Dion Almaer

★ Delight developers in ecosystem-enhancing, easy to copy ways Some thoughts on the role of platforms (especially open ones like the Web) and how to think about long term progress. blog.almaer.com/delight-develo…

This is awesome. Specs provide the layers of context that are needed to help LLMs do what you really want. You don’t want one mega spec, you want composability. Go SpecTree Go!

Curious to see the Textual Toad come to be.

"The future of software development might just be jazz. Everyone improvising. Nobody following the sheet music." worksonmymachine.substack.com/p/nobody-knows…

I've posted a couple of Substack notes recently about my thoughts on AI coding workflows. - The Navigation Principle (substack.com/@wycats/note/c…) - Using READMEs for "Specs" (substack.com/@wycats/note/c…)

🔔 Gemini 2.5 Flash-Lite is now stable and generally available. This model offers the lowest latency and cost in the Gemini 2.5 family, making it our fastest and most cost-efficient option for developers building production-ready AI applications.

“When your goal is to reason with an LLM about implementing or fixing some code, you need to provide extensive information to the LLM” antirez.com/news/154

Announcing Code Sandbox MCP, a simple code interpreter for your AI agents. Code Sandbox MCP is a simple, self-hosted code interpreter that gives AI agents like Gemini the ability to execute code in a local, containerized environment you fully control. How it works: 1. Starts a…

wait, i legit slept on google’s AI suite. gemini writes, flow films, notebooklm researches, stitch designs, project mariner automates. i filmed a 38-minute tutorial with google’s VP of gemini and it just rewired my whole workflow.

Official results are in - Gemini achieved gold-medal level in the International Mathematical Olympiad! 🏆 An advanced version was able to solve 5 out of 6 problems. Incredible progress - huge congrats to @lmthang and the team! deepmind.google/discover/blog/…

I see a world in which any startup can submit their RL environment to a big lab to fine tune a custom agent on. This would unlock a lot of user experiences for users that foundation labs are currently structurally unable to deliver on

The Product Manager’s Guide to Not Slowing Things Down: Velocity isn’t the goal, it’s how you work. kathykorevec.substack.com/p/the-product-…

Think I've settled on JSX with Preact as the templating language for this Tailwind CSS course I'm working on. Lets me use standard HTML attributes (class vs. className, dash-cased SVG attributes) so non-React users will almost not even notice it's React 👍🏻

We're working across products to try and make this better for devs. What are your best practices or examples for AGENTS.md? Let us know!

Today we are rolling out our first Gemini Embedding model, which ranks #1 on the MTEB leaderboard, as a generally available stable model. It is priced at $0.15 per million tokens and ready for at scale production use!

“People don't listen to their friends, they listen to their competitors.” scripting.com/2025/07/13.htm…

Thrilled to welcome @windsurf_ai founders @_mohansolo & Douglas Chen and some of the brilliant Windsurf eng team to @GoogleDeepMind. Excited to be working with them to turbocharge our Gemini efforts on coding agents, tool use and much more. Great to have you on board!