SanjaySingh-v1.0

🐏 Himalayan Shepherd Son
🤖 AI so easy, Mummy learns it😄
🏕️ Building India's 1st Conservation Glamping
1️⃣ Person → 💰1 Billion $ | The journey is LIVE 🔴

[Joseph Tsar @joseph_tsar_]

Full Course on How To Speak Clearly

[Xudong Han @Xudong07452910]

今天已经看到不少人在分享 Stanford 新开的 CS329Z: Engineering AI Agents。

不想再重复介绍课程，我反而更好奇：

Stanford 到底把什么算作 2026 年的「Agent Engineering」？

看完整 syllabus，RAG、Tool Use、MCP、Agent Framework 基本都放在前半程，更像 Agent 的基础能力。

真正让我注意的是后半段。

第一节课就把 Agent Engineering 的核心工程问题概括成：

Decomposition、Data、Evaluation。

后面的课程也沿着这条线展开，从 Agent Data、Trajectory、Memory、Optimization，一直讲到 Evaluation、Coding Agent、Long-running Agent 和 Reliability。

作业设计也很值得关注。

第一个作业让学生从零搭 Agent；第二个直接变成 Evaluate an Agent：自己设计 benchmark、grader、LLM-as-Judge 和 error analysis。

Evaluation 甚至连续讲了两周。

一个很有意思的细节，是课程专门讲 pass@k 和 pass^k：前者看「多试几次能不能成功一次」，后者看「连续跑很多次能不能都成功」。

对 Agent 来说，这其实对应两种完全不同的可靠性。

Multi-Agent 那一节也挺克制。

一边讲 orchestration、handoff 和 collaboration，一边专门讨论协调失败和错误传播，reading 里甚至放了一篇：

《Don't Sleep on Single-agent Systems》

至少从这份 syllabus 看，「更多 Agent」并没有被默认等同于更好的系统。

所以我更愿意把这门课看成 Stanford 对 2026 年 Agent 技术栈的一次切片。

RAG、工具调用这些能力正在变成底座，后面更值得继续往下挖的，是 数据怎么来、系统怎么评、Agent 怎么优化，以及它能不能在复杂、长时间的真实环境里稳定跑下去。

这几个问题，感觉也会是接下来一两年 Agent 研究持续关注的方向。

https://t.co/s3i3GISGWb

[Tehxi @yajnshri]

Breath is the King of Mind by Renowned Yoga Guru Sh. B.K.S. Iyengar

[Amisha Yadav @amisha_explains]

🇮🇳 Officially from INDIAN IITs 🇮🇳

No fees. No entrance exam. No catch.

Here are 28 courses that are better than most paid bootcamps:

1. Artificial Intelligence: Concepts and Techniques: https://t.co/xQLCBAt71z

2. Artificial Intelligence: Foundations and Algorithms: https://t.co/oSYHBy2CAo

3. Cloud Computing: https://t.co/s3eKQ4kO5I

4. Computer Vision: https://t.co/sw0ODnKZkb

5. Corporate Finance: https://t.co/AMT78OsBlw

6. Cyber Security and Privacy: https://t.co/iqFHf8TPii

7. Data Science for Engineers: https://t.co/bc2P6F5Jsd

8. Data Structures and Algorithms Design: https://t.co/kuixtbaTCD

9. Data Structures and Algorithms using Java: https://t.co/X8UsoouHO6

10. Data Structures and Algorithms Using Python: https://t.co/ShnmBRxGR5

11. Database Management System: https://t.co/dFPgDFun9i

12. Deep Learning - IIT Ropar: https://t.co/DTlEKFvlJa

13. Digital Marketing in the AI era: https://t.co/cY4BAnSMKI

14. Ethical Hacking: https://t.co/Vt6zK7tKzl

15. Generative AI for Computer Vision: https://t.co/GoFZOqhaOm

16. Introduction to Large Language Models (LLMs): https://t.co/cb50yjV59U

17. Introduction to Internet of Things: https://t.co/zikEcWbfQ2

18. Introduction to Machine Learning: https://t.co/yOyY1M8YnB

19. Introduction to Operating Systems: https://t.co/yq2e4gr2S9

20. Learning Analytics Tools: https://t.co/8ZTz0nJwlf

21. Mathematical Foundations of Generative AI: https://t.co/yQJ5crEMkP

22. Natural Language Processing: https://t.co/NW57eofXXt

23. Operating System Fundamentals: https://t.co/qWX8UeZIYK

24. Programming in Java: https://t.co/OAyT6kLYAz

25. Programming in Modern C++: https://t.co/0E3lquqUpM

26. Programming with Generative AI: https://t.co/MItYoWr2bF

27. Python for Data Science: https://t.co/mCcMpCOsA3

28. Reinforcement Learning: https://t.co/9m0pEVhORd

Repost so someone else can find this roadmap, and pls consider following
@amisha_explains for more content around AI, Beauty, and businesses.

[Thomas Ricouard @Dimillian]

Astra is very good at 3D modeling, and I can't wait for all of you to experience it, for now here is a little walkthrough on how I built the demo house for our launch blog post. From a Blender scene to a Unreal Engine 5 walkable experience. 🧵

[Ronin @DeRonin_]

omg, this dude is killing it at robotics

he has explained step-by-step how to build your own Microduck (cool pet project for portfolio)

gonna try to build it next weekend

follow this hidden talent

[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

[DAN KOE @thedankoe]

https://t.co/klNSdaI7aa

[Avid @Av1dlive]

i've spent 70+ hours using Grok Bot.

call me insane... but i think this is the closest thing to AGI i've ever seen.

give this article to your bot. it'll be the most productive thing you do this week https://t.co/p6pycvbih6

[Remi Fabre @RemiFabreRobot]

We sold over $2,500,000 worth of Microducks in the first 24 hours. Is this one of the most successful robot launches in history?

Insane achievement by @antoinepirrone and the team!

[clem 🤗 @ClementDelangue]

BIG ANNOUNCEMENT FROM HUGGING FACE TODAY:

We're unveiling Microduck 🐥🤖

It's a tiny $399 open-source robot you can teach new tricks with reinforcement learning. It can walk, pick things up, get back up when it falls, and even roller-skate.

Welcome to the era of open-source affordable robots to democratize physical AI and world models!

🤗🤗🤗

[Dan Kornas @DanKornas]

NVIDIA AI Infrastructure & Operations (NCA-AIIO) - Complete Certification Exam Course

What you will learn:
- Build a practical understanding of this part of deep learning
- Understand how models learn, optimize, and improve from data
- Prepare data and features so the modeling work has a solid base
- Connect retrieval, embeddings, and generation into practical AI systems
- Work with sequence, text, and language-modeling problems
- Move from notebooks toward deployable, monitored AI systems

Link is in the reply 👇

♻️ Share this with your network if you found it useful or insightful.

[Naval @naval]

One of the best recent podcasts on AI. Neil has a gift for explaining all of the jargon and insights simply.

(I’m not involved, just found it unusually educational).

[Tech with Mak @techNmak]

Someone built a 503-lesson AI engineering curriculum and open-sourced the whole thing.

It’s called AI Engineering from Scratch.

20 phases. Roughly 320 hours. Free. MIT licensed. Python, TypeScript, Rust, and Julia.

What makes the curriculum interesting is how it teaches =>
You don’t start with the framework.
You build the smaller version yourself first.

> Backprop from the math.
> A tokenizer from scratch.
> Attention from scratch.
> An agent loop from scratch.

Then you use the production library for the same idea.

The lesson structure is consistent:
MOTTO → PROBLEM → CONCEPT → BUILD IT → USE IT → SHIP IT

And every lesson leaves you with something reusable:
→ prompts for AI assistants
→ SKILL[.]md files for coding agents
→ agent implementations
→ MCP servers
→ runnable code

The repo describes the end result as a portfolio of 503 artifacts you built and understand.

It also has an AI-native way to actually take the course.

Run:
npx skills add rohitg00/ai-engineering-from-scratch

Then /start-learning.

A 10-question placement quiz finds your starting point and creates a personalized LEARNING[.]md.

From there:
→ /learn teaches the next lesson through concept, math, code, and quiz
→ /course-guide finds the exact lesson for a topic
→ /check-understanding quizzes you on a phase

It works with Claude Code, Cursor, Codex, OpenClaw, Hermes, and other agents that understand SKILL[.]md.

And there’s another useful part

The repo also includes a free, open-source Claude Certification Academy covering all four official Claude certification tracks =>
→ Associate Foundations
→ Developer Foundations
→ Architect Foundations
→ Architect Professional

Each route includes blueprint-mapped lessons, runnable labs, a diagnostic, capstone work, and a full-length original practice exam.

There’s even a /claude-certification tutor that can help choose a track, teach the lessons, run labs, review artifacts, and save progress. (not affiliated with Anthropic)

The full curriculum:
→ Phase 0 - Setup & Tooling
→ Phase 1 - Math Foundations
→ Phase 2 - ML Fundamentals
→ Phase 3 - Deep Learning Core
→ Phase 4 - Computer Vision
→ Phase 5 - NLP
→ Phase 6 - Speech & Audio
→ Phase 7 - Transformers Deep Dive
→ Phase 8 - Generative AI
→ Phase 9 - Reinforcement Learning
→ Phase 10 - LLMs from Scratch
→ Phase 11 - LLM Engineering
→ Phase 12 - Multimodal AI
→ Phase 13 - Tools & Protocols
→ Phase 14 - Agent Engineering
→ Phase 15 - Autonomous Systems
→ Phase 16 - Multi-Agent & Swarms
→ Phase 17 - Infrastructure & Production
→ Phase 18 - Ethics, Safety & Alignment
→ Phase 19 - Capstone Projects

Phase 19 alone contains 17 end-to-end products plus 9 deep-build tracks.

There’s even a six-volume EPUB/PDF edition generated from the same core curriculum.

503 lessons is a lot, but the structure is what makes this repo useful => understand the mechanics first, build them yourself, then move up to the abstractions.

GitHub Repo: https://t.co/E2Rg09gnrR

[ISRO @isro]

Milestone Wins – Flight to the Moon
Chandrayaan 3’s Vikram Lander placed India near the Moon's south polar region and became milestones of ingenuity, precision and resolve.

Join us via livestream for NSpD2026 Celebration:
https://t.co/nJUzQe7ERi

#NationalSpaceDay2026 #Chandrayaan3 #VikramLander #SoftLanding

[Daniel Belfort ⊙ @danielbelfort]

Introducing Offprint™

The physical magazine made from everything you "save for later" but never actually go back to read.

Find online, read offline.

Print media is back!

[Roan @RohOnChain]

My friend applied to 150 tech jobs in two years. No MIT. No Stanford.

Last month Anthropic offered him $750,000.

I asked him how he broke in from zero.

He sent me the exact video that got him in. A 3-hour course to build a full LLM from scratch.

A developer teaches you exactly how LLMs like ChatGPT and Claude are actually built.

I watched it last night.

Halfway through, I realized I could break into an AI lab in weeks, not years.

Bookmark this and read the article below.

• 00:00 - intro to building LLMs
• 17:58 - LLM tokenization
• 01:06:03 - LLM embedding vectors
• 02:46:55 - LLM transformer architecture
• 03:25:05 - GPT architecture

[Avinash Singh @AvinashSingh_20]

Visa sponsorship + Remote jobs + Freshly funded startups!

European Tech Visa Sponsorship
https://t.co/0cG1NsgTzh

US Tech Visa Sponsorship
https://t.co/4GobLk06zi

Japan Tech Visa Sponsorship
https://t.co/QZvaDwuNAA

150 Remote Companies Hiring
https://t.co/Nyr3R9x3AG

100% Remote Hiring Companies
https://t.co/oxlA04UU5w

400+ San Francisco Bay Area companies
https://t.co/LzP3zns3K7

1500+ Recent Funding Startups
https://t.co/yNRpm5oueq

Indian Startups Freshers DS/ML/AI Hiring 2026
https://t.co/IEaPKmJIe3

Indian Robotics Startups Database
https://t.co/7MV7NQdszm

1300+ Hidden Companies Hiring List
https://t.co/clu9NJAzAN

Bookmark it!

[Vaibhav Sisinty @VaibhavSisinty]

An Indian engineer in Bengaluru got tired of potholes destroying his car.

So he built an app with Codex that detects them, finds who is responsible, and files a complaint.

Here is how it works. He installed a dashcam with GPS and an accelerometer. The app records everything while he drives.

A vision model classifies every pothole by size. Small, medium, large.

But here is the clever part. Most Indian roads are under warranty.

The contractor who built them is legally responsible for fixing potholes for free.

His app searches through 2,900 government contracts, finds the exact tender number, identifies the officer responsible, attaches the photo and geolocation, and generates a ready-to-file complaint.

One drive to work. 12 potholes detected. 12 complaints ready.

--
vc: @gkcs_

[Andrew Ng @AndrewYNg]

New: A map of the most important skills in AI Engineering. https://t.co/VVkn1Dqp1N