Category: Daily Reading List

On my first normal workday in a week-and-a-half, it was mostly all meetings. Lots of catch up. But I got to spend the last bit of the day working on a presentation about “how to use AI without losing your soul.” You can tune into it on Wednesday if you’d like!

[blog] Earning taste and judgment. Worried that recent college grads are either unemployed or underemployed? You should be. Addy looks at what humans are uniquely good at (taste/judgement) and how early-career people that build it up.

[blog] Evals belong in your CI/CD pipeline. I’m a sucker for opinionated statements like this. Gets my attention. This is reasonable advice. Although, my evals haven’t been super fast, so I wonder if you’d run them on every build.

[blog] ADK Agent Identity on Google Cloud: Separate Identities, Separate Blast Radius. This topic is probably “stage two” after you’ve already built an agent. But understanding agent identity is critical if you ever hope to put an agent in production.

[blog] Turn User Signal Into What You Build Next. It might not just be about mixing up the SDLC stages. Rather, it’s putting “user feedback” higher on the priority list and putting prototypes/variants in the hands of users faster.

[blog] The Risk of Exposed Cloud Functions and How to Harden. The serverless hype train has reached the station, but that doesn’t mean many of us don’t still build scale-to-zero apps and functions. These are real concerns to mitigate.

[blog] “Loop Engineering” Is Mostly Just Broken SDLC Wearing a Costume. Adron bringing out the big guns. Instead of using loops (error correction) to stumble through a process, craft workflows with clear signals.

[blog] Three Agents Behind One Door: Multi-Tenancy in A2A. Multi-tenancy is new in A2A and makes it possible to have multiple agents behind one Agent2Agent server.

[article] DoorDash Uses Envoy and Valkey for a 1.5M RPS Proxy Cache with 99.99999% Availability. Cool mix of caching, invalidation, failure handling, request coordination, and performance optimizations in one system.

[blog] Making highly available, multi-region Cloud Run services just got easier. Run the same web app across regions with automatic failover? Not trivial. But we made it easier.

[blog] Building scalable AI agents with modular prompt transpilation. Are you thinking about prompts like software artifacts? This post encourages a more rigorous approach to decomposition and construction.

[article] AI OK in Linux development, says Torvalds. If you don’t like AI being used in Linux, fork it. More here.

[blog] Designing APIs for Agents. Don’t just cater to humans and what we can comprehend. This piece encourages us to think differently for our AI consumers.

[blog] Google is a Leader and positioned furthest in Vision and highest in Execution in the 2026 Gartner® Magic Quadrant™ for Conversational AI Platforms. Fantastic showing, and a very valuable technology category to start using at your own company.

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

Just got home after a 24+ hour travel day back from India. No wifi on the second (11 hour) segment, so I’ll have some catching up to do! Great trip, though.

[blog] What 10 autonomous film crews taught us about agent teamwork. This might be the wildest thing you read all week. What happens when you create film crews out of AI agents, and then assign an AI documentary crew to capture the journey? You get this.

[article] How I Cut an AI Agent’s Token Use by 94%. Super smart. Over time, you might realize that your natural language agent skill can be “compiled” into a deterministic script.

[blog] My thoughts on the future of Go in the AI era. So very well laid out, and a major reason that Go may become the preferred language for this AI era.

[blog] How to manage AI investments in the agentic era. From OpenAI. It looks at how to identify (and quantify?) value from your AI investment.

[blog] NotebookLM is now Gemini Notebook. We like renaming things. This product keeps getting better and more integrated, so I get the reasoning.

[blog] Kimi K3: Open Frontier Intelligence. Yeesh, you could make the case that this is now the best coding model, and it’s an open one. Running this yourself is fairly impractical (for now), but I imagine the API will be super popular.

[blog] Four math concepts to improve your vibe coding. I was told there’d be no math today. But Karl pulled me in with this post.

[blog] Model Routing Is Simple. Until It Isn’t. Some interesting perspective up on the Hugging Face blog. You’ll increasingly crave a model router. But there’s more to it than you think.

[blog] Evolving Spec-Driven Development: Conductor Now Supports Antigravity. This spec-driven extension for the Gemini CLI now works as a plugin for Antigravity and agentic harnesses.

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

It’s my final day in India, as I fly home in a few hours (3am local time). Great trip. I’ve got a pile of things to think about, and another handful of things to go build.

[article] The Prototype Is a Question, Not a Product. Just because we can do lightning-fast prototypes now, it doesn’t mean we shouldn’t have a strong hypothesis and scorecard before we start. Great reminder here.

[blog] The (no longer) missing multi-agent pattern: triggering dynamic workflows from an agent. Very cool example of a runtime-determined orchestration that responds to dynamic requirements.

[article] 5 Trends That Defined AI Engineering at World’s Fair 2026. Great lineup at this event. These trends represent the current best thinking.

[article] How I use HTMX with Go. Long tutorial, but something I hadn’t tried before. And I was shocked to see all the positive vibes in the Hacker News conversation about this pairing.

[blog] Claude at scale on Google Cloud: Frontier AI, built for enterprise production. We’re not just serving a great model here; you’re automatically getting our built-in scale, security, and cost-management features.

[article] Microsoft is reportedly training salespeople to talk down OpenAI and Anthropic. We all take different approaches. This is certainly one. Although the claim about selling the only “end to end system” is … something.

[blog] Which Doc Format is Best for AI Specifications? I had to regroup after reading the first sentence—that’s a LOT of specs. It seems we’re still figuring out which document formats represent the best way to transmit and store agent-friendly data.

[blog] What does “playing politics” mean for software engineers? Its naive to believe “I’ll just do great work and everything will work out.” Do great work. But also have self-awareness about the system you’re operating in.

[article] The Frontend Verification Gap in AI-Assisted Development. Your favorite AI tool probably creates some very nice looking frontend UIs. What does “correctness” look like though, and how do you test the frontend’s behavior?

[blog] Let Your Agent Run Its Own Code Inside a Cloud Run Sandbox. I’m very bullish on this, and have already seen some pretty wild reference apps within Google.

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

Our Google Developer Expert community is pretty remarkable. 8% of them are in India, and I got to spend all day today with them. And, I also snuck in some experimentation with agent skills.

[article] The AI Didn’t Fail. The Deployment Did. That AI-powered café almost went bankrupt. Was it because the model went crazy? No. Casey shows that the AI didn’t have the context, guardrails, and memory that a competent builder would include.

[blog] Building Service Topology at Scale: Architecture, Challenges, and Lessons Learned. Do you have an accurate topology of the dependencies between your systems? Few do. Here’s how Netflix built a streaming system to try and solve it.

[article] DeepMind CEO calls for an independent standards body to regulate frontier AI. Important proposal that seems to have universal support among AI leaders.

[blog] Control the ideas, not the code. You’ve got a limited number of usable hours in the day. Salvatore has a post worth reading and contemplating.

[blog] You Just Hired a Million Bad Employees. There are like six hot takes in this post. Are humans now cheaper than software? This seems to assume that token costs will remain flat. But there’s an important point here about wasted cycles.

[article] Google rolls out sovereign AI stack for India’s regulator sectors. I didn’t completely embarrass myself in this interview.

[blog] “We Are 90% Done” Is the Most Expensive Sentence in Software Delivery. That last 10% seems to take most of the time. A lot of work hides in there, as called out in this post.

[article] The real AI race may no longer be at the frontier. We’ll see. Open models are rightfully having a big moment right now. That won’t go away.

[blog] Systems Engineering Playbook: Optimizing Qwen 3.5-397B MoE on Ironwood (TPU7x). Speaking of open models, here’s how we optimized one through some creative techniques.

[article] Google’s Genkit Ships Agents API with Detached Turns and Human-in-the-Loop for TypeScript and Go. Great InfoQ writeup about this unique SDK for adding agent capabilities to your full-stack apps.

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

Today was a whirlwind. What a great crowd at Google I/O Connect in Bengaluru. After keynotes and breakouts, I got to spend my afternoon with media, analysts, startup founders, podcasters, and influencer-types. I’m officially talked out.

[paper] Always-OnAgents:A Survey of Persistent Memory, State, and Governance in LLMAgents. We’re going to see more always-on agents instead of primarily doing one-and-done sessions. But what does that mean to all the persisted state and the risks that come with that? Interesting analysis.

[blog] On Arguing Against Jevons. Why do we keep “flinching” when faced with Jevon’s Paradox that efficiency and lower cost of a thing results in more consumption of that thing?

[blog] Amazon SQS turns 20: Two decades of reliable messaging at scale. This was one of the first cloud services I ever used. Coming from a messaging background, I was blown away by the simplicity of having queues in the cloud.

[article] Five studies that are changing how I think about AI in software engineering. Each of these papers takes on a different question, and this article summarizes the key points from each. But most data shows that we’re building code faster than we can verify it or deliver it!

[article] Enterprise AI is entering an evaluation gap: Agents are gaining autonomy faster than companies can verify them. Echoing my previous point! There’s not a single solution.

[article] Engineering teams will shrink as AI shifts responsibilities. You’ll likely see more, smaller teams replacing the fewer, bigger ones. Give high-leverage tools to an empowered team of 3-8 and you’re set.

[blog] Securing the AI supply chain on GKE: Introducing k8s-aibom for automated AI BOMs. It’s open source, and looks at your deployed AI serving infrastructure to produce live bills-of-material.

[article] “Every IC engineer is really a front line manager now.” But are they productive? Indeed. This is a new type of work and new set of skills for every individual contributor. I’m not sure either are being well-addressed by management.

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

I’m getting ready for tomorrow’s big Google I/O Connect event here in India. We spent today doing rehearsals, meeting with startup founders, and community people. And, reading the handful of items you’ll find below.

[article] Stop being the code review bottleneck. Great advice here, including sample prompts you can steal for yourself. Using agents as pre-reviewers is a good pattern. Involve humans where necessary.

[article] 85% say code review is the new bottleneck. Here’s what the AI coding narrative leaves out. Another item about code reviews. It also reiterates why pre-merge AI reviews are part of the solution.

[blog] Why write code in 2026. Good argument here. You write code so that you remain an active participant in the software. And it helps you think through a problem.

[blog] I let Google Antigravity 2.0 write a microservice from scratch while I ate lunch, and I’m mildly terrified. It feels weird when you don’t have to babysit a looping agent. Especially when you notice decent product thinking and UX sense.

[blog] Why TypeScript 7.0 Was Rewritten in Go (and what it means for your dev stack). What a tremendous piece by Steve. It’s time for everyone to take a look at Go, and realize why it’s purpose-built for the era we’re in.

[blog] Microsoft joins Google in backing Go for AI agents — OpenAI and Anthropic lag. See, Microsoft gets it too. Their agent framework just introduced Go support.

[article] Agentic AI strains legacy IT systems. Maybe rewrite them in Go 🙂 New demands require a new approach to infrastructure.

[article] 57% of enterprises have watched AI agents be confidently wrong. The fix is an agentic context layer, but who has one? Most don’t have a trusted context layer. Those companies will likely struggle with trustworthy results and production-scale deployments until one is in place.

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

I can’t remember a smoother long-distance travel day. Flights all on time, decent wifi, got a little sleep, and went from the airplane landing in India to my hotel in less than an hour. Personal record!

[article] Why AI Coding Agents Still Need Clear Specs. Create fully complete specs before firing up your Ai agent, or just /goal and let your agent figure it out? This talks about the importance of validating the spec (even a small one) before setting the coding agent loose.

[blog] Solve harder problems with AlphaEvolve, now available to everyone on Google Cloud. This is an impressive system that we’ve used ourselves to design our TPUs, improve Cloud Spanner storage, and improve scientific research.

[blog] Multi-Tenancy Isn’t About Databases. Good reminder that multi-tenancy is about isolation, and you don’t necessarily have to use separate databases per tenant.

[blog] The Antigravity Loop: From Chat to Autonomous Engineering. Romin looks at the spectrum of autonomy for agent modes. I like the examples of when you use each.

[article] Own the Outer Loop. Engineers need to own accountability for their systems, regardless of how much AI owns the agent inner loop.

[blog] The Future is Stateless: MCP 2026 Draft Spec Arrives in MCP Toolbox. Assume that we’ll keep seeing some of the agentic protocols evolve for performance and interop reasons.

[blog] From Dirt to Tokens: How AI is Collapsing the Walls Between Apps and Infrastructure. Instead of building apps on other people’s infrastructure, many companies are looking to control costs by managing their own AI infrastructure.

[blog] LiteRT.js, Google’s high performance Web AI Inference. Check this out. AI client side in the browser. Very impressive results.

[article] AI skills now listed in 73% of tech job postings. Hiring managers need people with the right skills. Right now, that’s AI skills.

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

Off to India this evening, and looking forward to a fun event in Bengaluru. Will Richard have airplane wifi? Nobody knows.

[article] Research: AI Is Changing What Employers Want from New Hires. Yes, this is what I want too: broader skills, abilities to synthesize information, and willingness to improve workflows.

[article] Agentic AI to disrupt $234B in SaaS spending: Gartner. I’d imagine we’ll still need many of the system-of-record SaaS products, but probably not all the UIs and add-ons. Just give me the core APIs.

[blog] Have you heard? Clickhouse is winning the observability wars! Sometimes those blazing the trail don’t get the glory; it’s those that follow. But Charity points out why a different approach to storage (and a different philosophy) for observability is distinctive.

[blog] Separating signal from noise in coding evaluations. A popular coding benchmark for AI may not be a great way to assess coding capabilities.

[article] Popular open source AI developer tool Ollama raises $65M, grows to nearly 9M users. Where you spend tech budgets is changing. Tools like this are wildly popular, for good reason.

[blog] Safely run AI-generated code in Cloud Run sandboxes. Now you can easily isolate untrusted code in a sandbox that’s part of your running service.

[article] JetBrains’ next move isn’t a better IDE — it’s a governance layer over Claude Code, Codex, and Gemini CLI. Same as above. Now you’ll be buying governance and orchestration layers on top of AI coding tools.

[blog] Human-in-the-Loop AI: Why ‘Ask the LLM to Confirm’ Isn’t Enough. This offers a useful perspective on approvals embedded in the prompt (and skippable) or in a deterministic gate (unavoidable). Which do you need?

[article] SpaceXAI launches Grok 4.5, its first built with Cursor’s help. Impressive model, and it’ll see good uptake. More here, and here.

[blog] Introducing Muse Spark 1.1. More model news, this time from Meta. Benchmark numbers look great.

[blog] Report: 83% of organizations need to upgrade their infrastructure to support agentic AI. Who are those other 17%? Especially if companies are going to seriously invest in self-hosted models, it’s critical to reset some core infrastructure.

[article] Do Hard (Meaningful) Things. Yes, don’t just do random hard things. Do hard things that matter.

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

It was a wild day of meetings, finishing some demos for next week, and playing whack-a-mole on my inbox.

[blog] In defense of AI mandates. You might not like ’em, but mandates can put needed muscle behind an important decision.

[blog] Orchestrating with Antigravity: A Crescendo of Agents (Part 2). I’m reading this in Riccardo’s voice, and enjoyed the journey he captured here.

[blog] Configuration as Code is a liability for security. This post says that configuration as data (fully materialized YAML in a database, not templates in git) is the way to go.

[paper] Gemma 4 Technical Report. Read this for some great details on our model architecture, pre-training, evals, and more.

[blog] Keeping AI Agents in Check: Guardrails and Callbacks in ADK. If you were working with a CLI, we’d be calling these agent hooks. Martin shows us how to use these callbacks/hooks to add reliable checks to your AI-led process.

[blog] Announcing TypeScript 7.0. The Go-based port is 10x faster than previous TypeScript. And it’s now ready to use.

[article] Former GitHub CEO’s startup Entire unveils its answer to the crush of AI coding agents. I’ve been sharing here for months that changes were coming to source control management. Related.

[blog] Great Products, Bad Companies. Companies with great products face some unique threats. Marty encourages us to recognize that and react.

[blog] The power of collaboration: How we can reduce traffic congestion. Cool experiment with legit outcomes.

[blog] Turning Data into Context: Google Cloud Storage (GCS) is Now Available in MCP Toolbox. Unstructured data matters too. Now this open source data-oriented MCP server lets you retrieve your object storage data.

[blog] Google Cloud named Leader in the 2026 Gartner® Magic Quadrant™ for AI Infrastructure. This offers a good look at who the legit players are in the competitive market of AI Infrastructure providers.

[article] CEOs fear they’re underinvesting in AI. There’s a lot of FOMO out there. But don’t use AI because of that. Apply it where it can make a real difference.

[blog] We terminated a TPU mid-training and it recovered in seconds: Introduction to elastic training with MaxText. Nervous that one node failure crashes your whole long-running AI/ML training run? Here’s a solution that doesn’t require a heavy outside orchestrator.

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

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: