Phillips

Interested in all beautiful things

[Phillips @PhillipsPan]

@MJTVHoPin 老年痴呆了？😱

[0xRafy @0xRafy]

https://t.co/PwUcHdewJJ

[Kin @KinGao476942]

https://t.co/dFoBkgY0Wx

[Kin @KinGao476942]

兄弟姐妹们，Grok bot 得学呀，这篇文章推荐每个人都认真看看，Grok bot 工程师写的，对个人开发来说，实现 10 倍提效不是梦，因为过去你可能只能同时管理 5～10 个窗口，现在借助 Grok bot @bot 和文章中分享的办法，你可以同时管理 200 个了窗口了！

[Andrew @s4yonnara]

https://t.co/YtaiuaNGjW

[Miles Ma @miles_mazy]

https://t.co/5Zh6GyBYxb

[Miles Ma @miles_mazy]

把 Hugging Face 入门整理成了一个 5 分钟视频：

从 Spaces 试应用，到 Models 找模型，再到 Datasets 看数据，常用入口都配了实际截图

我的建议是，先试出喜欢的效果，再找背后的模型，确认软件兼容后再下载

模型怎么选、文件怎么下载、LoRA 怎么搭配、好用的资源怎么收藏，这条视频带你走一遍

[何頻 @MJTVHoPin]

押注美國衰落是中國最危險的戰略；中國最大的弱點不是芯片，是福利；中國的優勢正在變成脆弱性｜大事火拍（20260826）
https://t.co/sKwC4gTRtX

[Codez @0xCodez]

SpaceXAI engineer, Lauren Tan:

"99% of people using GrokBot just for 1% of its real power. They run 1 agent without "loop" & "graph"

I'm running a team of 20+ GrokBot agents, fully autonomous. I have a Chief of Staff agent, a PM agent and 20+ workers - that's the new stack of engineer"

In a 1-hour session, a SpaceXAI engineer showed how to build a team of effective AI agents from scratch

this is worth more than a $500 agentic engineering course

watch this workshop today, then read how to build a fleet of GrokBot agents in the article below

[meng shao @shao__meng]

Computer Use 好用，但很贵！Codex 如此，Grok Bot 也是如此！

那怎么做，能既用上 Computer Use 的能力，又降低 token 消耗成本呢？

@SamSokolin 的思路是对的：
让 Bot 只用一次 Computer Use 完成任务，同时抓取过程中的网络请求，然后把这些请求整理成一段脚本。���后重复执行同一任务时，直接调 API，不再走图形界面。

# 为什么这样做？

1. 省 token。 Computer Use 模式的成本结构决定了它很贵：每一步都要截屏 → 模型看图理解 → 决定点哪里 → 再截屏验证。一个"打开发票页、筛选未付款、导出 CSV"的任务可能要几十轮视觉推理。而脚本执行不需要模型参与，几乎零 token。

2. 快得多。 GUI 操作受页面加载、渲染、模型推理延迟的多重限制。直接调 HTTP 接口只受网络往返时间限制，通常从分钟级降到秒级。

3. 复用现有登录态。 推文里"with the same auth"是关键。Bot 在浏览器里已经登录了，抓包时能拿到 session cookie / token，脚本沿用这些凭证，不需要额外申请 API key 或走 OAuth。这也是为什么这种方式对没有公开 API 的 web 应用同样适用——任何网页前端最终都是在调后端接口。

# 看看 Sam 展示的工作流

① 下达任务
"去 acme. com/invoices，筛选未付款，导出 CSV"

② Bot 执行并抓包
一边点击，一边记录了 4 个请求：GET /api/session、GET /api/invoices?status=unpaid、GET /api/invoices/columns、POST /api/export/csv

③ 生成脚本并加入 routine
输出一段 Python requests 脚本，复用 session、拉取未付款发票、POST 导出、写入本地文件；之后作为定时任务（routine）自动跑

整个流程：访问站点并交互 → 抓取网络请求 → 生成脚本 → 放进 routine 运行 → 返回结果。

# 使用 Agent 重要分工��则
视觉操作用于"探索"，代码用于"执行"！

[Grok Bot @bot]

Grok Bot for Enterprise is available today.

It’s free for all Grok and Cursor enterprise customers for the next two weeks.

https://t.co/rZPvHvfI5B

[Jackywine @Jackywine]

https://t.co/nlg7yJheQp

[Tony出海 @iamtonyzhu]

https://t.co/lik5rlHwmO

[Kai-Fu Lee @kaifulee]

I recently spoke with Bloomberg Weekend about three questions I think will define the next decade: Where is AI headed? How will organizations change? And what happens to jobs as AI becomes increasingly capable? These questions are also at the heart of my new book, AI Native, eBook/Audiobook launching Sep 15.
https://t.co/BmXAwVHsze

[Elon Musk @elonmusk]

I’m so glad this is being done!

An important next step is making it autonomous.

[黄小木 @ai_xiaomu]

obsidian=知识库，而知识库将会是 agi 的底座，chatgpt6 的翅膀⬇️

[Jensen Huang @JensenHuang]

Exciting day for NVIDIA and @huggingface.

Open models strengthen safety and cybersecurity, accelerate innovation and diffusion, and enable sovereignty. They allow every developer, startup, university, industry and country to build with, customize and benefit from AI.

Thank you @ClementDelangue for coming to me.

NVIDIA is going to be a great home for Hugging Face, its community and the future of open models. 🤗

https://t.co/q8Om2Xc5ye

[meng shao @shao__meng]

Grok Bot 产品最佳实践：用一支 Bot 团队完成研究、PRD、设计与交付

SpaceXAI 产品团队 @KevinKNeilson 和 @roshan_s 做了一场 45 分钟的工作坊，主讲 Grok Bot 产品最佳实践，展示了一种新的产品团队组织方式：PM 第一次拥有一支真正"向自己汇报"的团队——只不过成员是 Bot​ https://t.co/g2iOUvIuj2

[Fei-Fei Li @drfeifei]

Next view prediction is the key to Atlas, enabling us to unify pixel-level generation and reconstruction. @jcjohnss @BenMildenhall @martin_casado and I had a deeper discussion on some of the most exciting technical innovations of Atlas, our newly released world model for spatial intelligence!

[Grok Bot @bot]

You can now add Bot templates from our marketplace.

We’re sharing Haggle Bot, our in-house procurement specialist. It negotiates vendor contracts, finds unused SaaS seats, and price-checks recurring purchases.

One week in, it's saved us over $100K.