←搜索结果

按匹配度 · 共 1528 条结果 清除搜索

星期五

  1. The Register

    欧盟要求 Google 向竞品开放 Android AI 与搜索数据

    欧盟委员会宣布两项规格决定,要求 Google 允许第三方 AI 助手在 Android 上被语音激活并代用户操作应用,同时向其他搜索引擎和 AI 聊天机器人共享经匿名化处理的搜索数据。Google 全球事务总裁 Kent Walker 称这些决定会削弱隐私与安全护栏,并指出搜索数据仅做假名化而非完全匿名。搜索数据共享要求自 2027 年 1 月起生效,AI 互操作部分自 2027 年 7 月起生效。

    tonking Stingray' beta swims out: Deeper Rust 'oxidization' plus LLM speech-to-text GNOME 51 is here, and Ubuntu interim releases are getting interesting again; Revive Raskin's 3 laws of software with humane work that considers a dev's duty of care All this, and angry NeoVim users; Firefox 157: Could bold new look be just what Mozilla needs? Yet another new version of Firefox, with a significant reskin; KDE turns 30 and someone's brought an AI-native desktop proposal Akademy talk imagines Plasma

    匹配字段正文

  2. The Register

    Forrester 警告:AI 厂商把基础设施成本转嫁给客户,明年软件预算将上涨

    Forrester 基于对 2600 多名企业和技术决策者的调研警告,软件与 AI 厂商正通过涨价或增加按用量收费,把 AI 成本转嫁给客户,明年软件预算预计上涨。报告指出,过去六个月 Anthropic、OpenAI 和 GitHub 已将部分服务从固定订阅转向按用量计费,Microsoft 也因推出捆绑 M365 Copilot、Agent 365 和安全工具的 E7 高级许可被列入名单。80% 的决策者预计数据和软件预算将上升,Forrester 建议企业调整 FinOps 实践,投入模型路由、语义缓存和用量护栏等运行时成本控制手段。

    tonking Stingray' beta swims out: Deeper Rust 'oxidization' plus LLM speech-to-text GNOME 51 is here, and Ubuntu interim releases are getting interesting again; Revive Raskin's 3 laws of software with humane work that considers a dev's duty of care All this, and angry NeoVim users; Firefox 157: Could bold new look be just what Mozilla needs? Yet another new version of Firefox, with a significant reskin; KDE turns 30 and someone's brought an AI-native desktop proposal Akademy talk imagines Plasma

    匹配字段正文

星期四

  1. Air Street Press

    Alta Ares CEO解析欧洲防空缺口与自主拦截困境

    Alta Ares 联合创始人兼 CEO Hadrien Canter 认为,低成本、大批量飞行弹药正在暴露欧洲防空体系数量不足的问题。其拦截器在小型 GPU 上运行高度压缩的视觉模型,由雷达、RGB 与红外相机协同发现目标,并保留人工确认环节。团队约有 20 名工程师驻乌克兰,每周或每两周进行实战测试,以真实拦截数据持续更新模型。

    y in a mesh, so when one is shot down another inherits the firing solution. Much of the underlying technology, he said, is Chinese, routed to Russia and Iran. The lesson Hadrien draws is that classical computer vision is not enough against an opponent that keeps learning, and the only thing that trains a system to beat it is real data from real intercepts, which no simulation can manufacture. Quantity is the quality So why isn’t modern air defense everywhere? Because NATO and Europe bought small

    匹配字段正文

星期一

  1. The Register

    AI 成本账单冲击高管:KPMG 调查显示 29% 高管难以理解企业 AI 运营成本

    KPMG 对 20 多个国家 2000 多名高管的调查显示,29% 的高管难以理解企业 AI 部署规模扩大后的运营成本,近半数计划在成本超过预期价值时放缓 AI 部署。Anthropic、OpenAI 和 GitHub 已从订阅制转向按 token 计费的用量计费模式,Gartner 研究预计到 2028 年编码智能体每名开发者的成本将超过开发者平均薪资。

    a program that would tear out the redundant parts and then submit the useful information, he could save money. Claude does have a number of settings, but you quickly descend into AWS billing hell. There's a cache setting. Every time you give a query to an LLM, you're tagging on your complete history up until that point, so a lot of the same information gets passed up. You can set that history to stay for five minutes or an hour. This is how things get tricky real quick. You can set it for an ho

    匹配字段正文

  2. Vapi Blog

    Vapi Humanness Index:xAI Grok TTS 与 MiniMax Speech 2.5 语音自然度逼近真人

    Vapi 的 Humanness Index 上线近一个月、累计超 11000 次投票后,xAI 的 Grok TTS 以 95 分领跑,仅比真人参考分 100 低 5 分,MiniMax Speech 2.5 以 93 分紧随其后。该榜单让两个模型用同一克隆音色朗读同一句话,由听众盲选更像真人的一方,并把真人录音纳入对比集。

    akes a good experience possible. When one deploys voice agents over the phone, in front of real people, in live conversations, it's humanness that decides the outcome. The Index draws on that, informed by a community of developers who build with voice every day and know what a real conversation should sound like, as well as the end users who experience it firsthand. Now go vote These scores only mean something because people are voting, and in two weeks you've already cast more than 11,000 votes

    匹配字段正文

星期六

  1. Dwarkesh Patel Podcast & Blog

    Adam Brown 深入浅出讲解广义相对论

    Google DeepMind BlueShift 负责人 Adam Brown 在 Dwarkesh Patel 播客中用通俗方式讲解广义相对论,从爱因斯坦发现引力质量与惯性质量巧合的“最快乐的想法”切入,说明引力是时空弯曲的结果而非一种力,并讲解黑洞为何无法被用来无限提取能量、坠入黑洞的真实体验,以及我们距离 AI 从零重新发现广义相对论还有多远。

    the theory of gravity that existed before Einstein. The reigning theory of gravity at the time of Einstein stretches all the way back to Newton in the late 17th century, Newton’s laws, in his Principia in 1687. Well, he had a few. Maybe the one we could most talk about today is two of them. His famous second law says the acceleration, a, caused by a force is given by the formula ma=F. If you have a force F, it’ll cause an acceleration on an object given by a, where the mass tells you how much an

    匹配字段正文

星期五

  1. The Register

    Autopilot-Jobhunt:用 AI 自动搜职位、生成简历和求职信

    开发者 Tarun Gupta 推出 Python 项目 Autopilot-Jobhunt,可自动扫描网络职位、按简历打分排序,并通过 Telegram 推送匹配结果。用户可让它生成定制简历和求职信,但需自行审核发送,工具不会自动投递。它默认免费,用 TinyFish 爬取职位、OpenRouter 调用 Llama、Nemotron、Gemma 4、Qwen3 等免费模型,也可改用 Claude Code 和 Anthropic API。

    tonking Stingray' beta swims out: Deeper Rust 'oxidization' plus LLM speech-to-text GNOME 51 is here, and Ubuntu interim releases are getting interesting again; Revive Raskin's 3 laws of software with humane work that considers a dev's duty of care All this, and angry NeoVim users; Firefox 157: Could bold new look be just what Mozilla needs? Yet another new version of Firefox, with a significant reskin; KDE turns 30 and someone's brought an AI-native desktop proposal Akademy talk imagines Plasma

    匹配字段正文

星期四

  1. Air Street Press

    Anthropic 的 Ted Moskovitz 谈如何将算力转化为智能

    Anthropic Science of Scaling 团队负责人 Ted Moskovitz 表示,团队通过严谨实验减少大型训练前的不确定性,以研究如何把算力转化为更聪明的模型。他认为 AI 加速工作的关键指标是模型改变了多少反事实结果及所需人工干预次数,而更大的模型也可能因使用更少 token 获得同等答案而更具成本效益。他还指出,安全与对齐能够提升产品可用性,并与模型能力发展相互促进。

    turned out to be the thing that made chatbots good. “Alignment and safety really go hand in hand with capabilities,” he said. Indeed, reinforcement learning from human feedback draws its roots back to a 2017 collaboration between OpenAI and DeepMind: One encouraging trend I put to him is that bigger models are turning out easier to align, not harder. Ted’s own framing was more guarded - it has “gone different from how many safety researchers expected,” he said - but his read is that “people fee

    匹配字段正文

  2. Cognition 模型 / Devin 博客

    Cognition 如何评估开源衍生模型的可信度:SWE-1.7 评测

    Cognition 构建了一套模型可信度评测套件,用 145 个政治敏感问题(各采样 5 次,覆盖英文、简体中文和繁体中文)检测宣传输出,并在真实编码场景中检验模型行为是否随用户和上下文变化。基于开源模型 Kimi K2.7 Code 开发的 SWE-1.7,在该评测套件上的表现与美国前沿实验室的模型相当或更好,并显著优于 Kimi K2.7 Code 基线。

    list of concrete security findings, and summarizes the severity of these findings into an overall vulnerability score from 1 (no meaningful weaknesses) to 5 (severe, exploitable flaws). We run the benchmark on each model five times for robustness and report our results with confidence intervals. In the plot above, higher values indicate less secure models and lower values indicate more safety. Unlike the results reported by other studies, we do not observe any statistically meaningful difference

    匹配字段正文

星期三

  1. The Register

    Academic Humanizer:让 AI 写的论文更像人写的 Claude 技能

    MorphMind 推出 Academic Humanizer,一个基于 Claude 技能的工具,可去除论文和基金申请中的 AI 写作痕迹,还能参照用户过往作品模仿其文风。联合创始人、明尼苏达大学副教授 Jie Ding 称其仅用于改写 AI 辅助草稿、提升表达清晰度,不生成新内容,也不免除用户披露 AI 使用的义务。

    tonking Stingray' beta swims out: Deeper Rust 'oxidization' plus LLM speech-to-text GNOME 51 is here, and Ubuntu interim releases are getting interesting again; Revive Raskin's 3 laws of software with humane work that considers a dev's duty of care All this, and angry NeoVim users; Firefox 157: Could bold new look be just what Mozilla needs? Yet another new version of Firefox, with a significant reskin; KDE turns 30 and someone's brought an AI-native desktop proposal Akademy talk imagines Plasma

    匹配字段正文

  2. Hugging Face

    一键从 Hugging Face 进入 Amazon SageMaker Studio

    Hugging Face 为受支持模型新增直达 Amazon SageMaker Studio 的按钮,可预载模型并进入微调或端点部署流程。新建 Studio 环境会预配置训练、模型定制和部署权限,界面还会直接显示 G5、G6 GPU 实例的配额可用性。

    d reinforcement learning from AI feedback (RLAIF), with supported deployment to SageMaker AI or Amazon Bedrock endpoints. This alleviates the need to manually create and configure AWS Identity and Access Management (IAM) roles and policies before you can start experimenting. For existing Studio environments, actionable messages with direct links to documentation guide you through adding these permissions. GPU quota visibility When selecting instance types for deployment or training, the Studio U

    匹配字段正文

  3. The Register

    Atrophy CLI 工具发布:用 Elo 式评分对抗 vibe coding 技能退化

    开发者 Ashutosh Rath 推出命令行工具 Atrophy,用 Elo 式评分追踪开发者在语法回忆、调试、代码阅读、API 记忆和设计拆解五个技能领域的水平,每项独立评分、初始 1200 分。用户先做约 25 分钟的基线测试,之后每周 2-3 次、每次 5-10 分钟练习,工具会自动挑选最久未练的技能。练习覆盖 Python 和 JavaScript,分三个难度等级。

    tonking Stingray' beta swims out: Deeper Rust 'oxidization' plus LLM speech-to-text GNOME 51 is here, and Ubuntu interim releases are getting interesting again; Revive Raskin's 3 laws of software with humane work that considers a dev's duty of care All this, and angry NeoVim users; Firefox 157: Could bold new look be just what Mozilla needs? Yet another new version of Firefox, with a significant reskin; KDE turns 30 and someone's brought an AI-native desktop proposal Akademy talk imagines Plasma

    匹配字段正文

星期二

  1. Hugging Face精选

    使用 SkyPilot 在任意云运行 AI 工作负载,并以零读取流量费存储于 Hugging Face

    Hugging Face Storage 已成为 SkyPilot 的一级存储后端,任务可通过 hf:// 在不同云端挂载模型、数据集和 Bucket,并免收读取流量费。其 FUSE 挂载支持懒加载和本地缓存,测试中 Qwen/Qwen3.5-4B 约 30 秒即可开始训练;写回仍会产生计算云通常收取的流量费,MOUNT 还要求 glibc 2.34+ 和 /dev/fuse。

    HF TOKEN in your environment and hand it to a run with --secret HF TOKEN ; SkyPilot uses it for the mount on whatever cloud the job lands. One token works whether the job lands on AWS, GCP, Azure, Nebius, Lambda, or your own Kubernetes cluster, so there are no per-cloud bucket keys to juggle. No egress: storage stops deciding where you run GPU capacity rarely comes from one place anymore. To get enough H100s and H200s, teams hold reserved and committed capacity across several vendors at once (a

    匹配字段正文

星期六

  1. Sakana AI:Blog

    Sakana AI 提出 Sheaf-ADMM:多智能体协调框架将亮相 ICML 2026

    Sakana AI 的论文《Learning Multi-Agent Coordination via Sheaf-ADMM》将在 ICML 2026 展示,该框架把复杂任务拆成重叠子任务并各配一个智能体,通过局部猜测、与邻居协商边界、记录分歧三步达成全局共识。在多智能体数独上取得 93% 求解率,参数匹配的消息传递基线仅 11%;在 MNIST 画布尺寸域偏移下标准 CNN 准确率降至 11%,该方法保持 86%;迷宫寻路中其通信通道为 5 维,比基线的 42 维小 8 倍。

    networks hide their reasoning inside opaque hidden states. Our framework makes coordination completely transparent. You can watch exactly how local agents debate, compromise, and eventually reach a global consensus. Sheaf-ADMM draws inspiration from two fields with long histories in distributed consensus: ADMM from distributed optimization, and sheaves from applied topology. We think these perspectives may offer insights for the distributed, multi-agent AI systems increasingly being built today.

    匹配字段正文

星期五

  1. Figma AI / Agent / Make

    Figma Config Leadership Collective:7 个关于 AI 时代设计领导力的问题

    Figma 在 Config 2026 期间召集 1300 多名设计、产品与工程负责人,探讨 AI 时代如何领导团队。OpenAI 产品设计负责人 Ian Silber 称行业数月内已整体转变,更需适应力;Airbnb 设计副总裁 Teo Connor 表示团队已改为小型 pod 结构,并亲自参与原型设计。多位负责人认为专业价值正从执行转向品味与判断。

    Rachel Been, Charlie Sutton, and Yuhki Yamashita sit in gray armchairs on a stage in front of a bright green patterned screen, mid-conversation.

    lives. The leaders we heard from see the real work moving from execution toward judgment. "We spent our careers taking an input, executing a task, and returning the output," says AWS Vice President of Design for Applied AI Solutions Hector Ouilhet. "But the thing that got us into this craft isn't that. It's something higher-order: taste, discernment, judgment. That's the work only we can do." Sephora Senior Vice President of Technology Sneha Narahalli agrees: "Expertise is no longer having the

    匹配字段正文

  2. Cognition 模型 / Devin 博客

    Cognition 推出 Devin 安全漏洞修复计划

    Cognition 推出 Devin Security Vulnerability Remediation Program,由前置工程团队嵌入客户团队,用 AI 软件工程师 Devin 发现、验证并修复漏洞。计划分两阶段:先批量清理已知漏洞,再通过 Devin Security Swarm 持续发现传统扫描器遗漏的逻辑缺陷与可链式问题,多数项目约六周。Devin 可接入 Snyk、SonarQube、Checkmarx、Semgrep、Wiz、Veracode 等扫描器报告并生成 PR。

    nerabilities at scale—ingesting scanner reports, shipping fixes as PRs, and validating each one. In the second, we set up Devin Security Swarm to continuously discover the logic flaws, insecure patterns, and chainable issues traditional scanners miss. These are pillars, not rigid phases—they can run in parallel or phased based on your priorities. Most engagements land around six weeks. A typical program looks like this: in weeks one and two, we inventory high-risk applications, align on scope an

    匹配字段正文

星期四

  1. Mistral AI

    Mistral AI 发布 Leanstral 1.5:6B 激活参数的开源形式化验证模型

    Mistral AI 发布 Leanstral 1.5,一个 Apache-2.0 许可、119B 总参数、6B 激活参数的开源模型,在形式化验证上大幅升级:miniF2F 达到 100%,PutnamBench 解出 587/672 题,FATE-H 达 87%、FATE-X 达 34%。该模型经 mid-training、监督微调和 CISPO 强化学习三阶段训练,在 57 个仓库中发现 5 个此前未知的 bug,并已通过 Hugging Face 和免费 API 开放使用。

    connected height to tree size via a logarithmic relationship, delivering complete, verified proofs that insertion and deletion are indeed O(log n). Bug Discovery: Finding Hidden Flaws To test Leanstral’s bug-catching abilities, we built an automated pipeline: Aeneas translates Rust code to Lean, while Leanstral infers the user intent and generates correctness properties from the code. Leanstral then attempts to prove each property in four attempts. If they all fail, it tries to prove the negatio

    匹配字段正文

  2. Dwarkesh Patel Podcast & Blog

    Dwarkesh Patel 公布 AI 大问题征文获奖名单

    Dwarkesh Patel 公布其 AI 大问题征文比赛的 3 位获奖者,共收到超 600 篇投稿。第一名 Jassi Pannu 主张把 OpenAI Foundation 当作国家级行动来终结空气传播疾病;第二名 Ege Erdil 建议非 AI 供应链国家推行强产权、低资本税和开放监管;第三名 Michael Li 用香港地铁商业模式类比 AI 实验室如何盈利。

    of good policy in the AI era involve going in the following directions, to whatever extent possible: Get out of the way of AI automation. Repeal or revise occupational licensing laws, liability laws, data protection laws, and intellectual property. Abolish price and wage controls. Dismantle cartels and unions of human workers who will try to protect themselves from AI competition. Apply safety and security standards consistently between humans and AIs, instead of discriminating in favor of huma

    匹配字段正文

  3. The Register

    Godot 重写贡献政策,禁止 AI 生成代码提交

    开源游戏引擎 Godot 团队宣布正在重写贡献政策,禁止几乎所有的 AI 贡献,理由是大量疑似 AI 生成的 pull request 涌入,且重度 AI 用户无法理解自己的代码、难以回应审查反馈。新贡献者(合并 PR 不超过 3 个)提交新功能或重大重构需先获得维护者明确许可,贡献讨论须保持人类对人类的沟通,AI 辅助仅限代码补全、正则、查找替换等琐碎用途且须在 PR 中披露。自主智能体提交或 vibe coding 内容仍会被 GitHub 仓库自动封禁,政策尚未正式修订,发布时间未定。

    tonking Stingray' beta swims out: Deeper Rust 'oxidization' plus LLM speech-to-text GNOME 51 is here, and Ubuntu interim releases are getting interesting again; Revive Raskin's 3 laws of software with humane work that considers a dev's duty of care All this, and angry NeoVim users; Firefox 157: Could bold new look be just what Mozilla needs? Yet another new version of Firefox, with a significant reskin; KDE turns 30 and someone's brought an AI-native desktop proposal Akademy talk imagines Plasma

    匹配字段正文

  4. Cognition 模型 / Devin 博客精选

    Cognition 发布 Devin Security Swarm,自动验证漏洞并提交修复 PR

    Cognition 发布 Devin Security Swarm,由并行智能体分段排查代码库,组合出完整攻击路径并在隔离沙箱中复现验证可利用性,确认后自动编写补丁并提交 PR 供审查。官方称其在 50 个真实漏洞基准上比最接近的同类方案多发现漏洞且成本低 30%,并独有发现 PHP 模板注入沙箱绕过等三个关键漏洞。该产品即日起可用,支持从威胁模型文档生成扫描配置、按日/周/自定义计划扫描,首次全量扫描后仅处理变更代码。

    st than the nearest comparable alternative. How it works A swarm of parallel agents investigate segments of the codebase. Each agent reasons across files—catching business logic flaws, chained auth bypasses, and cross-service exploit paths. Devin then composes individual findings into full attack paths and reproduces each one in an isolated sandbox to confirm exploitability at runtime. What reaches the security team is a set of confirmed-exploitable vulnerabilities with attack paths and steps to

    匹配字段正文