一个基于市场验证的 AI 初创公司产品构建的端到端智能体基准，从真实采用的产品工作流中提炼任务，而非研究者预设任务。在统一智能体框架下，最强模型也仅能完成约 30% 的任务，复杂指令遵循和领域专业知识是主要失败来源。该基准揭示了当前通用智能体在真实用户任务上的能力边界。 🔗 阅读原文 via AIHOT · https://aihot.virxact.com/items/cmszieman0euzrodptihyllom
作者基于Reddit上"让Claude真正开始思考"的帖子，引入逻辑学中的"钢人论证"（steelman）概念，并自创了一个"双向钢人Prompt"。该Prompt通过重述真实问题、强化正反双方观点、找出关键变量、逼AI给出明确判断四步，旨在避免模型谄媚，帮助用户找到问题最本质的答案。作者以自己选择公司司庆日的真实案例演示了使用效果。 🔗 阅读原文 via AIHOT · https://aihot.virxact.com/items/cmsxwd7a40coxroz06g25sfaf
The model maker added $18 billion in annualized revenue in two months.
宇树科技发布「超人」机器人，并定 8 月 19 日科创板上市 8 月 17 日，宇树科技发布了一款人形机器人新品「超人」，原地跳高 2m，极限速度 12.66m/s（0.85m 腿长），超越全世界全人类原地跳高和奔跑速度纪录。 宇树科技表示，全新整机刚用 3 个多月研发出来，未来几个月还有很大的完善空间。 早在今年 4 月，宇树 H1 机器人就跑出 10m/s 刷新世界纪录，逼近博尔特速度极限。如今「超人」再次打破纪录，并成功超越人类极限。 宇树科技 8 月 12 日宣布，旗下人形机器人累计生产下线约 18,000 台（仅包括仿生双足人形的数量，不含其他类人形机器人或轮式底盘人形）。 同日，宇树科技发布首次公开发行股票科创板上市公告书称，公司股票将于 2026 年 8 月 19 日在上海证券交易所科创板上市。 根据此前公告，宇树科技发行价格 150.80 元/股，中一签（500 股）需缴款 7.54 万元。（来源：格隆汇） OpenAI 与英伟达扩大计算潜在机会，到 2030 年计算规模或达 6000 亿美元 8 月 17 日，根据英伟达发布的新闻稿，OpenAI 将在 PORTS-P…
Artificial intelligence startup Groq Inc. today announced that it has raised $350 million in additional funding. The Series A round was led by returning backer Disruptive. Grok stated that Nvidia Corp. plans to join the round later down the line, but didn’t specify how much the chip giant will invest. The cash infusion comes less […] The post AI cloud operator Groq raises $350M more in funding appeared first on SiliconANGLE.
Cyber resilience is becoming a business imperative as AI accelerates attacks, expands the enterprise attack surface and gives autonomous systems greater access to sensitive data and critical operations. Preventing every disruption is no longer a realistic measure of security success, even for mature organizations. As attack windows contract, enterprises must also understand which operations are […] The post Three insights you may have missed from theCUBE’s coverage of Black Hat USA appeared firs…
Google 开源了基于 ADK 和 Gemini 的零信任客服与退货智能体示例，演示如何防御提示注入等攻击。该架构在 LLM 上下文之外通过三层硬性安全机制保障：硬件支持的加密签名确保数据库写入不可抵赖、gVisor 沙箱隔离动态代码执行、确定性语义网关校验业务逻辑。系统提示词只是软约束，无法作为安全边界。 🔗 阅读原文 via AIHOT · https://aihot.virxact.com/items/cmsxv0p3v0bntroz0227a6b3j
Cursor 今日起向所有付费计划用户开放 Origin 代码托管的早期测试版，提供仓库、拉取请求、代码浏览及 GitHub 同步功能。用户可创建以 cursor.com/codebase/ 为前缀的仓库，或将 GitHub 仓库同步至 Origin，双向同步评论与审查。Vercel、Depot 和 Buildkite 集成已可用，智能体功能即将推出。 🔗 阅读原文 via AIHOT · https://aihot.virxact.com/items/cmsxsvb5s0a4troz0z2q2nu3h
The vendor is part of a wave of native-AI startups that are attracting strong funding.
This story originally appeared in The Algorithm, our weekly newsletter on AI. To get stories like this in your inbox first, sign up here. Flock, the police-tech giant known for its network of some 120,000 automatic license plate readers around the US, announced some changes to its platform last Thursday. The updates are meant to prevent…
404 Media 通过在一本珍本图书中放置追踪设备，首次揭露亚马逊未公开的购书行动：批量购入大量书籍，扫描用于 AI 训练数据，随后销毁。追踪显示这些书最终被送往亚马逊的一处人工智能训练中心。 🔗 阅读原文 via AIHOT · https://aihot.virxact.com/items/cmsxld6es03o8roz0jblf3ii9
Teams customize their models to hit their targets for latency, speed, memory, and compute. With the open NVIDIA Nemotron family of models, developers can find... Teams customize their models to hit their targets for latency, speed, memory, and compute. With the open NVIDIA Nemotron family of models, developers can find the right-sized model for their needs. The new Nemotron 3.5 Lightning NVFP4 checkpoint, for example, preserves accuracy while unlocking up to 4x faster throughput. It’s compressed…
The model can run locally on a laptop and demonstrates how Alibaba plans to continue pioneering open source.
一份面向希望减少技术环境中侵入式 AI 的用户的操作指南，涵盖 Adobe Acrobat、Android/Gemini、Apple Intelligence、Chrome、Edge、Firefox、DuckDuckGo、Google Workspace、Slack、WhatsApp 及 Windows 11/Copilot 等平台。 🔗 阅读原文 via AIHOT · https://aihot.virxact.com/items/cmsxkakju0300roz0zx6cvdeu
The debate over whether we are in an artificial intelligence bubble took a new turn this week. Despite the ballooning AI spending, Dave Vellante (pictured, right), chief analyst for theCUBE Research, contends that any bursting point may be far off. Now that Nvidia Corp. Chief Executive Jensen Huang, has committed $500 billion to establish independent […] The post On theCUBE Pod: AI bubble debate heats up and neocloud earnings challenge doubters appeared first on SiliconANGLE.
Run an inference server yourself for low latency, on-prem data, or offline use; use a hosted API otherwise. Start one locally with Docker.
Anthropic 的 CI 工程师用 Claude Tag 构建了值班智能体，作为 CI/CD 故障的一线响应者。Claude 在事故发生后中位 14 分钟发布首份基于证据的分析，最快案例中 3 分钟内验证修复并确认错误率恢复基线。该方案通过 Slack 频道、Datadog 或 Grafana 工具访问及 GitHub 技能文件实现，Anthropic 已发布通用设置套件供其他团队部署。 🔗 阅读原文 via AIHOT · https://aihot.virxact.com/items/cmsz1zjiy013pro3f1e0aafra
Anthropic 公布两项实验：Claude（Mythos Preview 和 Opus 4.8）针对 15 个靶点设计蛋白质结合剂，成功 14 个，命中率达 22.6%-35.1%。 🔗 阅读原文 via AIHOT · https://aihot.virxact.com/items/cmsz8gh3n06e4rodpl8l2syek
Higgsfield Inc., a generative artificial intelligence video and image creation platform for professionals, today announced that it has raised $400 million in early-stage funding, quadrupling the company’s valuation to $5.4 billion since its previous funding round in January. The Series B round was led by DST Global, with participation from Tribe Capital, Growth Equity at […] The post Higgsfield raises $400M at $5.4 billion valuation to scale video and image generation platform appeared first on …
Humanoids desperately need to stop making YouTube videos and get a job already, and Persona AI is one of the few humanoid companies which seems to be entirely focused on making that happen. Persona’s approach has been to carefully select a job that is economically viable for robots right now, and they’ve found one that was also the job of one of the very first industrial robots ever sold: welding. IEEE Spectrum first spoke with Persona two years ago, shortly after it was founded by Nic Radford a…
Wispr AI Inc., the developer of a cross-platform, AI-powered dictation platform, today announced it has raised $280 million in a Series B funding at a $2 billion valuation. Menlo Ventures led the round alongside existing investors Notable Capital, NEA, Neo Ventures, 8VC and MVP Ventures. New investors Acrew, Forerunner, Goodwater, Peak XV, Together Fund, and […] The post Wispr raises $280M to power up natural speech-to-text using AI appeared first on SiliconANGLE.
The impact of AI on software development has been both profound and ever-evolving. Last year, I wrote about AMD’s plans to use AI not just for generating new lines of code, but also for other steps in the software development lifecycle (SDLC), such as triaging problems, debugging code, and testing the software. At the time, we were hoping for a 25 percent productivity boost from AI use over the course of two or three years. But with each new release, the capabilities of large language models (LL…
The robot is part of the companies' expanded partnership to shift physical AI from concepts to real-world deployments.
Artificial intelligence startup Palona AI, which is building an intelligence operating layer for physical businesses, is ready to show what it can do after closing on a $20 million round of early-stage funding today. The Series A round, which included converted simple agreements for future equity, was led by Ardenwood Ventures and saw participation from […] The post Palona raises $20M in funding to bring AI automation to brick-and-mortar businesses appeared first on SiliconANGLE.