Daily Reading List – June 17, 2025 (#570)
Long list today. There’s a lot going on! Let me know what sort of content resonates with you the most.
[blog] We’re expanding our Gemini 2.5 family of models. Our Gemini 2.5 Flash and Pro models are now generally available, and there’s a new 2.5 Flash-Lite that’s fast and efficient.
[blog] Modern (Go) application design. Here’s a good look at structures and how they help us build maintainable processes and systems.
[blog] The “Trust, But Verify” Pattern For AI-Assisted Engineering. Skipping AI assistance entirely is dumb. Implicitly trusting everything AI gives you is also dumb. Addy offers a helpful framing that should help us treat AI the right way.
[blog] Expanding ADK AI agent capabilities with tools. Agents need tools to do most anything interesting. Guillaume looks at the various types of tools you can use in the Agent Development Kit.
[blog] MCP and Agentic AI on Google Cloud Run. Your agents and MCP servers need to run somewhere. Services like Cloud Run offer a lightweight, flexible option.
[blog] Build and Deploy a Remote MCP Server to Google Cloud Run in Under 10 Minutes. Related to the above, and a good walkthrough by Jack.
[article] How Kraft Heinz measures AI project value. What does success look like? I like learning from AI users and what works for them.
[blog] Beyond the Buzz: Practical AI with Cloud Databases using MCP and ADK. One more on agents (I think). Billy does a good job explaining why you’d want to use an MCP server that connects you to various databases.
[article] The cracks in the OpenAI-Microsoft relationship are reportedly widening. Yikes. I’m glad we have a more integrated stack that welcomes partners, but doesn’t depend on them.
[article] Do reasoning AI models really ‘think’ or not? Apple research sparks lively debate, response. The AI geeks are fighting! Apple published a paper which some agreed with, and others didn’t. I also liked how Simon thinks about it.
[article] The Anti-Metrics Era of Developer Productivity. Good post. You’ll never find one suitable “productivity metric” for tech teams, but maybe you’re looking for the wrong thing.
[article] Why most developers hate their tech stack. Here’s a metric for you. How many of your folks like or dislike your tech stack? How does that prevent them from doing their best work?
Want to get this update sent to you every day? Subscribe to my RSS feed or subscribe via email below: