O'Reilly

"This 2,800-word essay (a 12-minute read) is about how to survive inside the AI revolution in software development, without succumbing to the fear that swirls around all of us. It explains some lessons I learned hiking up difficult mountain trails that are useful for wrestling with the coding agents. They apply to all knowledge workers, I think." Check out new article by Ed Lyons on our #Radar: https://lnkd.in/egsvMNbG

"Imagine you deploy an autonomous AI agent to production. Day one is a success: The demos are fantastic; the reasoning is sharp. But before handing over real authority, uncomfortable questions emerge." Check out #Radar article by Artur Huk to learn more: https://lnkd.in/dcvQDbqY

Over the last few months, almost every leadership conversation somehow lands on #AI. But the real question I keep hearing from leaders is “Okay, but how do I actually use this in my day-to-day work?” That is why this upcoming session caught my attention: “Generative AI and Claude Cowork for Leaders and Managers” by O'Reilly Media, Inc. It is being led by Dave McKeown, someone who has spent years helping leaders move from just managing work to actually developing people and driving outcomes. What I like about this one is that it is practical: - Using Claude to set clear, data-backed goals. - Figuring out what to delegate (and what not to). - Preparing for tough feedback conversations without overthinking them. - Building accountability without micromanaging. - And actually creating a 30-day leadership action plan (not just notes you forget later). - It also shows how AI can work beyond chat across your workflows, documents, and recurring tasks. Date: April 30 If you are a leader, manager, or even an aspiring one trying to figure out how AI fits into your role, this is worth exploring. Happy to share more details. DM me. #Leadership #GenerativeAI #AIinLeadership #FutureOfWork #PeopleManagement #Managers #LeadershipDevelopment #OReilly #AIForWork #DigitalTransformation #CTO #CIO #CISO #CDO

Most engineers can build a system that works. Far fewer can build one that keeps working — at scale, under failure, across years of changing requirements. That gap is exactly what Designing Data-Intensive Applications is about. The first edition by Martin Kleppmann became one of the most-recommended engineering books of the last decade. Not because it listed tools, but because it explained why tools make the trade-offs they do. The 2nd edition — now co-authored with Chris Riccomini — just landed on polish Amazon today. Here's what's new and why it matters: The world shifted. Cloud-native architectures replaced on-prem defaults. Vector databases appeared. LLM pipelines became production problems. The 2nd edition catches up — with updated chapters on cloud data warehouses, vector embeddings, stream processing, and a brand new chapter on the philosophy of building ethical, responsible data systems. 672 pages. 14 chapters. Covers everything from replication strategies and sharding to distributed transactions and consensus algorithms — the stuff that makes or breaks systems at scale. If you work with data at any layer of the stack, this belongs on your shelf. Thanks Martin Kleppmann, Chris Riccomini, and O'Reilly for continuing to raise the bar for engineering literature. One question for you: what's the single hardest distributed systems problem you've had to actually solve in production — not in theory?

Last week on O'Reilly 𝐀𝐈 𝐂𝐨𝐝𝐞𝐜𝐨𝐧 several topics were discussed in the 𝐀𝐠𝐞𝐧𝐭𝐢𝐜 world. These 3 topics I found especially useful, which describe the setup, management and evolution in the agentic workspace. 𝐎𝐫𝐜𝐡𝐞𝐬𝐭𝐫𝐚𝐭𝐢𝐧𝐠 𝐂𝐨𝐝𝐢𝐧𝐠 𝐀𝐠𝐞𝐧𝐭𝐬 🥁 by Addy Osmani https://lnkd.in/dhzaWvxB https://lnkd.in/dekEBrkK -> THE 8 LEVELS OF AI-ASSISTED CODING -> SUBAGENTS -> AGENT TEAMS -> CLOUD ASYNC AGENTS -> THE RALPH LOOP -> 10 PRO MOVES FOR 10X VELOCITY A great resource to see where the industry is heading. 𝐓𝐡𝐞 𝐥𝐞𝐯𝐞𝐥𝐬 𝐨𝐟 𝐀𝐈 𝐚𝐝𝐨𝐩𝐭𝐢𝐨𝐧 ⬆️ 🆙 https://lnkd.in/djxVsdeF 𝐂𝐨𝐝𝐞 𝐅𝐚𝐜𝐭𝐨𝐫𝐲 🏭 by Ryan Carson https://lnkd.in/dqsV6SKY In the conference he showed us his setup 🏆 -> OPENCLAW -> CLOUD SANDBOXES -> ORCHESTRATION -> AUTOMATIONS -> TEAM CONTEXT -> SENTRY -> SLACK This is the future of agentic coding. I started building one over the weekend and running from time to time out of tokens 🥲 𝐓𝐡𝐞 𝐌𝐲𝐭𝐡𝐢𝐜𝐚𝐥 𝐀𝐠𝐞𝐧𝐭-𝐌𝐨𝐧𝐭𝐡 🧞 by Wes McKinney https://lnkd.in/d8-XdY3K Here I liked the 𝐚𝐜𝐜𝐢𝐝𝐞𝐧𝐭𝐚𝐥 𝐜𝐨𝐦𝐩𝐥𝐞𝐱𝐢𝐭𝐲 term, which basically describes what will happen if you let agents build code without some guardrails, verification, and constraints, basically everything that is done by our tools and processes, and that after around 100-200k lines of code the agents collapse under their own built complexity and that the system is not scalable anymore. 𝐌𝐲 𝐭𝐚𝐤𝐞𝐚𝐰𝐚𝐲… As AI and the surrounding ecosystem become more and more advanced, it’s important to start thinking about a high level setup and management of AI workers, and also that a local machine setup is not required anymore using cloud environments. There is more to discover 🧐 What fancy AI workflow did you use? #mcp #agents #openclaw #ai #python #generativeai #cursor #claude #oreilly #ai #artificialintelligence #code #softwaredevelopment #productivity #automation

🚀 𝗬𝗼𝘂𝗿 𝗔𝗜 𝗽𝗶𝗽𝗲𝗹𝗶𝗻𝗲 𝗺𝗶𝗴𝗵𝘁 𝗯𝗲 𝗹𝘆𝗶𝗻𝗴 𝘁𝗼 𝘆𝗼𝘂 𝗮𝗯𝗼𝘂𝘁 𝗶𝘁𝘀 𝗼𝘄𝗻 𝗿𝗲𝘀𝘂𝗹𝘁𝘀. 🚀 𝘔𝘪𝘯𝘦 𝘸𝘢𝘴. 𝘈𝘯𝘥 𝘵𝘩𝘳𝘦𝘦 𝘈𝘐 𝘢𝘥𝘷𝘪𝘴𝘰𝘳𝘴 𝘮𝘪𝘴𝘴𝘦𝘥 𝘪𝘵. This one really got me. My AI blackjack simulation that ran hundreds of hands to compare different strategies was reporting great results. Then I actually looked at the data and a lot of runs were obviously wrong. Turns out the pipeline had a bug where the AI verification step was passing almost everything regardless of the actual results. I asked three different AIs to review the pipeline and none of them caught it! The only thing that exposed it was checking the aggregate numbers, which just didn't add up. The real pass rate was way lower than what the system was reporting. That's scary, right? If you let probabilistic behavior into a step that should be deterministic, everything will look plausible and the system will happily report success. You have no way to know something's wrong until you go looking for it. 𝘊𝘩𝘦𝘤𝘬 𝘪𝘵 𝘰𝘶𝘵 𝘰𝘯 𝘵𝘩𝘦 𝘯𝘦𝘸 O'Reilly 𝘴𝘶𝘣𝘴𝘵𝘢𝘤𝘬: https://lnkd.in/e2JjngTU

There are still some spots left in my LLM Safety and Security course with O'Reilly. Come to this session to learn how large language models, AI assistants, and AI agents work in the context of a larger system. For nearly 20 years, technology solutions have been focused on optimizing tasks. Think iPhone and Android apps, SaaS services, and now AI agents. But we need to bring back the old school thinking of how a given task affects a larger system around it, and how it applies to large language models and AI agents. This is the mindset that keeps airplanes flying safely, and there are some lessons to learn about system safety. Sign up now! https://lnkd.in/gdtPBMf9

Check out my Google Cloud Associate Data Practitioner course on O'Reilly Media. This course is designed to help you prepare for the Google Associate Data Practitioner certification, a valuable credential for those looking to design and build data processing systems in the cloud. The Google Cloud Associate Data Practitioner exam demonstrates skills and knowledge in using Google Cloud technologies for data analysis and management. What you’ll learn and how to apply it: - Understand the exam objectives for the GCP Associate Data Practitioner certification. - Grasp basic data engineering best practices specifically related to Google Cloud. - Learn about extracting and loading data into appropriate Google Cloud storage systems. - Explore important storage services in GCP, including Cloud Storage and Bigtable. - Identify services for structured data in GCP (Cloud SQL, Cloud Spanner, BigQuery). - Identify services for unstructured data in GCP (Cloud Storage, Filestore). - Identify services for semi-structured data in GCP (Cloud Datastore, BigTable). - Apply security measures and ensure compliance with data privacy regulations. - Discover various data management tools in GCP, such as DataPrep and Data Catalog. - Learn visualization techniques and utilize GCP Services like Data Studio. - Identify high availability and disaster recovery strategies for data in Cloud Storage and Cloud SQL. - Gain insight into AI/ML services for BigQuery ML and AutoML. #CloudComputing #DataEngineering #DataScience #GoogleCloud #GCP #DataAnalytics O'Reilly Pearson Digital Crest Institute For more information, visit: https://lnkd.in/eFxh7HPx