Google AI@GoogleAI · 5月1日69

Last week, we made Gemini Embedding 2, our first natively multimodal embedding model, available to the general public. Since then, developers have used it to build video analysis tools, visual shopping assistants, and more. But you might be wondering... what is an embedding model? 🤔 Let’s break it down! 1. What is it? Think of an embedding model as a "universal translator." It takes text, images, video, and audio data and turns them into a long string of numbers, like a unique digital fingerprint. 2. How does it work? Historically, search has been text only. Now, instead of just matching data by keyword, Gemini Embedding 2 maps multiple modalities in the same space based on meaning. It "feels" the connection between a video of a soccer goal and the words "game-winning shot" without needing tags. For example, "ocean" and "waves" are placed close together, but "ocean" and "toaster" are miles apart. 3. How can you use it? Developers have been using it to incorporate smarter search functionality into their builds. This means creating tools where you can snap a photo of a product and type "find this in yellow," or search through thousands of hours of video by describing what happens in a scene. 4. Ready to try it out for yourself? You can start using it today via the Gemini API or the Gemini Enterprise Agent Platform.

译谷歌上周正式向公众发布了其首个原生多模态嵌入模型Gemini Embedding 2。该模型如同“通用翻译器”，能将文本、图像、视频和音频数据转化为独特的数字向量。其核心突破在于不再依赖关键词匹配，而是基于语义将不同模态的数据映射到同一空间，从而理解内容间的深层联系。开发者已利用该模型构建视频分析工具、视觉购物助手等应用，实现通过拍照或描述场景进行智能搜索的功能。模型现可通过Gemini API或Gemini Enterprise Agent平台使用。

OpenRouter@OpenRouter · 4月30日52

alphaXiv x OpenRouter is live! When an AI paper mentions a model, alphaXiv now turns it into a preview: provider, description, use-case rankings, and a direct link to the OpenRouter model page. Go from research to model in one click with @askalphaxiv

译alphaXiv x OpenRouter 现已上线！ 当 AI 论文提及某个模型时，alphaXiv 现在会将其转换为预览：提供商、描述、用例排名，以及直达 OpenRouter 模型页面的链接。 通过 @askalphaxiv 一键从研究直达模型。

meng shao@shao__meng · 4月30日57

Glean 发布 Waldo：自研"智能搜索专用模型"，作为前置环节运行在前沿大模型之前，专门负责检索规划，把"找资料"和"做推理"这两件事拆开 @glean https://www.glean.com/blog/waldo-launch Waldo 基于 NVIDIA Nemotron 3 Nano（30B/3B MoE），采用 instruct 模式而非 reasoning 模式，以压低延迟。使用 Thinking Machines Tinker API 做 LoRA 微调 为什么做 Waldo · 观察：企业 AI 任务无论多复杂，几乎都从搜索开始（多轮迭代检索 → 综合作答） · 痛点：让前沿模型同时做"检索规划"和"深度推理"，是用最贵的算力干最机械的活——慢且贵 · 思路：高频、定义清晰的子任务（如搜索）应该交给专门训练的小模型；前沿模型只做综合与生成 架构关键决策 · Waldo 作为 第一步 运行，而非作为前沿模型的 sub-agent 被调用。 · 子智能体方案需要 3 次串行推理；Waldo 前置最优情况下只需 1 次前沿模型调用 · Waldo 调用工具集（Glean Search、员工搜索、Web Search），完成后不生成自然语言，而是把检索好的上下文"原状"交给前沿模型，让前沿模型像自己搜过一样直接作答 · 不替代 Glean 现有的语义搜索与企业知识图谱，而是在其之上做"规划层" 衍生能力：自适应推理路由 Waldo 跑完后，其自身执行轨迹（调用次数、命中文档数、是否稀疏、是否需越界工具）天然成为路由信号，用来决定下游前沿模型应启用多深的推理档位。模型"读自己的活"来评估任务复杂度。 实际效果 · 单次 LLM 调用：Waldo 比默认推理模型（GPT-5.4 medium）快 10×+（~250ms vs ~3s） · 端到端集成后：延迟降低 ~50%，Token 消耗降低 ~25%，质量无回退 · 约一半查询走"快路径"，根本不需要前沿模型的完整能力

译Glean推出自研的智能搜索专用模型Waldo，旨在将企业AI任务中的“检索规划”与“深度推理”分离。Waldo基于NVIDIA Nemotron 3 Nano构建，作为前置环节运行，专门负责调用Glean Search、员工搜索和Web搜索等工具进行检索规划，并将检索到的上下文直接交给前沿大模型进行综合作答。这种架构避免了让昂贵的前沿模型处理机械的搜索任务，使单次调用延迟降低至约250毫秒，比默认推理模型快10倍以上。集成后，系统端到端延迟降低约50%，Token消耗减少约25%，且约一半查询可走“快路径”无需调用完整前沿模型。该设计印证了专用小模型在处理重复性任务上的效率优势，是智能体AI架构的重要演进。

Chubby♨️@kimmonismus · 4月30日64

Google just quietly proved the AI monetization thesis. Quite interesting earnings! Cloud revenue up 63% to over $ 20b gen AI product revenue growing nearly 800% year-over-year. The backlog nearly doubled to $460 billion in a single quarter. The number of $100M–$1B deals doubled. Multiple billion-dollar-plus contracts signed. But the real story is Search. The prevailing narrative was that AI would cannibalize Google's core business, people get answers from chatbots, stop Googling. The opposite is happening. Search ad revenue grew 19%, queries hit an all-time high. Google turned the biggest existential threat to its business into a growth accelerator.

译谷歌最新财报有力反驳了AI将侵蚀其核心业务的论调。其云收入增长63%至超200亿美元，生成式AI产品收入年增近800%，大额合同储备翻倍。关键转折在于搜索业务：搜索广告收入增长19%，查询量创历史新高。这表明AI非但没有取代传统搜索，反而成为其业务的增长加速器，成功将生存威胁转化为发展动力。

Sundar Pichai@sundarpichai · 4月30日63

Q1 earnings are in: 2026 is off to a terrific start. Our AI investments and full stack approach are lighting up every part of the business: Search queries are at an all-time high with AI continuing to drive usage. Google Cloud revenue grew 63%, Gemini models have incredible momentum, and it was our strongest quarter ever for consumer AI subs, driven by @GeminiApp. Thanks to our partners + employees around the world. Much more to share on our earnings call in 20 minutes… and at Google I/O in 20 days!

译谷歌2026年第一季度业绩表现强劲，AI投资与全栈策略正全面推动业务增长。公司搜索查询量因AI驱动创下历史新高，Google Cloud收入同比增长63%。Gemini模型发展势头迅猛，以GeminiApp为代表的消费者AI订阅业务也创下季度最佳纪录。公司即将举行财报电话会议，并将在20天后的Google I/O大会上分享更多进展。

向阳乔木@vista8 · 4月29日53

姚老师和张凯经过大量数据研究分析写的论文，还有一手实战经验。 用科学的方法做GEO，像用数据洞察做增长一样。

译姚老师和张凯的GEO论文已在全球最大论文平台arxiv完成审核并发布，这是全球第二篇GEO专项研究。论文基于今年3月最新数据，涵盖大量Prompt、引用和AI抓取记录，采用科学方法进行GEO分析，类似数据驱动的增长洞察。研究成果以正式报告形式呈现，源数据已开源在GitHub。作者表示，如果对社区有帮助，将继续抓取更多数据进行专项研究并开放成果。

TestingCatalog News 🗞@testingcatalog · 4月28日57

Sonar 2 is now available on Perplexity web 👀 > Sonar models are Perplexity’s in‑house LLMs, optimized specifically for fast, web‑grounded search and answering Which base do you think was used for Sonar 2? DeepSeek V4, Kimi K2.6, or Qwen?

译Sonar 2 现已在 Perplexity 网页端上线 👀 > Sonar 模型是 Perplexity 自研的大型语言模型，专门针对快速、基于网络的搜索与回答进行了优化 你认为 Sonar 2 是基于哪个模型开发的？DeepSeek V4、Kimi K2.6 还是 Qwen？ [引用 @sethsaler]：Perplexity 推出的 Sonar 2。👀 @testingcatalog @btibor91

阿绎 AYi@AYi_AInotes · 4月26日49

什么是 GEO，怎么做 GEO，怎么让你的龙虾这些 Agent 帮你搭 GEO 系统？老哥这篇文章讲的都很清楚，感兴趣的可以看看

译一篇由@Lonely__MH分享的文章详细阐述了GEO的概念、构建方法与实践应用。核心内容是指导读者如何利用AI Agent（文中以“龙虾”等代称）来搭建GEO系统。文章提供了从理论到实操的完整路径，旨在帮助感兴趣的开发者或研究者系统地掌握这一技术。

Ethan Mollick@emollick · 4月19日

This paper shows people are asking a lot of medical questions of AI already, but we have little evidence of how good or bad this is. Most of the published research uses old models & compares to doctors. How do new models compare to the info people would have gotten without AI?

译这篇论文显示，人们已经在向 AI 询问大量医疗问题，但我们几乎没有证据表明这究竟是好是坏。 大多数已发表的研究使用的是旧模型，并与医生进行比较。新模型与人们在没有 AI 的情况下会获得的信息相比如何？

TestingCatalog News 🗞@testingcatalog · 4月18日

Perplexity is working on Workflows for Perplexity Computer. Workflows will have a specific UI to capture additional context from users in order to better execute the task. Every workflow here was a startup before 👀

译Perplexity 正在为 Perplexity Computer 开发 Workflows。 Workflows 将有特定的 UI 来获取用户的额外上下文，以便更好地执行任务。 这里的每个 Workflow 以前都是一家初创公司 👀

Chubby♨️@kimmonismus · 4月17日

What the heck? Perplexity is cooking lately!! Need to check it out asap

译什么鬼？Perplexity 最近势头很猛！！得尽快去看看

Jeff Dean@JeffDean · 4月16日

Delighted to see that Ricardo Baeza-Yates (@PolarBearby) is this year's winner of the @TheOfficialACM Luiz Barroso Award (an award named in honor of my longtime Google colleague). Ricardo is widely regarded as one of the world’s foremost researchers in information retrieval, and many of you may have read is excellent textbook Modern Information Retrieval (co-authored with Berthier Ribeiro-Neto). He has also played a pivotal role in strengthening the Latin American computing community. Ricardo's accomplishments and selection would make Luiz proud! https://awards.acm.org/about/2025-barroso

译Ricardo Baeza-Yates荣获ACM Luiz Barroso Award。作为信息检索领域顶尖学者，其合著的《Modern Information Retrieval》是该领域经典教材。他长期致力于推动拉丁美洲计算社区发展，贡献卓著。该奖项以Google资深工程师Luiz Barroso命名，旨在表彰在计算领域取得杰出成就的研究者。

TestingCatalog News 🗞@testingcatalog · 4月14日

Perplexity grew from $100M to $500M with only 34% growth in team size and fully pivoted to Perplexity Computer. Computer for businesses 👀

译Perplexity 营收从 1 亿美元增长到 5 亿美元，团队规模仅增长 34%，并全面转型为 Perplexity Computer。 面向企业的 Computer 👀 [引用 @AravSrinivas]：Perplexity 最初是我们自用的小型企业工具。当时我们只有 4 个人，没有收入，但手边有 AI。 向 Computer 的转型实际上是一个圆满回归。创始人正在用它来发展对经济和社区有重要意义的公司。 看到它现在以重大方式助力小型企业和初创公司，令人欣慰。 Perplexity 仍然是一家初创公司。我们刚刚将营收从 1 亿美元增长到 5 亿美元（5 倍），而团队规模仅增长 34%。2026 年将在同样的小团队基础上实现营收翻倍。而我们才刚刚热身。 这里的每个人都在小型企业工作，我们构建的一切都是为了那些建设者。

DogeDesigner@cb_doge · 4月14日

NEWS: Grokipedia traffic is surging. The latest data shows a sharp spike in visits. More and more websites and blogs are now citing @Grokipedia articles. Share Grokipedia links and cite Grokipedia on your websites and blogs.

译新闻：Grokipedia 流量激增。最新数据显示访问量急剧上升。 越来越多的网站和博客正在引用 @Grokipedia 文章。 分享 Grokipedia 链接，并在您的网站和博客上引用 Grokipedia。

TestingCatalog News 🗞@testingcatalog · 4月11日

Perplexity started providing early access to previously announced Personal Computer! > Personal Computer is an always on, local merge with Perplexity Computer that works for you 24/7. Business automation will be one of the most hot AI topics in 2026. It is coming 👀

译Perplexity Personal Computer 启动早期测试，这是一款本地运行、7×24小时在线的AI助手。有用户称已投入2万美元用于商业自动化，4周内实现业务流程全面自动化。

Deedy@deedydas · 4月11日

Oh, how the tables have turned. OpenAI / ChatGPT is now running sponsored search ads on Claude keywords.

译OpenAI 与 ChatGPT 开始在搜索引擎针对 Claude 关键词投放赞助广告。昔日被拿来与 Claude 比较的 ChatGPT，如今也要购买竞品关键词广告位，形势彻底逆转。

Andrej Karpathy@karpathy · 4月5日

Something I've been thinking about - I am bullish on people (empowered by AI) increasing the visibility, legibility and accountability of their governments. Historically, it is the governments that act to make society legible (e.g. "Seeing like a state" is the common reference), but with AI, society can dramatically improve its ability to do this in reverse. Government accountability has not been constrained by access (the various branches of government publish an enormous amount of data), it has been constrained by intelligence - the ability to process a lot of raw data, combine it with domain expertise and derive insights. As an example, the 4000-page omnibus bill is "transparent" in principle and in a legal sense, but certainly not in a practical sense for most people. There's a lot more like it: laws, spending bills, federal budgets, freedom of information act responses, lobbying disclosures... Only a few highly trained professionals (investigative journalists) could historically process this information. This bottleneck might dissolve - not only are the professionals further empowered, but a lot more people can participate. Some examples to be precise: Detailed accounting of spending and budgets, diff tracking of legislation, individual voting trends w.r.t. stated positions or speeches, lobbying and influence (e.g. graph of lobbyist -> firm -> client -> legislator -> committee -> vote -> regulation), procurement and contracting, regulatory capture warning lights, judicial and legal patterns, campaign finance... Local governments might be even more interesting because the governed population is smaller so there is less national coverage: city council meetings, decisions around zoning, policing, schools, utilities... Certainly, the same tools can easily cut the other way and it's worth being very mindful of that, but I lean optimistic overall that added participation, transparency and accountability will improve democratic, free societies. (the quoted tweet is half-ish related, but inspired me to post some recent thoughts)

译AI正赋能民众反向提升政府透明度与问责制。历史上政府使社会"可读"，而AI让民众具备解析政府海量数据的能力。政府问责的瓶颈并非数据公开，而是处理原始信息的智能——如冗长法案虽法律透明却难以实用理解。AI不仅赋能专业记者，更让普通民众能解析预算、立法差异、游说关系等复杂信息，地方政府场景同样适用。尽管技术存在双刃剑风险，但作者对民主社会因参与度和透明度提升而改善持乐观态度。

Satya Nadella@satyanadella · 3月30日

Introducing Critique, a new multi-model deep research system in M365 Copilot. You can use multiple models together to generate optimal responses and reports.

译介绍 Critique，M365 Copilot 中全新的多模型深度研究系统。 你可以同时使用多个模型来生成最佳回复和报告。

Sundar Pichai@sundarpichai · 3月12日

We’re bringing new capabilities powered by Gemini models to @googlemaps. With Ask Maps, get answers to complex questions about any place you want. For example it can help with complex requests like "Find me the best 3-hour family hikes in the Grand Tetons and a spot for a packed lunch”. Will try this next time I'm there:) Rolling out now in the US and India.

译Google Maps 上线 Gemini 驱动的 Ask Maps，支持复杂地点查询，例如同时查找大提顿国家公园3小时家庭徒步路线和野餐地点。该功能已在美国和印度推出，作者表示下次去那里要试试这个功能。

DeepSeek@deepseek_ai · 12月10日

🚀 DeepSeek-V2.5-1210: The Grand Finale 🎉 🌐 Internet Search is now live on the web! Visit https://chat.deepseek.com/ and toggle “Internet Search” for real-time answers. 🕒 🧵(1/3)

译🚀 DeepSeek-V2.5-1210: 盛大终章 🎉 🌐 联网搜索现已上线网页版！访问 https://chat.deepseek.com/ 并开启"联网搜索"获取实时回答。🕒 🧵(1/3)