motyl.dev Weekly #1 (Mar 2 – Mar 8, 2026)
A curated digest of what I found worth reading this week. Since this is first issue, I'm keeping it short and focused on AI.
Hello and welcome to the first issue of motyl.dev Weekly! I'm Greg, a software engineer and architect (in this AI era we'll all have to be architects, if we want to survive on the market). I'm building motyl.dev to share my thoughts on software engineering, AI, and productivity. I read all the news AI slop so you dont have to ;) and I am selecting best articles to share here.
Originally this project was meant to be AI powered summary of my own mailbox that is full of various tech newsletters. I wrote article about this the other day - Inbox Overload to Audio-Friendly Newsletters: My MCP Journey. You can read more about it there.
The News section is updated daily, and I will try to keep this weekly digest published every Sunday. You are welcome to participate in next newsletter content selection - all you have to do is to vote on articles that you find interesting by clicking on the upvote icon. It's that simple.
Okay enough with the introduction. I am so excited to have you here. Let's dive in!
AI
Building Claude Code with Boris Cherny Boris Cherny walks through how Claude Code evolved from an internal side project into a core engineering tool at Anthropic. The most striking detail: he runs five parallel Claude instances to ship 20-30 PRs a day, and argues that plain glob and grep beat every fancy RAG approach for agentic code search.
Google launches Gemini 3.1 Pro Preview Google's new mixture-of-experts model topped the Artificial Analysis Intelligence Index and hit state-of-the-art on ARC-AGI-2, GPQA Diamond, and Humanity's Last Exam — all at the same price as its predecessor. The competition for model quality is getting serious.
OpenClaw Founder to OpenAI... Now What? Peter Steinberger, creator of the 200k-star OpenClaw agent framework, is joining OpenAI to work on personal agents. OpenClaw moves to a foundation with OpenAI's backing. The community is cautiously optimistic but already stress-testing every word of that promise.
The AI workspace cookbook A practical breakdown of AI workspaces into five composable ingredients: instruction files, context folders, agentic memory, skills, and MCP servers. The real insight is that directory structure itself is context engineering, and each ingredient compounds the others.
Coding
Breaking the Re-render Chain: Our Migration from Context to Zustand A hands-on migration story from React Context to Zustand that documents the performance wins and architectural tradeoffs involved in taming cascading re-renders.
I actually tried AI coding and it's worse than I thought An intresting YT video worth your time. Not everyone is sold on AI-assisted development, and the friction points described here — hallucinated APIs, broken context, wasted debugging time — are real problems the tooling still needs to solve.
Architecture
The Future of Software Engineering with AI: Six Predictions "The Pragmatic Engineer" is newslteer I honesltly recommend. In this issue you can find six predictions about the future of software engineering with AI and some intresting statistics like 92% of developers use AI coding tools monthly, but outcomes diverge sharply: organizations see either 2x more incidents or 50% fewer, depending on their engineering foundations.
Tools
A Guide to Browser DevTools - The Network Monitor A solid walkthrough of the browser Network Monitor that covers the fundamentals most developers take for granted. Worth bookmarking if you've never systematically learned what each column and filter actually does.
AI Tooling for Software Engineers in 2026 A survey of nearly 1,000 engineers shows Claude Code went from zero to the top AI coding tool in eight months. 95% of respondents use AI tools weekly, and Anthropic's models dominate coding tasks by a margin that isn't even close.
Other
Deep Blue Simon Willison and the Oxide and Friends podcast coined a term for the psychological dread software developers feel as AI encroaches on their craft. Named after the chess computer that defeated Kasparov, "Deep Blue" captures something many in the industry feel but few want to admit.
Curated by Grzegorz Motyl. Subscribe for weekly updates.