Category: Daily Reading List

I’ve been working for days (unsuccessfully) on a piece of a demo I’m doing next week in India. Fortunately our product and engineering teams jumped in, and we found some service bugs. So it wasn’t ENTIRELY my fault. But today I got it working, and it’ll be the highlight of my day.

[blog] Survey Says: Agentic AI Penetrates the Enterprise, but Some ROI Questions Remain. I look forward to these reports from Battery Ventures. They’re more useful to me than the giant dev surveys.

[blog] Go 1.26 Quietly Fixed the Things That Were Actually Annoying. Well good. I’m glad our team does the work that makes your code better, without you needing to do much of anything.

[blog] This Is How To Overcome Burnout: 5 Secrets From Philosophy. Some valid things to try if you feel stuck on the hamster wheel.

[blog] Write code, not specs. Here’s a contrarian take. Instead of maintaining two “systems” (specs and the realization of that spec in code), this person sees the code as the source of truth for patterns, requirements, and such.

[article] On measuring engineers as individuals. When you record and measure individuals based on output metrics, a bunch of the background collaboration work disappears as a result.

[blog] Architecting Stateful Agents with the Gemini Interactions API. You’re managing client-side state for your back-and-forth with an LLM or agent? There’s a better way.

[blog] Is the PRD dead? How to choose the right level of documentation. If traditional requirements docs aren’t the answer, what is? It’s worth talking about this with your team. If you’re still doing it the old way, you might not be seeing the velocity improvements you hoped for.

[article] Andrej Karpathy, Google and Garry Tan agree Markdown is the answer, but they’re not solving the same problem. We’re doing a lot of reading and writing from Markdown. The moat is shifting!

[paper] MCP Server Architecture Patterns for LLM-Integrated Applications. This paper calls out five patterns (such as MCP as resource gateway or stateful session store) as well as anti-patterns and cross-cutting concerns.

[article] Want Workers to Reskill? Show Them Who They Can Become. Ah, that’s harder than just saying “go learn these new things.” But better.

[blog] Expanding Managed Agents in Gemini API: background tasks, remote MCP and more. Good improvements for this fully-managed server-side agent service.

[article] IT unemployment dips below 3% for the first time this year. Maybe the new jobs aren’t in the companies or departments that we assume they are?

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

I enjoyed the time off, and am also happy to be back into routines. The tech world didn’t take much of a break, and I’m working through a reading backlog.

[blog] Most rewrites serve the engineer, not the business. Spicy headline, reasonable take. And using AI doesn’t help you understand the “why” of the decisions embedded within the existing system.

[blog] Where does Antigravity look for Agent Skills? We’ve made this confusing and I look forward to the point when agent skills can be easily discovered and stashed for later use. Mete does a good job here steering us to to the right places.

[article] 5 Traits That Set the Best Leaders Apart. Terrific advice. Some of these traits are easier to adopt than others.

[blog] How to Count Gemini Tokens Locally. As we all become more token conscious, it’s not a bad idea to understand the ins-and-outs of counting tokens.

[article] New Alibaba AI framework skips loading every tool, cutting agent token use 99%. We’re going to need to see more of this in all the AI frameworks. Instead of upfront loading of every tool (and all that context), we should have a lazy loading process.

[blog] What’s New in A2A: v1.0, a Python DX Glow-Up, and a Fresh New Look. A2A has staying power. You’re seeing it supported across programming languages and commercial products. This was a pretty major 1.0 release.

[blog] The State of AI in the SDLC: A Roadmap for Scaling. I buy it. As you chase your first customer, you’re just focused on the code (and use case). Then you have to think about process, and after that, organization.

[article] 10 moments that defined AI’s turbulent first half of 2026. Nobody would have guessed all of these back at the start of the year.

[blog] Of Skills and Loops with AI Assistance. If you invest up front in encoding your knowledge into skills and crafting agent loops, you’ll see some fairly dramatic speedups.

[blog] Building Gin: Simple Over Easy. Sometimes the throwaway thing you built along the way turns into the main thing. Great story about the start of Go’s most popular web framework.

[blog] Tom’s opinionated guide to skill building 101. Excellent. Skills don’t just serve software developers. They can make other disciplines, like technical writing, significantly better.

[article] The twilight of the chatbots. Looping agents with good direction don’t require a lot of human intervention. That’s different than early mainstream AI days where we engaged with chatbot interfaces.

[blog] Anomaly detection using dynamic thresholds and two-year-long alerts in Cloud Monitoring. Sometimes static thresholds or short lookback windows don’t give you the right signals. We’ve now made it easier to create alert policies over wide sets of historical metrics.

[article] The Pulse: a new trend, smart model routing. Yes, take advantage of tools that can pick the right model for the task. This can save you a ton of money.

[article] The website of the future may assemble itself for every visitor. I suspect that this generative UI pattern is going to show up in a LOT of web and mobile experiences in short order.

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

I had a great day off. The kids had a summer camp for the morning, so I took a joyful, meandering drive through the mountains and towards the beach. One beach-adjacent taco stand later, I wandered back. This is why I’ll never want a self-driving car.

[article] 7 reasons experienced EMs get stuck. This may be for engineering managers, but these reasons apply to nearly any manager position.

[blog] A return to two-pizza culture. Amazon’s Werner Vogels looks at the shift in when you write down your plan, and how teams work with AI.

[blog] Learning faster with Antigravity. How do you build a frontend for a backend you’re unfamiliar with? Andrew did a learning loop, and created a skill to make this process repeatable.

[article] Z.ai launches ZCode to challenge Cursor, Claude Code and GitHub Copilot in AI coding. The hottest AI coding model just got a coding UX. Smart fast-following offering.

[blog] Bringing speed and strong cost performance to the market with Gemini Omni Flash and Nano Banana 2 Lite. Speaking of cool models, these two are remarkable for those doing media creation.

[blog] Guide to the OWASP MCP Top 10. Here are the security issues you need to watch out for with MCP. Some of these you can directly mitigate, and others might require some proxies or gateways to protect your users.

[blog] AlloyDB AI Functions – now with revolutionary performance boosts and cost savings. We’ve made it cheaper and faster to make LLM calls directly from your SQL statements.

[blog] 🤖 Ten do’s and don’ts for LinkedIn posts, according to actual research, based on pre-slop posts. Solid AI-summary from Coté here. Do these things to perform better on LinkedIn and stop hating yourself.

[blog] Why we built ADK 2.0. A big part of this was bringing deterministic execution to agents with “workflows.” Good post on when to use this new abstraction.

[article] When Developing an AI Strategy, Beware the Urgency Trap. Purpose matters more than anything else. Get that right! Rushing headlong in the wrong direction rarely pays off.

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

We’ve somehow got both tomorrow and Friday off for our Independence Day holiday. I’ll still probably do a reading list both days. I’m sure the anticipation will keep you up tonight.

[article] “Code should be regenerated, not maintained”: Codeplain makes the case for spec-driven development. I’d imagine that for most people, this approach breaks a lot of things. Which might not be a bad thing.

[article] Five tools to bolster your AI coding stack. Good list. Covers things like testing environments, validation tools, and security services.

[blog] Get started with the Claude apps gateway for Google Cloud. If you care about performance, you probably already point to Anthropic models hosted on Google Cloud. Now you can run their new apps gateway here too and get enterprise management.

[article] How top PMs increase their leverage with AI. This has a framework you might apply to a few different disciplines, not just PM.

[blog] Driving the Agent Quality Flywheel from Your Coding Agent. Build better agents by running better evals. Here’s a skill you can install that gets you further along.

[blog] GenPage: Towards End-to-End Generative Homepage Construction at Netflix. Cool details from Netflix on their generative UIs, with fine-tuned models underneath.

[article] .NET’s long-term support is not long-term enough, dev complains. I’m sure it’s tough to keep up with nonstop language updates. I’d be surprised if the answer will be “support for longer” or “slow down.” Related.

[blog] AI shouldn’t shrink headcount. It should shrink teams. There are downstream implications of the “smaller, faster teams” push. Doesn’t mean you shouldn’t do it, but be prepared.

[blog] Gemini Spark updates: macOS launch, connected apps and more. Personal agents are here for everyone. It’ll take time to figure out what to do with them!

[blog] Hugging Face and Cerebras bring Gemma 4 to real-time voice AI. Cool demo from the team at Hugging Face.

[blog] Faster Gemma 4 on MLX with multi-token prediction. Another Gemma-related item from the community, this time from Ollama.

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

There’s a lot today about the strategy around using and rolling out AI at scale.

[blog] Mastering Gemini Computer Use : A Comprehensive Hands-on Guide. Instead of creating brittle scripts that try and parse UI elements, what if you worked with a model that could “see”? That’s computer use, and what Romin explains here.

[article] The future of engineering at Nationwide, Comcast, TD, and HPE. These are all wonderful takeaways. I’d be surprised if any large enterprise—I haven’t met one—has truly done any of this at scale, but it’s the right foundational beliefs.

[article] Markdown Is Becoming the Application Layer of AI Apps. Fascinating take. Maybe not wrong. But it’s a shift for people to put even more domain behavior into prose.

[blog] Beyond Gemini Enterprise: Google’s Hidden AI Superpower. A great model is a must-have. We’ll keep making those. But it’s about more than just an LLM, and Google’s got a unique value proposition up and down the stack.

[article] Agent Memory. Excellent deep dive into the various types of agent memory and considerations for each.

[blog] Cloud CISO Perspectives: How Google Cloud Security uses AI internally. The use of a self-reflective loop seems important here.

[article] The AI jobs debate just got messier. Sometimes the data refuses to bend to the narrative. Or, we just assume nobody knows what’s going on right now. I’m voting for the latter.

[article] Claude Code turned every engineer into three. Now companies need more product thinkers. Yes, knowing syntax or obscure domain knowledge isn’t the most valuable thing today. Instead, can you understand higher-order problems and own the journey towards the goal?

[blog] Build reliable multi-agent applications with ADK Go 2.0. Discover our new graph-based workflow engine, built-in human-in-the-loop, and dynamic orchestration. If you’re using Go for AI workloads (great choice!), this is the best way to build AI agents.

[article] Your Talent Strategy Has to Keep Up with Your AI Transformation. Be smart about how you approach a plan for new talent, and how you invest in your current staff.

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

I liked some of the manager-focused pieces today, as that’s an under-served audience right now. There’s plenty of content for how individuals use AI, but much less about the impact and role of managers.

[blog] Ask an AI expert: What exactly is the full stack? The central Google team asked me some questions about full-stack AI. Here’s the result.

[article] Uber’s journey of measuring AI impact on developer productivity. I didn’t listen to the episode, but the takeaways published here are solid.

[blog] Five essential testing patterns for AI agent development. Great post from Karl. It’s easy to get AI to generate shallow tests for your codebase. Are you asking for and seeking a more comprehensive testing strategy?

[article] Where is the flattening of engineering organizations actually happening? Some of these results may surprise you. Reorganization is much more common than layoffs, and the top is flattening faster than other layers.

[article] AI Adoption Is Overloading Your Middle Managers. We need *more* investment in middle management because of the AI transition?

[blog] Paging Charity! How can engineering leaders avoid becoming Bond villains? The number one job of every manager/leader is to win at business. Not build great teams, not coach the next generation. Those things matter a ton. But you won’t get the opportunity to do those things if the team doesn’t matter.

[article] AI Tools Accelerates Coding, but Not Overall Software Delivery, GitLab Research Finds. When I see headlines like this, it’s an indicator that these teams haven’t yet done the required changes to teams and workflows.

[blog] Locally test your Cloud Run service code that writes to Google Sheets. You can’t test everything locally when dealing with cloud services, but many things you can.

[article] CEOs, CIOs clash over AI’s value. Oooh, enterprise drama. Each side probably expect different outcomes, or at least on different timelines.

[blog] Antigravtiy CLI Plugin for Claude Code. Mix models, mix harnesses, whatever. You don’t need to use one stack for everything, and it may be prohibitive to do so!

[blog] The Problem is Prompt Debt. Less prose, more firm direction. And fewer hand-written prompts, more generated.

[article] The AI era requires a different kind of experimentation. Food for thought. Have you thrown our your own playbook and assumptions? Might be time.

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

Most of my “building” nowadays is for fun, for blogs, or for customer demos. But today I had to take a spreadsheet of data and enter records one-at-a-time into a web app. No thanks. Broke out the Google Antigravity CLI and took a few minutes to build a Playwright script that looped through the rows and filled out the web form. Achievement unlocked.

[blog] Learn Anything With My /teach Skill. Neat skill from Matt. He talks about it for Claude Code, but there’s nothing specific to that harness. Run this in your favorite environment to get a personalized teacher.

[blog] Writing Loops, Not Prompts, Explained. Skip the “write loops not prompts” hysteria. There’s a place for automating recurring work. But prompts are still part of the equation.

[paper] Developer Productivity in the Age of Generative AI: A Psychological Perspective. My colleague wrote up his Master’s Thesis on an interesting topic and it’s available here to read.

[blog] Hidden Technical Debt of AI Systems: Agent Harness. What an insightful read. I hadn’t thought about production versus training harnesses. The idea of the harness melting into the model is something we’ve had as well.

[blog] Moving Past “Demos Over Memos”: The Future of AI in Enterprise Product Development. Fair points from Jason, and I should change how I refer to this idea. We need MORE than just vibe coded demos to explain ideas, understand market fit, and build alignment.

[blog] How to apply professional design principles in AI app development. Really good. It’s easy (and I’m speaking to myself) to just accept the stylized UI that comes back from the AI tool. But this post points out the eight areas where you can push back to get a stand-out result.

[blog] Make AI Boring Again. Smart take, as always. Learn these tools, figure out the best use, and make it boring.

[blog] The Coming Divide: AI-Native or Left Behind. I get it. This one is a choice. You can choose to use AI or not. Just like you can choose to read books or not. But then own the outcome.

[blog] Optimizing cloud economics with linear elastic caching. Good thinking, useful experiment, and legitimate results.

[article] The Two-Organizations Problem. There’s the presented view of the org (via dashboards, townhalls, etc) and the lived org (how we experience it day to day). Those are not always the same.

[blog] Building in the Age of Collaborative Coding. I haven’t seen a ton written about how mature teams work together with AI agents and tools.

[blog] Demystifying A2UI: How to Make AI Agents “Speak UI” in Your App. I’m seeing more where generative UIs—giving LLMs the ability to lay out a screen based on provided components and data—make sense.

[blog] Conference talks in an AI driven world are likely to get boring, or need to dramatically change. Let’s watch presenters wait for prompt responses! That sounds super fun. Not really. We’ve got to rethink tech events to ensure they don’t turn crazy boring.

[blog] Here’s how Gemini can help you avoid jetlag. I was very skeptical of this claim. So I tried it on my upcoming trip to India. Wasn’t half bad.

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

How well my favorite sports team plays should have no bearing on my happiness. Alas, that’s not the case. But, my Padres won three in a row, which partially explains my good mood this week.

[blog] Why Are Agents Buying the Same Answer Twice? That’s a darn good question. Devin wonders why we’re computing stable answers over and over again, versus storing and recalling them.

[blog] Who Wrote That Query? You need the right visibility if you’re going to troubleshoot effectively. I like that our MCP Toolbox for Databases now makes it possible to trace a SQL query back to the agent prompt.

[site] ThoughtWorks Radar – April 2026. I don’t know how I missed this back in April. But scan through what technologies are ready for adoption and trial.

[blog] Resilience Patterns Can Make Your System Less Resilient. Can you make your system worse by applying retries, circuit breakers, and fallbacks? Yes, yes you can.

[blog] Brownfield modernization with Antigravity. New stuff is fun, but we’re all going to spend most of our AI coding effort on existing systems.

[blog] Configuration belongs in a database. Maybe a hot take? But the argument here holds water with me.

[article] AI coding token costs are on track to rival human payroll. I’m not sure it’s wise to try and project current behavior too far out. We’re still figuring out these tools, models, and pricing schemes. Related.

[article] Stop Getting Good at Protocols. Get Good at Agent Experience. Be good at both? Even after reading this, I’m not sure you can be great at AX without understanding the ways to make it happen. But I get the point to not obsess over fast-changing implementations.

[article] Demand for AI-ready coders skyrockets in 5 years. So the roles are changing, not disappearing? I think everyone not invested in the doomer outcome expected this to be true. Related.

[blog] Ping Pong Pairing with Antigravity. Aja applies a proven technique but with AI agents instead of humans.

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

We’re in process of doing our quarterly “check-ins” towards our personal goals. It’s a great moment for me to reflect and appreciate what my team has accomplished already!

[blog] How To Measure Development Productivity? Should you measure activity? Or output? Those are flawed categories, although they can have value at the team level. This post reminds us to consider desired outcomes and measure that.

[blog] Introducing computer use in Gemini 3.5 Flash. When you need/want agents that can “see” and reason across browsers and desktop, this sort of capability will be intensely valuable. Great to see it natively integrated into the Gemini Flash model.

[article] The Mom-and-Pop SaaS era has arrived. How many great ideas never got built because the cost of realizing them with software was too high? Not anymore.

[article] Lost confidence. When you can’t honestly quantify your confidence about the outcome, how do we proceed? Some really useful techniques called out here.

[article] Europe’s cloud sovereignty push may backfire. Related to the above point, I can’t predict how this goes. Doesn’t seem like it’ll go well.

[blog] Build Cross-Language Multi-Agent Team with Google’s Agent Development Kit and A2A. Cool demo, including downloadable source code. How do you make agents built in different languages work together? Here’s how.

[article] The AI Productivity Bill Comes Due in Production. The bottleneck probably moved downstream, and the work itself changes shape to be more about supervision.

[article] Kubernetes in the Age of AI. It’s a runtime for model inference, a host for agents, and probably more.

[blog] Is it thinking or is it broken? Building transparent AI chat UIs with Genkit. Is this thing on? You might ask that when waiting to see if the AI service is stuck, thinking, or something else. Here’s a pattern for showing a heartbeat.

[blog] Skills Over MCP. Can we ship the manual with the product? That’s what this working group is trying to figure out.

[article] OpenClaw and Hermes agree on what an agent is. They disagree on what controls it. Interesting. Jani helps us understand the philosophical difference between the two personal agent platforms.

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

I’ve had a new blog post bouncing around in my head for a couple of weeks, and I built out some of the necessary proof-of-concept last night. Now, to block time to write the thing. That’s really the most important part of blogging, the writing.

[blog] Old Software Was Fast Because It Had No Choice. Is modern infrastructure too forgiving? Do we allocate extra headroom because we can? Instead of throwing capacity at every problem, maybe we can design smarter.

[article] Help Employees Get Better—Not Just Faster—with AI. This may change how I talk about this topic. Don’t make AI adoption just about creating AI fluency; it’s a change in how you express experience and judgement.

[blog] How to Become Ridiculously Good at Kubernetes. I like the point that average engineers know the commands while expert engineers know the failure modes.

[blog] Interactions API: our primary interface for Gemini models and agents. It’s GA, and the best way to interact with Gemini models and agents. Get built-in support for background work, state management, and more.

[blog] Don’t rely on instructions, use Agent Hooks to enforce guardrails. Straightforward point. An agent can ignore instructions, but the hook runs every time.

[blog] Securing the future of AI agents. This is from Google DeepMind, and links to a pretty great paper on how to think about agent security.

[blog] In-place pod restarts: Boosting efficiency and workload reliability in Kubernetes v1.35. This is a fairly important feature for those who don’t want massive control plane thrashing when pods restart.

[blog] How Netflix Simplified Batch Compute with Kueue. Today’s got a lot of Kubernetes content. I’m not sure why. But, it’s a great reminder of how important your underlying infrastructure really is.

[blog] How we built a Flutter-powered AI coffee shop. I find that “fun” examples of technology are often more educational than the professional ones.

[article] AI isn’t solving cybersecurity workforce woes. Let’s see. A majority of cybersecurity people feel like calling it quits, AND we don’t have enough people to even fill the open roles right now. This is fine. Everything’s fine.

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