Douglas Pinheiro @dgsdeveloper

Publiquei meu primeiro livro/e-book independente (self-published), e é de graça! Se chama "Competente o Suficiente". ailair.akitaonrails.com/pt-br/book/ Digamos que é a "novelização" de todo o trabalho que eu este ano e vinha publicando no meu blog, mas agora num formato que - eu espero - seja mais fácil de digerir, com começo, meio e fim. Glossário, Bibliografia, tudo que um livro precisa ter! Quem se interessar mais, também publiquei o "making-of" no meu blog: akitaonrails.com/2026/10/13/aut… E está tudo open source no GitHub: github.com/akitaonrails/a… Curioso pra saber o que vocês acham. Este ainda é 1.0-preview3, vou continuar ajustando e esse é o momento pra feedback e sugestões!

keyboard_arrow_left Previous keyboard_arrow_right Next

Fabio Makita @AkitaOnRails

2 days ago

Se eu compilasse tudo que pesquisei e experimentei sobre IA na forma de livro/ebook, vocês teriam interesse? 🤔

68 5 471 23K 45

Finalmente! Saiu o resultado do novo LLM Coding Benchmark V4!! Foram 10 dias rodando 24/7, milhões de tokens e dezenas de horas de compute pra conseguir o melhor benchmark de programação prática que eu consegui até agora. Re-testei 39 das principais LLMs do mercado, de Astra e Fable, até Deepseek ou GLM, até Qwen 3.8 Max e 27bi Local! Ficou tão longo, que dividi em 2 partes. Pra entender o histórico, os objetivos, a metodologia, siga pra parte 1: akitaonrails.com/2026/09/15/nov… Pra ir direto pros resultados, tabela de ranking, análises, prós e contras de cada categoria de modelos, siga pra Parte 2: akitaonrails.com/2026/09/15/nov… A diferença deste V4 em relação aos anteriores é que em vez de mandar fazer um app, agora eu simulo mini-sprints (7 no total), e simulando um programador malicioso externo adicionando código malicioso e bugs sutis pra ver quem consegue chegar até o fim com um app que funcionaria em produção (nada de leetcodes ou similares). E por isso cada execução pode levar mais de 1h (ou 12 horas no caso do Qwen 3.8 27bi local). Espero que gostem!

We're publishing our most detailed threat intelligence report to date. It covers how people tried to misuse Claude—for cyberattacks, influence operations, surveillance, biology, and building weapons—and how we found and stopped them. We disrupted every operation in the report, and used the lessons from them to strengthen our safeguards. Where appropriate, we also shared what we found with authorities and other AI companies. These cases are not typical: we’re highlighting some of the most sophisticated misuse we’ve seen. But they’re especially important to discuss, because they show us where AI misuse is headed, where our safeguards work, and where they need to improve. We’re publishing this report so others can spot the same activity on their own platforms, and so we can give the public a clearer view of how emerging threats develop. Read the report: anthropic.com/threat-intelli…

Apple avoided foldables for years. Then dropped the iPhone Duo yesterday. The shit people are doing with it is insane. 10 wild videos: 1. Netflix on a 7.6" screen

9 engineering blogs I keep coming back to: 1. Netflix TechBlog 2. Uber Engineering 3. Cloudflare Blog 4. Engineering at Meta 5. LinkedIn Engineering 6. Discord Engineering 7. AWS Architecture 8. Slack Engineering 9. Stripe Engineering

AI Engineering — Ultimate Roadmap (SAVE THIS)! ├── 01. LLM Fundamentals │ ├── Transformers & Attention │ ├── Inference & Decoding │ ├── Tokenization │ └── Context Window & Limits │ ├── 02. Prompting & Context │ ├── Prompting & In-Context Learning │ ├── Context Engineering │ ├── Structured Outputs & Schemas │ └── Constrained Decoding │ ├── 03. Representations │ ├── Embeddings │ ├── Similarity Search │ ├── Vector Databases │ └── Hybrid Retrieval │ ├── 04. RAG │ ├── Chunking │ ├── Hybrid Search │ ├── Reranking │ ├── Query Rewriting │ └── GraphRAG │ ├── 05. Tools & Agents │ ├── Tool / Function Calling │ ├── Agentic Workflows │ ├── Agent Memory & State │ ├── MCP │ ├── A2A / Interoperability │ └── Human-in-the-Loop │ ├── 06. Evaluation │ ├── Evals │ ├── Test Datasets │ ├── Online Metrics │ └── Tracing & Debugging │ ├── 07. Safety │ ├── Guardrails │ ├── AI Security │ └── Prompt-Injection Defense │ ├── 08. LLMOps & Serving │ ├── Model Selection & Routing │ ├── Caching, Latency & Cost │ ├── vLLM / TGI │ ├── Quantization & Batching │ └── Production Lifecycle │ ├── 09. Adaptation │ ├── Fine-Tuning │ ├── PEFT │ ├── LoRA / QLoRA │ └── Distillation │ ├── 10. Advanced │ ├── Multimodal AI │ ├── Knowledge Graphs │ ├── AI App Architecture │ └── Product Integration Learn to ship AI systems, not just prompt models.

System Design Concepts Ranked by Difficulty 🟢 Load Balancing — Easy 🟢 CDN — Easy 🟢 Caching — Easy 🟢 API Gateway — Easy 🟢 Rate Limiting — Easy 🔵 Message Queues — Moderate 🔵 Database Sharding — Moderate 🔵 Replication — Moderate 🔵 Reverse Proxy — Moderate 🔵 Consistent Hashing — Moderate 🟠 Microservices — Hard 🟠 Event Driven Architecture — Hard 🟠 Distributed Transactions — Hard 🟠 Service Discovery — Hard 🟠 Circuit Breakers — Hard 🔴 Distributed Systems — Very Hard 🔴 Consensus Algorithms — Very Hard 🔴 Leader Election — Very Hard 🔴 Distributed Consensus — Very Hard 🔴 Fault Tolerant Systems — Very Hard Which concept deserves 🔴? 👀

AI/ML GOLDMINE 📌 20 repos that cover everything from ML basics to LLMs, agents, and production AI. 1. Machine Learning for Beginners github.com/microsoft/ML-F… 2. Hands-On Machine Learning github.com/ageron/handson… 3. Made With ML github.com/GokuMohandas/M… 4. fast.ai github.com/fastai/fastbook 5. PyTorch github.com/pytorch/pytorch 6. Hugging Face Transformers github.com/huggingface/tr… 7. LLMs from Scratch github.com/rasbt/LLMs-fro… 8. LLM Course github.com/mlabonne/llm-c… 9. Hands-On Large Language Models github.com/HandsOnLLM/Han… 10. AI Agents for Beginners github.com/microsoft/ai-a… 11. GenAI Agents github.com/NirDiamant/Gen… 12. AI Engineering Hub github.com/patchy631/ai-e… 13. Awesome LLM Apps github.com/Shubhamsaboo/a… 14. RAG Techniques github.com/NirDiamant/RAG… 15. vLLM github.com/vllm-project/v… 16. llama.cpp github.com/ggml-org/llama… 17. Annotated Deep Learning Papers github.com/labmlai/annota… 18. OpenAI Cookbook github.com/openai/openai-… 19. AI Engineering Portfolio Projects github.com/landedjobs/ai-… 20. Awesome AI Engineering github.com/Guidely-org/aw… Don't try to finish all 20. Pick the ones that match where you are:

🚨 BREAKING: Google just launched free AI courses. No sign-up fees or prior skills required. Here are 10 courses you don't want to miss:

How does Claude's text watermark work? Anthropic recently shared their intent to watermark text so they can identify AI-generated text. This post is based on my understanding of how it works. LLMs produce text word by word. At each step, they generate probabilities for the next likely word. Instead of sampling randomly from those words, the watermarking trick changes which words are allowed to be picked. How to watermark a response? Step 1: The model produces probabilities for the next word. Step 2: Normally a random number generator picks one of the good candidates. With watermarking, a keyed function takes a secret key plus the previous few words and decides which candidates are valid to pick from. Step 3: This repeats for the whole response. Places where there are multiple plausible choices carry the watermark signal. How to detect a watermarked text? Step 1: For any candidate word in the text, we check whether it is a valid choice based on the secret key and the few preceding words. If the word is valid, that is counted as a match. Step 2: Run this across the entire text. Watermarked text matches far more often. The overall match rate can be treated as an AI-generated score. I’m personally getting quite annoyed by the false negatives from all these AI text detection techniques, especially for technical writing. What's your thoughts on AI text detection? Do you think AI text detection is useful, or will it create more problems?

Claude Academy is now live. Whether you're figuring out what AI is or already using Claude every day, there's a path that meets you where you are. The courses and tutorials are free and open to anyone at academy.claude.com

📢Meet Qwen3.8-Max — our most capable model to date. Next week, the open weights of Qwen3.8-Max will be released, and Qwen3.8-27B is also going open-weights to meet you all!🎉 Qwen3.8-Max, a new bar for coding and cowork at 2.4T parameters: - Autonomous coding: 10+ days of self-evolving development, from empty folder to production without hand-holding, complete project trace in the GitHub：github.com/qwen-code-dev-… - Real work, real results: Production-quality deliverables across hundreds of professions. - Long-horizon mastery: System-level autonomous planning with closed-loop adaptive learning, driving 500+ turns of chip design optimization and 365 days of e-commerce strategy. - Native multimodal intelligence: Vision isn't just input — it's a continuous feedback loop for planning, execution, and self-correction. 💰Pricing: Input: $2.0 / M tokens Output: $6.0 / M tokens Implicit Caching: $0.25 / M tokens Start building with Qwen3.8-Max! 🚀 📖 Blog: qwen.ai/blog?id=qwen3.8 ✅ Qwen Studio: chat.qwen.ai/?models=qwen3.… ⚡ API: qwencloud.com/models/qwen3.8…

El ingeniero que construyo Claude Code acaba de soltar un video de 28 minutos donde te enseña a escribir prompts que realmente funcionan He visto cursos de 300 dolares que no llegan ni a la mitad de lo que explica en los primeros 10 minutos Archivos CLAUDE.md, atajos de memoria, sesiones paralelas y patrones de prompting que cambian el juego Todo en un solo video y completamente gratis Da igual si eres principiante, desarrollador o ya llevas meses usandolo Esto te va a hacer usar Claude como nunca antes

This Friday, July 17th at 7pm UTC, the Story of Java is dropping on CultRepo. Join James Gosling in the chat! youtu.be/ZqGSg4b_cZA?si…

When writing a specialization, be careful about its location, or to make it compile will be such a trial, as to kindle its self-immolation A sneaky bit of C++ lore from Andrew Koenig, who was a Bell Labs computer scientist and a founding member of the C+++ Standards Committee. Watch the full C++ documentary here: youtu.be/lI7tMxzSJ7w?si…

Union types are coming to C#! Unions model closed sets of data shapes, as commonly seen in e.g. wire protocols. In this #MSBuild session, we explore the clean expression of intent and the confidence and elegance that unions lend to consuming code. 🎥 buff.ly/yigYfF4

Starlink V3 satellites have >10X bandwidth of V2 and there’ll be >10X launched, which means >100X more bandwidth. Also, altitude will be 350km vs 550km, so min latency can be cut in half. Light travels 300km/ms in space, so physics round trip min latency drops to <5ms.

Andrej Karpathy (formerly OpenAI) just announced he's joining Anthropic. His own words: "I think the next few years at the frontier of LLMs will be especially formative. I am very excited to join the team here and get back to R&D." Co-founder of OpenAI. Former Director of AI at Tesla under Elon Musk. PhD from Stanford under Fei-Fei Li. The man who coined "vibe coding." One of the most respected AI researchers alive. He left OpenAI in 2024. Built an AI education company. Now he's walking into Anthropic's R&D lab. This is not a PR hire. This is one of the architects of modern AI choosing which lab he thinks is going to matter most in the next decade. OpenAI just lost its founding talent to the competition. Again.

keyboard_arrow_left Previous keyboard_arrow_right Next

Mario Nawfal @MarioNawfal

5 months ago

Boris Cherny, the creator of Claude Code at Anthropic, just explained how to write prompts that actually work CLAUDE.md files, memory shortcuts, parallel sessions, and prompting patterns all in one video and completely free

41 587 3K 526K 6K

Você está cursando TI mas acha que sua faculdade não é profunda o suficiente? Eu analisei a grade curricular do MIT, stanford, harvard, e a minha (UFABC) e compilei 46 livros usados de verdade nessas universidades nas mais diversas matérias da computação. (1/3)

keyboard_arrow_left Previous keyboard_arrow_right Next

🚀 DeepSeek-V4 Preview is officially live & open-sourced! Welcome to the era of cost-effective 1M context length. 🔹 DeepSeek-V4-Pro: 1.6T total / 49B active params. Performance rivaling the world's top closed-source models. 🔹 DeepSeek-V4-Flash: 284B total / 13B active params. Your fast, efficient, and economical choice. Try it now at chat.deepseek.com via Expert Mode / Instant Mode. API is updated & available today! 📄 Tech Report: huggingface.co/deepseek-ai/De… 🤗 Open Weights: huggingface.co/collections/de… 1/n