Keep on to blur preview images; turn off to show them clearly
She literally collected 900+ open-source AI tools and resources to teach you how to become a better AI engineer. Completely for free! From fundamentals, to RAG, evaluation, memory, reinforcement learning, and fine-tuning ✨ Just read the code, the docs, and the examples.
Product & Devs Growth @Cyfrin | Ex @Alchemy | Created @cyfrinupdraft and @AlchemyLearn | Robotics | Making web3 mainstream
Yes, 100x yes! The AI-generated code result should be reviewed in context of the original prompt. If the prompt is unknown, the intention is unclear, too, so we may now know *why* AI made certain decisions in the code, was it intentional or a hallucination.
~20 yrs in web-dev, now mostly Laravel. My Laravel courses: https://t.co/HRUAJdMRZL My Youtube channel: https://t.co/qPQAkaov2F
Seems like Twitter @x recently changed 'For you' as the default tab when visiting instead of 'Following'. How do I change this back? For you is completely useless...
Privacy Experience @PrivacyEthereum | Events @ethereum | OSS 🛠️ Indie Maker @eth_gastracker https://t.co/xNlTAwRRp1
Mino: AI Web Agent Platform Mino 是 @Tiny_Fish 推出的 AI 驱动网页自动化平台，通过自然语言描述任务，让 AI Agent 在真实浏览器中可靠地导航、执行复杂多步流程，并大规模访问无 API 的网站数据。 核心问题与解决方案 许多高价值网站数据隐藏在登录墙后、动态 JS 页面或复杂交互流程中，这些网站往往没有公开 API。传统方法包括： · 脆弱的网页抓取脚本：网站布局稍变即失效 · 消费级 AI Agent（如 Claude Computer Use 等）：每步动作都依赖大模型推理，依赖截图分析，导致 慢、贵、不确定 Mino 的核心创新在于“学习后固化”： · 首次运行：使用 AI 模型推理导航路径、理解页面并完成任务。 · 后续运行：将学到的流程转化为确定性脚本，以毫秒级速度执行，无需反复调用大模型。 这使得 Mino 在生产环境中更可靠、更快速、更低成本，同时随着使用次数增加，系统会不断优化和“记住”模式。 主要特点 · 自然语言控制：只需用英文描述任务目标（如“提取前 2 个商品名称和价格，并以 JSON 返回”），无需编写 CSS 选择器或 XPath。 · 真实浏览器自动化：支持登录、处理动态内容、填写表单、完成多步交互（如预约、结账）。 · 抗检测与隐身模式：内置 stealth 浏览器，绕过 Cloudflare、DataDome 等反机器人机制。 · 实时流式输出：通过 SSE 实时观察自动化执行过程。 · 代理支持：可指定地理位置路由请求。 · 大规模并行执行：适合高并发数据提取或跨数千网站的操作。 · 结构化输出：任务完成后直接返回 JSON 等格式数据。 · 与其他 AI 集成：支持 MCP，便于与 Claude 等大模型结合。 适用场景与优势 Mino 定位于 企业级 应用，特别是需要可靠、大规模网页操作的场景，如： · 实时价格监控、竞争情报采集。 · 旅行/酒店预订数据聚合。 · 健身/外卖平台实时可用性检查。 · 任何涉及“隐藏网页”的业务情报提取。 相比传统工具，它提供更高可靠性（99.99% 上线率）、更广覆盖范围和更低成本；相比消费级智能体，它更注重生产稳定性而非单次交互体验。 Mino 文档
邵猛，中年失业程序员 😂 专注 - Context Engineering, AI Agents. 分享 - AI papers, apps and OSS. ex Microsoft MVP 合作 - 私信/邮箱：shaomeng@outlook.com 📢 公众号/小红书: AI 启蒙小伙伴
RT @charles_irl: i think TCP is actually The Cooked Protocol - lost streaming audio/video to WebRTC over UDP - lost AI/HPC to IB verbs ove…
gpu enjoyer at @modal. he/him. ex @full_stack_dl, @weights_biases (acq. @CoreWeave), phd Berkeley @Redwood_Neuro. try https://t.co/SYWVMCb7OB
I have been going back and forth with a huge company that wants to use SaaS Pegasus for more than a month. They started with a "negotiation" vendor. Got 3-4 other parties involved. Lawyers. Multiple EULAs. Red lines. Have been using AI constantly to even understand what we are discussing. I really thought it'd eventually end with a big enterprise-deal payoff, but it's looking like they're gonna get a $999 license like everyone else. Lesson learned.
Went from CTO to solopreneur, now building an empire one business at a time. Growing https://t.co/o8SD3o1h6p, https://t.co/NVUx34raAP, and https://t.co/2uIpQjuNAC