Sonny PhD @Machine_Enginee

5 useful questions if you want to learn LLM application development this year: 1. What should I build first? 2. How do I know if my RAG system works? 3. When do I need agents? 4. What should I evaluate? 5. How do I move from a notebook to a working application? We cover these kinds of questions on LLM Zoomcamp. LLM Zoomcamp is a free, hands-on course where we build practical LLM applications with RAG, agents, vector search, evaluation, monitoring, and more. The live cohort starts on June 8, 2026. Join here: github.com/DataTalksClub/…

Anthropic CEO Dario Amodei: "The cheapest way to use Claude is also the smartest. Most devs do the exact opposite" this is one of the best interviews I've seen in a long time in this interview he breaks down exactly how a system changes everything: - the memory and context features that turn Claude into a second brain - the knowledge architecture most users don't know Claude can build - the integration layer that connects Claude to your actual workflow - why typing one question at a time is the most expensive way to use Claude if you've been using Claude for months and still start every conversation from scratch with zero context, you don't have a Claude problem. you have a system problem instead of another show tonight, watch this make sure to bookmark it before it gets lost in your feed full guide in the article below

Khairallah AL-Awady @eng_khairallah1

a week ago

x.com/i/article/2060…

33 53 350 620K 1K

Your technical books shouldn’t die as forgotten PDFs. book-to-skill is a Claude Code skill that turns a technical book or document into a structured skill for studying, referencing, and using while you work. It helps you query a book from inside Claude Code by extracting chapters, glossary terms, patterns, cheatsheets, and a main SKILL.md, then loading the right chapter files on demand. Key features: • Multi-format input – supports PDF, EPUB, DOCX, TXT, Markdown, reStructuredText, AsciiDoc, HTML, RTF, and MOBI/AZW formats • Skill output structure – generates SKILL.md, per-chapter Markdown files, glossary.md, patterns.md, and cheatsheet.md • On-demand chapters – loads chapter files only when you ask about a topic instead of putting the whole book in context • Technical-book extraction path – can use Docling for code blocks and tables, or faster text extractors for prose-heavy PDFs • Claude Code workflow – query the installed skill with commands like /your-book-slug replication or /your-book-slug ch05 It’s open-source (MIT license). Link in the reply 👇

500-hour AI infrastructure engineering curriculum github.com/ai-infra-curri…

@DoloresGMorris You might be a better and ethical researcher Dolores. If you look at the majority of STEM, and medical research, you will see this as a normalized trend of high self-citations mostly forced by academia, and sometimes greed to have high citation count!

"Matrix Calculus for Machine Learning and Beyond" is an interesting set of free lecture notes for understanding the mathematics behind modern deep learning. It covers gradients, Jacobians, Hessians, matrix-valued functions, backpropagation, optimisation, and many of the mathematical structures used in machine learning and AI models. One interesting aspect is that the material maintains a strong university-level rigour while remaining highly visual: the notes include numerous diagrams, graphs, geometric interpretations, and intuitive explanations of matrix calculus applied to neural networks. It is a valuable resource not only for students studying machine learning, but also for anyone who wants to build a solid foundation in computational linear algebra and optimisation. ocw.mit.edu/courses/18-s09…

New: Solana Summer School Spend your summer getting smart on the greatest blockchain on earth. Solana Summer School includes: - 5 weeks of live lessons and office hours with @SolanaFndn DevRel - Special guest lecturers from the ecosystem - Dedicated student resources - Plus something special for those who graduate Open to all current college students (must register with school email). Starts June 15. Space is limited. Apply now: luma.com/g8b5fy9s

Started to aggregate ZK resources I personally used for learning. There's a lot of "awesome ZK" repos, but some of them are lists of lists of lists. If you know anything worth adding, issue a PR or send me a carrier pigeon with a link github.com/0xluk3/ZK-reso…

💯 this is why I really like Learning mode in Claude Code I personally use this for all my side projects and it keeps me so much sharper, great if you want to use Claude Code but still stay hands-on! /config → Output style → Learning

Addy Osmani @addyosmani

3 weeks ago

x.com/i/article/2055…

109 604 4K 621K 5K

// Code as Agent Harness // 100+ page report on all things related to agent harnesses. (bookmark it) In particular, the survey summarizes methods and applications of code as agent harness. This paper makes a strong case that code-as-harness might be the key to moving us towards a broader science harness engineering. Is code all you need? Maybe. Regardless, the paper argues that future systems must have the following four properties: executable, inspectable, stateful, and governed. Paper: arxiv.org/abs/2605.18747 Learn to build effective AI agents in our academy: academy.dair.ai

@alexmamyy Brother, what glasses are those? They are cool, I wana buy them!

The Forward Deployed AI Engineer (FDE) Playbook: Structural Mismatches Seeking the AI Last Mile - Why OpenAI and Anthropic are Investing in Consulting Companies

Jane Street pays $650,000 a year for quants. Stanford just released the exact RL-for-trading bible for free. 16 chapters. 0 to algo trader. Asset allocation, market making, American option exercise, full Python code & Colab notebooks. Bookmark & give it a weekend.

Jim Simons agreed to do exactly one long-form interview about his life in math, codebreaking for the NSA, and building Renaissance Technologies. The Simons Foundation filmed it and dumped it on their website behind a Vimeo player nobody could find. Two hours and 40 minutes. Someone reuploaded it to YouTube on a channel with 483 subscribers. 749 people have watched it. The deepest Simons interview on the internet. And almost no one knows it exists. Bookmark it tonight. Then read the article below.

zostaff @zostaff

3 weeks ago

x.com/i/article/2054…

25 52 468 690K 1K

Collection of 127 scientific diagrams for physics and ML github.com/janosh/diagrams

For researchers interested in U.S. banking and finance, we are sharing a new data resource! It contains information on bank balance sheets, bank runs, and bank failures from the 19th century to the present:

x.com/i/article/2054…

Building my own document DB and came across this resource If you’re trying to learn database internals / storage engines in Rust, this site is incredibly good

Very excited to finally release TorchLean publicly! I also wrote a longer blog on why I think this matters: robertj1.com/torchlean_veri… Thread below :)

Prof. Anima Anandkumar @AnimaAnandkumar

4 weeks ago

TorchLean codebase is now available! TorchLean is a Lean 4 framework for verified neural-network software. It supports typed tensors, runnable training, graph IRs, verified autograd, Float32/IEEE semantics, CROWN / IBP-style verification, certificate checking, PyTorch interop,

3 25 97 14K 44

"Mathematical Theory of Deep Learning" is an excellent free resource for anyone interested in the mathematical structure underlying modern deep learning systems. The book introduces the theory of deep neural networks through approximation theory, optimization theory, and statistical learning theory, three of the central pillars of the field. What makes it particularly interesting is its attempt to balance rigor with accessibility, focusing on the essential ideas needed to understand modern AI systems without sacrificing mathematical depth. Despite this clarity of exposition, the book is clearly oriented toward a specialized audience. It is also an enormous cultural contribution and an extremely valuable free resource for students, researchers, and anyone interested in studying deep learning more rigorously. arxiv.org/abs/2407.18384