Category: Daily Reading List

Day two of vacation included me hitting golf balls and eating lunch overlooking the ocean. And a couple of unexpected work calls. But overall, definitely feeling better and more energized already.

[article] An Interview with Google Cloud CEO Thomas Kurian About the Agentic Moment. I’m absurdly biased, but I simply don’t see other CEOs demonstrate command of issues the way Thomas Kurian does in long-form interviews like this.

[blog] AI in DevOps: Why Adoption Lags in CI/CD (and What Comes Next). It’s an interesting point of view. I’m not sure the conclusion is right. Are devs slow to add AI to CI/CD because of trust issues, or because the current landscape of products isn’t suited for AI workloads or surfacing the right AI integration points?

[article] “Developer loyalty is at zero right now”: Google doesn’t care which AI coding tool you use. Thomas comes across measured and cool in the topmost interview. I come across as a wildcat in this one. I guess we balance each other out.

[article] Vibing, Harness and OODA loop. You had me at OODA loop, as I’m a big Boyd fan. Just because AI helps us “act” faster, doesn’t mean we should skip the other important parts of the loop.

[blog] Agent Chaos to Engineered Intelligence – Key Takeaways from Google Cloud Next 2026. Paul’s a good industry analyst and has a extracts a few important points about our story for developers.

[article] What 4 engineers with 10+ years of experience say about staying relevant in the AI era. Everybody’s trying to figure out what work looks like. I don’t trust anyone who acts too confident. Read things like this to observe the thought process of others.

[blog] Agent Memory Patterns. This post calls out a few places to store mutable memory. It focuses on files, memory blocks, and skills. I’m not sure how you’d categorize external durable memory services.

[blog] Building real-world on-device AI with LiteRT and NPU. On-device AI will keep growing in popularity. I’m friends with the PM of this product and am trying to learn more about this space.

[article] The End of One-Size-Fits-All Enterprise Software. What workflows do you want to control yourself? You’ll build, compose, collaborate, or buy software to drive it.

[blog] We’re donating Agent Payments Protocol to the FIDO Alliance to support the future of secure, agentic payments. You love to see it. Open standards are useful in domains like this, and we’re playing our part.

[blog] 8 Cursor Rules for Go Developers — 2026 Edition. Good, structured code still matters. These are some rules to stash in your agentic IDE to get cleaner Go code.

[article] Google begins putting the guardrails on agentic AI. It’s not about flashy features right now, but who’s containing agents and making them manageable.

[blog] The Future of Agentic AI: Inside Microsoft Agent Framework 1.0. Glad to see Microsoft betting on the A2A Protocol for their Python/C# agent framework.

[blog] It’s all about the angle: Your photos, re-composed. Super cool. It uses AI to alter the perspective in existing photos.

Want to get this update sent to you every day? Subscribe to my RSS feed or subscribe via email below:

I’m on vacation this week, hoping to recharge after last week’s shenanigans. But that also means more time for reading. So besides spending today at the San Diego Zoo, I read enough to publish a reading list.

[blog] From “Hello World” to “Hello Agents”: The Developer Keynote That Rewired Software Engineering. This is quite the writeup from last week’s developer keynote at Google Cloud Next.

[article] Google Cloud Next 2026: Pichai and Kurian on Why Cloud Is Now an Operating System, Not a Service. The recaps keep coming, and this one makes some astute observations about the competitive landscape.

[article] The 20 Software Engineering Laws. I read this over the weekend, but it’s worth including in today’s list. Great look at familiar (and unfamiliar?) principles that might shape how you approach engineering work.

[blog] Stitch’s DESIGN.md format is now open-source so you can use it across platforms. Don’t sleep on this. Design tools might get the interop spec they need.

[blog] ReasoningBank: Enabling agents to learn from experience. Unlike other approaches, this strategy analyzes and uses successful and failed experiences to improve agents after deployment.

[blog] Roadmaps: Preparing To Self-Destruct In Three, Two, One … The era of intense annual planning and rigorously maintained roadmaps is on pause. Maybe dead. Unless you’re doing physical construction, you should be focusing on shorter time horizons and continuous planning.

[article] Google CEO Sundar Pichai says 75% of the company’s code is AI-generated. Hey, it’s me (and my buddy Sundar) quoted in this Fast Company article.

[blog] A good AGENTS.md is a model upgrade. A bad one is worse than no docs at all. Very actionable advice here on how to help (or hurt!) your agent trajectories with these documents.

[blog] All your agents are going async. This is focused on human-agent patterns (like with your coding tool). But those building agentic systems are already working through async patterns for triggering, state management, and gathering responses.

[blog] This Is How To Improve Your Writing: 6 Expert Insights. Every single one of these insights is spot on. So many great thoughts go unseen because the thinker can’t communicate them effectively.

[blog] AI threats in the wild: The current state of prompt injections on the web. What are indirect prompt injection patterns, and what do they look like in the wild? Our Google Security team does a sweep.

[article] AWS Ends WorkMail and Moves App Runner to Maintenance Mode. For some reason, no other cloud besides Google Cloud can get traction with app container runtimes. So strange.

[blog] Next ’26 Hands-On: 10 Codelabs to Build Featured Tech. We added well over a hundred new codelabs to our catalog last week to help devs go hands-on with all the tech they saw at Cloud Next. Here are some highlights.

Want to get this update sent to you every day? Subscribe to my RSS feed or subscribe via email below:

Skipped yesterday’s post because I had virtually zero time to read thanks to Google Cloud Next activities. Watch the keynote replay and tell me what you think!

[blog] Day 2 at Google Cloud Next: A marathon developer keynote. This is a super recap of our unique developer keynote yesterday where we stayed with one theme for all the technical demos.

[repo] Race Condition. That gorgeous marathon planning system we showed off on stage yesterday? Yup, we open sourced all of it. Have fun!

[article] Cognitive debt: The hidden risk in AI-driven software development. What is cognitive debt, where does it impact you, and how can you mitigate it? This article goes fairly deep.

[blog] DeepSeek-V4: a million-token context that agents can actually use. Another impressive model family from the DeepSeek team. More here.

[blog] Introducing Gemini Enterprise Agent Platform, powering the next wave of agents. Major advances in multiple directions here. Check out this post for a flyover view.

[article] Google’s Gemma 4 shines on local systems – both big and small. Here’s an assessment of our latest open models and if they perform as advertised.

[article] The modern data stack was built for humans asking questions. Google just rebuilt its for agents taking action. Openness is a differentiator for us, along with our view of semantic context.

[blog] The future of data lakehouse: Open and interoperable for the agentic era. On the topic of openness, here’s more on where we’re taking the lakehouse.

[blog] Introducing the Google Cloud Knowledge Catalog. Does a “universal context engine” sound interesting to you? It does to me.

[article] The Pulse: ‘Tokenmaxxing’ as a weird new trend. You want to gamify token usage, go ahead. It’s great for Google’s revenue. But I wouldn’t suggest it. Focus on desired outcomes, not weak proxy metrics.

[blog] Next-gen FinOps for the AI era. Speaking of cost controls, I’m happy that Spend Caps are coming to services in Google Cloud.

[article] Google splits its TPU line in two for the agentic era. We don’t always share the same opinions as other vendors, and here’s a case of divergence from AWS.

Want to get this update sent to you every day? Subscribe to my RSS feed or subscribe via email below:

Amazing first day of Google Cloud Next. Keynotes were fun and had a ton of productive customer chats.

[blog] Welcome to Google Cloud Next ‘26. Great roundup of what we’re announcing today, and why it matters.

[blog] The new Gemini Enterprise: one platform for agent development, orchestration, and governance. You deserve nice things! We’ve juiced-up our platform for building agents and doing work.

[blog] Gemini Enterprise for the agentic task force: introducing long-running agents, agentic collaboration spaces, advanced governance, and more. This post goes deeper into the Gemini Enterprise app and how we all can do more productive and interesting work.

[blog] Announcing Spanner Omni: Your infrastructure, Google’s innovation. The best data service in the cloud available … anywhere? Yup, you can now bring Spanner with you to run on any infrastructure.

[blog] What’s new with Databases: Powering the agentic future. Plenty of new and exciting things from our Data team.

[blog] What’s new in Cloud Run at Next ‘26. We haven’t let up on our serverless investment, and it’s a one-horse race right now. Cloud Run continues to introduce valuable and innovative capabilities.

[blog] What’s new in GKE at Next ‘26. Kubernetes is already a fundamental component for many companies, and with this AI surge, it’s finding new ways to be useful.

[blog] What’s next in Google AI infrastructure: Scaling for the agentic era. Compute, networking, and storage are still the bread-and-butter of a hyperscaler, and ours should be second-to-none.

[blog] Inside the eighth-generation TPU: An architecture deep dive. Big improvements, and two distinct systems.

[blog] Level Up Your Agents: Announcing Google’s Official Skills Repository. Fantastic work from my team here. We built a foundational set of agent skills you can use with your favorite agentic harness.

[article] 15 principles for managing up. Really good. “Managing up” has a bad rap, but it’s important to effectively and clearly communicate with the boss.

[article] Eclipse Foundation offers enterprise-grade open source alternative to Microsoft’s VS Code Marketplace. Already popular, this registry has over 10k extensions. I’m glad Google is a supporter.

[blog] Introducing the Builders Hub from the Google Developer Program. Tremendous work from my team to build this new experience. It pulls together more of Google’s dev products into a single place.

[article] SpaceX is working with Cursor and has an option to buy the startup for $60B. Wouldn’t have predicted that, but it also makes sense given the ambitions of those involved.

[blog] A Guide to 5 Agent Payment Protocols. It’s wild there are five of these already. But they don’t completely overlap. This post tries to separate the use cases for each.

[article] Employers say they struggle to find workers with the right AI skillset. This is specific to graduate hires. Universities need to quickly reset to ensure their churning out people who have the skills the market is willing to pay for.

Want to get this update sent to you every day? Subscribe to my RSS feed or subscribe via email below:

Google Cloud Next is nearly here, and today’s pre-event activities were hopping. I got to meet with our Google Developer Experts, have a handful of customer meetings, and spend time with my team.

[blog] Building Gemma 4 Local-Powered LLM Apps with Go and Yzma. Solid example of building an LLM app with Go and running the open Gemma model on a local machine.

[blog] most devs ignore git worktree. heres why they’re wrong. I’m sold. This is a short post, but it clearly explains the git worktree model and how to use it.

[blog] Agent Harness Engineering. Is it a model problem? Or is it a configuration problem with the harness? Lots of smart insights in this post.

[blog] Drop to Demo: Gemini CLI Subagents. If you’re truly going to get into agent-first engineering, then get familiar with coordinating a team of agents.

[article] GitHub pauses new Copilot sign-ups as agentic AI strains infrastructure. Why are so many AI services all-of-a-sudden dealing with capacity issues? This article makes the point that our interactions with AI have gotten longer and more sophisticated. That consumes a lot of compute!

[blog] How to fine-tune LLMs in 2026. Educational post. I’m not a machine learning engineer, but most of us could understand what this is all about.

[article] The engineer AI can’t replace. Just because we produce code faster, doesn’t mean we stop thinking about it. This post explores “taste” and the experience to know what good output looks like.

[blog] Building Event-Driven Data Agents with BigQuery, Pub/Sub, and ADK. I really liked how this whole solution came together in the post. The flow from real-time stream processing to agent-based investigation was smooth.

[article] What’s 🔥 in Enterprise IT/VC #494. Good edition this week. After finishing it, I was struck by how few things are “known” right now in IT. What’s the right pricing model for AI? Who are the builders? How to manage AI budgets? Which trend has staying power? Nobody knows. Just learn faster than others.

{article] GitHub Acknowledges Recent Outages, Cites Scaling Challenges and Architectural Weaknesses. Our industry depends a lot on a healthy GitHub. Here’s hoping that they’re successful in fixing their reliability issues.

Want to get this update sent to you every day? Subscribe to my RSS feed or subscribe via email below:

I’m in Las Vegas this week for Google Cloud Next ’26, and spent the last two (full) days in various keynote rehearsals. Fun stuff with people I enjoy hanging out with.

[blog] The Agent Stack Bet. Important post. Have we given agents too much … agency? Addy calls out four architectural bets to make to get back in control.

[blog] How to hire people who are better than you. Great stuff if you’re trying to improve who and how you hire. Which I think is all of us!

[blog] From operational to analytical: The unified Spanner Graph and BigQuery Graph solution. I haven’t used many graph databases, but it seems convenient to have these capabilities built into database engines I already use.

[article] When Your Ambition Starts to Exhaust You. These are five helpful questions to ask yourself.

[guide] Multi-agent private networking patterns in Google Cloud. Very good content here for anyone working through multi-agent setups.

[article] ‘Tokenmaxxing’ is making developers less productive than they think. I understand why some think this is a good proxy metric for meaningful adoption internally. But it’s far from an indicator of excellence.

[blog] Designing synthetic datasets for the real world: Mechanism design and reasoning from first principles. Here’s a new Google Research project that produces better synthetic datasets with the help of AI.

Want to get this update sent to you every day? Subscribe to my RSS feed or subscribe via email below:

Heading to Las Vegas tomorrow for the prep ahead of Google Cloud Next 2026. Today was all about clearing out my queue of work beforehand. I also updated my bio here to reflect a change in my job.

[article] Salesforce launches Headless 360 to turn its entire platform into infrastructure for AI agents. The first move from most SaaS vendors was to offer their own agents and agent platforms. Build gravity. The second play is to ensure their platforms can play in any agent ecosystem.

[blog] A new way to explore the web with AI Mode in Chrome. People seem to like this change, judging from posts and comments I saw online.

[article] Canva won’t make you use its new AI. Nice. AI should be one way you interact with a product. Not the ONLY available way.

[blog] A year of open collaboration: Celebrating the anniversary of A2A. This community accomplished a LOT in the past year. The protocol itself improved, vendors adopted it, and agent frameworks supported it.

[blog] Build to Learn vs Build to Earn. Whether you’re a product manager or not, know when you’re doing discovery and learning, and when you’re actually building the real product.

[blog] A2UI v0.9: The New Standard for Portable, Framework-Agnostic Generative UI. I’m going to spend more time with this. It’s gotten much easier to empower agents to create UIs.

[article] How does codebase familiarity shape collaboration with coding agents? I didn’t find this research surprising. Core maintainers are more diligent on the codebase than casual contributors who offload more work to AI agents.

[blog] Bringing AI Closer to the Edge and On-Device with Gemma 4. Post from NVIDIA with all sorts of details about this open model and how to run it.

[article] Why AI readiness training fails. I think we all know that any training that isn’t applied within two weeks is forgotten. Can we rethink how we’re upskilling people with AI.

[blog] Streaming Gemini 3.1’s expressive new TTS model in Java. Sweet example of this frankly amazing text-to-speech update for Gemini.

[article] Sleep Deprived. I don’t think more of people who brag about not sleeping. That seems like a dumb decision. I’ll be tested next week in Vegas, but still plan to prioritize rest.

Want to get this update sent to you every day? Subscribe to my RSS feed or subscribe via email below:

No time to build personal projects today, but it was rewarding to check in on next week’s announcements, launches, and the very cool stuff my own team is shipping at Google Cloud Next.

[article] Reluctantly Influential: Inside Lenny Rachitsky’s Demandingly Chill Life. Such a good piece about someone I admire. He’s a more successful version of me 🙂

[blog] Multi-Agent A2A with the Agent Development Kit(ADK), Cloud Run, Agent Skills, and Gemini CLI. This is a good combo to get familiar with. Even if you swap out components, they represent smart categories to invest in.

[article] Where will developer wisdom come from? Do you still need wisdom about how to build software? Or is AI wiser about how to write good code? Provocative article.

[blog] Building End to End AI Agents with ADK Dev Skills. I made a big mistake not using these ADK skills when building an agent last weekend. The process was fine, but this would have sped me up even more.

[article] AI Is Writing Our Code Faster Than We Can Verify It. Review every line, or give up entirely? There’s a third choice, as outlined here.

[blog] MCP Toolbox v1.0: The Open-Source Framework for Secure Agentic Data Access. This open project got better month over month, and now it’s a great offering for those bringing data to AI agents.

[article] Why Leaders Need “Power Skills.” Hard skills get you promoted, but “power skills” keep you there. This article looks at how you connect with others and demonstrate higher-order leadership.

[blog] Jaspr: Why web development in Dart might just be a good idea. We used this ourselves to rebuild the Dart and Flutter websites.

[blog] We rebuilt Flutter’s websites with Dart and Jaspr. And here is that particular story. I’m tempted to build something with Jaspr this weekend.

[article] Cost concerns loom as agentic AI advances in software development. The costs will feel high while we’re in the in-between stage of still relying solely on people while introducing AI that changes that dependency.

[blog] Prepay for the Gemini API to get more control over your spend. Prepay billing in the Gemini API may. help with cost concerns.

[article] Anthropic releases Claude Opus 4.7, narrowly retaking lead for most powerful generally available LLM. New Anthropic model for you. And we used this chance to ship multi-region endpoints for this model in Vertex AI.

[article] OpenAI takes aim at Anthropic with beefed-up Codex that gives it more power over your desktop. I wonder how the growth of these “everything apps” squeeze out other tools over time.

[blog] Building a Scalable RAG Backend with Cloud Run Jobs and AlloyDB. If you do it right, you don’t need expensive, always-on hardware to do some pretty special things.

[blog] Your Dependencies Are Someone Else’s Attack Surface. Good post. If you have humans or agents coding, have the right security checks in place.

Want to get this update sent to you every day? Subscribe to my RSS feed or subscribe via email below:

I had lunch with a friend today who showed me how he built an agent team with Claude Code, and the resulting software he completed in days instead of months. It was eye-opening and inspired me to go back to the office and start recreating his work with Gemini CLI subagents.

[blog] How Do We Get Developers to Read the Docs. Are you serving both the “skim it to find an answer” and “show me every decision” users in your docs?

[article] AI discovers bugs faster than teams can respond. The answer will have to be “respond better” not “slow down your bug discovery.” Right?

[blog] Gemini 3.1 Flash TTS: the next generation of expressive AI speech. I’m looking forward to how better AI voice generation changes learning experiences. Imagine being able to learn in your language and vibe of choice!

[blog] Anti-patterns in event modeling – Passive-Aggressive Events. A wonderfully geeky middleware architecture post, and I’m down for it.

[blog] Chrome Skills: Getting Started with Google Cloud examples. Now this makes sense to me. Great examples from Romin that show why Chrome Skills might actually be a big deal.

[article] Surviving AI Price Wars Without Destroying Your Business. A whole lot of meaningful advice in here for those looking to price AI products for buyers.

[blog] Subagents have arrived in Gemini CLI. Great feature that lets you isolate work and build out agentic teams that can tackle work independently.

[article] Why great teams can’t be copied. Love this perspective. We mistake the outputs of trust with the inputs. If you don’t build the right foundation, all those “successful” practices from prior teams fall flat.

[blog] AWS Interconnect is now generally available, with a new option to simplify last-mile connectivity. Good deal for AWS customers, especially those using Google Cloud and looking for a private connection. Our own docs are here.

[article] Should You Develop Your Leadership Strengths—or Fix Your Weaknesses? You have a strategy here, whether you’ve recognized it or not. Do you spend most of your time investing in your strengths, or bolstering your weak areas?

[blog] Set up Claude Code with Vertex AI. Where are you getting your Anthropic models from? If you care about performance, it’s probably from us. Here’s how to point Claude Code at Anthropic models running on Google Cloud.

[blog] The Gemini app is now on Mac. Installed it. Looks great, and an example of us working differently to ship quickly.

[blog] The RedMonk Programming Language Rankings: January 2026. Working with publicly available data—which admittedly won’t tell you everything—the RedMonk crew checks in with their regular language rankings.

Want to get this update sent to you every day? Subscribe to my RSS feed or subscribe via email below:

A (mostly) fun and frantic day. But the opposite is no good, so I’ll take it.

[blog] Agentic Engine Optimization (AEO). The guide you didn’t know you needed. Addy provides a checklist and more for helping you structure your content in a way that agents can benefit from.

[blog] Deep Dive into Google Cloud Pub/Sub Single Message Transforms and AI Inference. I looked at this feature yesterday, but Romin did a better job of coverage last week.

[blog] Commercial vs Internal Products. Good reminder of the different audiences and urgency for both types of products.

[article] 43% of AI-generated code changes need debugging in production, survey finds. Maybe doing 100% agent-generated code will work out for people. Or, it creates a ticking time bomb. Only time will tell.

[blog] Productive procrastination. Our brains like novelty and try to protect us from negative emotions. How do we recognize this truth, while still doing the work we need to do? Solid advice here.

[blog] Gemini Robotics ER-1.6 enhances reasoning to help robots navigate real-world tasks. The robots are getting smarter with this latest (and safest) robotics model.

[blog] A Simple Coding Agent in a Loop with LangChain4j, Jbang, and Gemini. Neat example. These types of agents aren’t THAT hard to wire up if you want to do it yourself.

[article] Hands-on with the Google Agent Development Kit. A fair review of this agent framework. I’m getting more bullish on it every month.

[blog] The “Dots” Just Got a Lot Easier to Connect: BigQuery Graph is Here! It’s the week before our flagship conference, and we’re still shipping stuff. This is a good one. Broader announcement post here.

Want to get this update sent to you every day? Subscribe to my RSS feed or subscribe via email below: