Daily Reading List – July 9, 2025 (#583)

When your day is typically full of meetings, does a light meeting day throw you off? It does for me, and I have to be very intentional about how I spend time. That was today, and I was able to get a few tasks done ahead of schedule.

[article] Hugging Face just launched a $299 robot that could disrupt the entire robotics industry. Looks great. Pre-order available now, not yet shipping or a ton of details.

[blog] Full-breadth Developers. Google is called out here as an anti-pattern, but I still really liked this post from Justin about bringing together the product (design) and execution skills into the developer role.

[blog] AI-Native Test Automation is Here. There’s been hand-wringing about AI slop resulting in more low-quality apps. But maybe, AI will help us add better testing than we have in the past?

[paper] Deep Research Agents: A Systematic Examination And Roadmap. You’ve got many choices when it comes to doing thorough research using an AI agent. This paper explores the core architectures and approaches of popular options.

[blog] Gen AI Evaluation Service — Computation-Based Metrics. What metrics should you use in your generative AI evals? Mete looks at computation-based metrics which are based on mathematics. He’s also got a great follow-on post that explains model-based metrics.

[blog] AI Tooling, Evolution and The Promiscuity of Modern Developers. Everything is up for grabs right now, as Stephen told me last week. He wrote a great piece today that you should read.

[article] Context Engineering: Going Beyond Prompt Engineering and RAG. If I’m the reason you keep seeing this term, my apologies. But I’m seeing it show up in many places.

[article] Spec-driven Development. Brian goes into depth on a topic I touched on last week. He does a terrific job explaining “agent docs” and doing explicit architecture and task management for the LLM.

[article] Against “Brain Damage.” This piece by Ethan reinforces why I’m gravitating to the spec-driven work that Brian highlighted above. It ensures that we’re still thinking, prioritizing, and steering creative work. And, not using AI as a crutch for everything, but still doing free thinking!

[article] Idle Thoughts On Programming and AI. A lot here, but it also builds on the last two items on the list. Software engineering feels like it’s changed more in the past 5 months than it has in the last 10 years.

[blog] Google Agent Development Kit (ADK): A hands-on tutorial. Terrific content here from Weights & Biases looking at building and evaluating AI agents.

[article] Tech unemployment rate hits lowest yet in 2025: CompTIA. Confusing, eh? With so much doom and gloom about layoffs and a bad tech market, it seems that open roles have simply shifted from big tech to many other places.

[youtube-video] Getting Started with Agent Development Kit Tools (MCP, Google Search, LangChain, etc.). This is a great conversation and I really like how Megan and Jack position the problem space and corresponding solutions.

[blog] Gemini CLI Tutorial Series — Part 4 : Built-in Tools. More tools discussion, this time looking at what’s included in the Gemini CLI. Part 3 looked at config settings.

[article] Why Senior Leaders Should Stop Having So Many One-on-Ones. Wow, this one made me stop and think. Should senior-level 1:1s be only focused on development, and leave comms and decisions only to group settings?

[blog] Unlock the Power of MCP Toolbox in Your Go Applications. Go developers can now use this SDK to access databases as tools within AI apps.

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