LLMs aren't just getting better at generating answers. They're getting better at working through problems. @rasbt and @marlene_zw break down reasoning models, inference scaling, training techniques, and their role in AI agents in this @acmeducation TechTalk: hubs.la/Q04yhnm80 Go deeper with Sebastian's book, Build a Reasoning Model (From Scratch): hubs.la/Q04yhncM0

AI can give you the Ruby syntax you ask for. A good book can teach you the things you didn't know to ask about. The Well-Grounded Rubyist, Fourth Edition goes deep into Ruby's ideas, mechanisms, and fascinating corners. Now in print: hubs.la/Q04ygZ9C0

Not every AI failure needs the same eval. Retrieval degradation, confidently wrong agents, and silent quality drift each fail differently. And your evals need to account for that. On Sept. 30th, Rush Shahani will show how, including a live walkthrough of an eval suite from Building Reliable AI Systems. It'll be at 11:30 AM ET — register here: hubs.la/Q04yg40M0

Save 50%, only on Sept. 25th! Today, you can save on on our new MEAP, Claude at Work by @EltonStoneman, and other related titles: hubs.la/Q04yft8_0 Learn how to make Claude a hyper-productive working partner with its agentic code generation to assistants that capture, execute, design, testing, and more.

Writing embedded Rust is only part of the job. You also need a toolchain that can take your code from development to the target hardware. Tomorrow, September 25th, Dr. David Cabanis will show how rustup, Cargo, GDB, QEMU, and Renode fit into a practical embedded Rust workflow. Register for the @DoulosTraining event: hubs.la/Q04y86Ct0 Learn more in his book, Embedded Software with Rust: hubs.la/Q04y86Bz0

An LLM app working in testing is great. Keeping it working when models change, users surprise you, and the system hits production is even better. Tomorrow at Serverless Toronto, Serj Smorodinsky and Brett Kennedy will cover how to baseline performance, define what "good" means, and use @DSPyOSS to optimize against measurable results. They'll also get into model swaps, cost and privacy tradeoffs, and ongoing evaluation in production. Join virtually Sept. 25th, 12–1:30 PM ET: hubs.la/Q04y8f670 Check out their book, Building LLM Applications with DSPy: hubs.la/Q04y8f680

Build a Reasoning Model (From Scratch) nitter.net/i/broadcasts/1qGvveXDq…

When software controls flight systems, railway signals, defibrillators, or autonomous vehicles, "it passed the tests" isn't enough. Memory Safe C++ shows you how to build verifiable, deterministic-by-design C++ systems and identify risks that testing can miss. Now in MEAP and 50% off through October 7th: hubs.la/Q04ybfxf0

Save 45%, only on Sept. 24th! Today, you can save on Bug Bash by Lina Zubytė and other related titles: hubs.la/Q04y86050 Learn how to run a time-boxed testing event with participants from across an organization.

More autonomous AI agents need better ways to see what they're doing and evaluate whether they're doing it well. Build an AI Agent (From Scratch) shows you how to build an observability and evaluation pipeline with OpenTelemetry and an automated LLM-as-a-judge system. It's on sale on Amazon: hubs.la/Q04y4x9K0

Using AI is useful. Knowing how it works under the hood gives you a much deeper understanding of what you're building with. Today's From Scratch Deal of the Day covers a wide range. Some examples include: • Reasoning models with Build a DeepSeek Model (From Scratch) • AI orchestration with Build an Orchestrator in Go (From Scratch) • And text-to-image generation with Build a Text-to-Image Generator (From Scratch) by @mark_h_liu Get these and more From Scratch titles for 50% off today: hubs.la/Q04y4tJ-0

That Manning to-read list getting a little out of hand? You can tackle it with a subscription. Get a year of access to Manning's digital library for $199.99. That's 20% off for today only. That'll make your backlog a little more manageable. The full details: hubs.la/Q04y3JsJ0

What does it actually mean to program a quantum computer? On Oct. 2nd, Mariia Mykhailova will explore how writing quantum algorithms as code helps you test ideas, reason about correctness, and put quantum computing concepts into practice. She'll also demo Workbench Quantum Katas, her latest project for hands-on quantum programming practice. Register for the 12 PM ET event: hubs.la/Q04y3M880

Save 45%, only on Sept. 22nd! Today, you can save on Acing the Frontend Interview by @JenniferFu88011 and other related titles: hubs.ly/Q04y377j0 Learn what employers are looking for when hiring frontend roles, and how you can show those qualities at interview.

Using Claude effectively isn't just about generating code. It's about learning how to manage AI agents across the development lifecycle. Claude at Work by @EltonStoneman shows you how to use @claudeai for specs, architecture, coding, testing, and more while keeping control of the process. Now in MEAP and half off through October 7th: hubs.la/Q04y32KD0

Hybrid search typically fuses keyword and semantic results. Wormhole vectors take a different approach: moving between sparse lexical and dense semantic vector spaces. At @OpenSearchProj on September 22nd, Trey Grainger will show how it works in OpenSearch and what it could mean for hybrid and agentic search. AI-Powered Search is also 45% off with code OpenSearchCon through October 6th: hubs.la/Q04xQfBn0

How do we use AI in education without losing sight of how people actually learn? That's at the heart of @wailsworkshop 2026, Dec. 9–11th in Paris, where researchers will explore responsible and equitable AI in learning sciences. Learn more: hubs.la/Q04xQb1C0 Dig deeper with our WAILS book bundle. Get 2+ books and save 50%: hubs.la/Q04xQb390

Useful AI output is only part of the problem. The systems around it need to be structured, predictable, and designed to handle increasingly complex work. Three books tackle different pieces of that: • Spec-Driven Development by @HariKrishnan83 for more structure for AI-assisted development: hubs.la/Q04xQ0H_0 • Designing AI Agents by @shidaren for better-designed agentic systems for complex tasks: hubs.la/Q04xQ0H00 • Context Engineering by @boni_gg for more deliberate control over the context AI works with: hubs.la/Q04xQ0J00 Get 40% off when you pick up 2+ Manning books this week.