Every week I share the top articles I have found and personally read on the internet. I only share the stuff I believe it’s worth your time.

100% signal, 0% slop.

I mainly read things from my public daily digest, which curates an ever growing list of sources. Here is this week’s best stuff 👇

🌀 How to Grow your Software Factory

13 min • Luca Rossi

This week I wrote a follow up from our popular The Era of the Software Factory, where I explored how to increase the leverage you get from AI, by creating, rules, modules, and making things compound. Co-authored with Rob Zuber, CTO at CircleCI.

🥇 Every Layer of Review Makes You 10x Slower

11 min • by Avery Pennarun

Awesome piece that many many people I know need to read. Every layer of review makes your process exponentially slower — and AI can speed up the coding part, but it does nothing against the time spent waiting for others to do stuff. Reducing reviews is less about AI and more about building a culture of quality and trust.

🥈 The Middle Loop

7 min • by Annie Vella

AI is shifting engineering work from creation to verification: Annie calls it the middle loop, sitting between the inner loop (writing code) and the outer loop (CI/CD, deploy). A useful mental model for how the day-to-day of engineering is evolving.

🥉 Don’t Let AI Write For You

2 min • by Alex Woods

Writing is fundamentally about structuring your thinking. Producing text is simply a (nice) byproduct of it. So, using LLMs to write for you undermines both your credibility and your ability to reason deeply. Use them for research and brainstorming, but the final words should be yours.

4) AI “Aha” Team Meetings

7 min • by Lara Hogan

There’s a growing gap between engineers who are thriving with AI tools and those who feel left behind. Lara held many interviews to learn more about how teams work today. She proposes regular team meetings where people share their AI aha moments: tips, tricks, and even mistakes.

And that’s it for today! If you find this list useful, please share it with your friends!

Share

You can also read the very same things I read every day via this free daily digest 📬

See you next week!

Sincerely 👋
Luca

Every week I share the top articles I have found and personally read on the internet. I only share the stuff I believe it’s worth your time.

100% signal, 0% slop.

I mainly read things from my public daily digest, which curates an ever growing list of sources. Here is this week’s best stuff 👇

🌀 The Telephone Game of Software

13 min • by Luca Rossi

This week I explored the steps that go from a feature simply being intent in a PM’s mind, to becoming actual code. Each of these steps can act as a lossy translation of the previous ones, and we need to prevent that.

🥇 Architecture Decision Record

8 min • by Martin Fowler

Maintaining ADRs is a great practice, which only got more useful with AI. Martin’s piece is a great primer for them, also linking to useful tools.

🥈 The 2028 Global Intelligence Crisis

25 min • by Citrini Research

I finally caught up with this viral imaginary report from 2028, that was published last month. Very useful to read to reflect on the second and third order effects of AI. It was a good test for my own beliefs: which predictions do I find reasonable? Which dubious?

🥉 Nobody Gets Promoted for Simplicity

8 min • by Matheus Lima

Tale as old as time — we most often reward complexity as a display of brilliance, and ignore simplicity. Engineers who over-build get a compelling narrative, while the ones who ship the simplest (and thus the best!) thing get… nothing.

4) Better, Faster, and Even More

10 min • by Rands

I am really enjoying Rands’ blog lately. It has become a blend of classic Rands’ management advice and occasional tales from his own AI experiments. This is one of the latter. I also love how practical he explains these — there are code snippets!

And that’s it for today! If you find this list useful, please share it with your friends!

Share

You can also read the very same things I read every day via this free daily digest 📬

See you next week!

Sincerely 👋
Luca

Every week I share the top articles I have found and personally read on the internet. I only share the stuff I believe it’s worth your time.

100% signal, 0% slop.

I mainly read things from my public daily digest, which curates an ever growing list of sources. Here is this week’s best stuff 👇

🥇 Finding Comfort in the Uncertainty

8 min • by Annie Vella

Annie joined the future of software development retreat, hosted by Martin Fowler and Thoughtworks, and reported back her thoughts, which I pretty much enjoyed. I love the eight themes she extracted and how she framed them.

🥈 GNU and the AI Reimplementations

9 min • by antirez

Antirez implicitly comments on the recent controversies about AI reimplementations (1, 2) by looking at examples from the past. I loved the historical perspective and also the new mental model he proposes: what matters now is design, novelty, and craft, not just the code itself.

🥉 Prioritize Relatively

2 min • by Boz

Super short post in Boz’s signature style, about how you should not prioritize things in isolation — but only prioritize them against each other.

4) Good Software Knows When to Stop

3 min • by Olivier Girardot

Good software knows the purpose it serves and doesn’t try to do everything. The hardest product discipline is recognizing when what you want to build next doesn’t fit your product vision — and that it should be a different tool instead.

And that’s it for today! If you find this list useful, please share it with your friends!

Share

You can also read the very same things I read every day via this free daily digest 📬

See you next week!

Sincerely 👋
Luca

Every week I share the top articles I have found and personally read on the internet. I only share the stuff I believe it’s worth your time.

100% signal, 0% slop.

I mainly read things from my public daily digest, which curates an ever growing list of sources. Here is this week’s best stuff 👇

🌀 The State of Product Development 2026

16 min • by Luca Rossi

This week we published our biggest research ever, in which we surveyed 340 teams on how they do product development, where the bottlenecks live, and where AI is (and is not) helping. Worth a read if you are interested in what teams actually do, without the hype!

🥇 Agentic Engineering Patterns

15 min • by Simon Willison

This is not a single article, but rather a growing collection of practices and patterns for getting the best results out of coding agents. Simon is an absolute authority on the topic, and I am excited that he has started to gather these systematically.

🥈 Judgment and Creativity Are All You Need

5 min • by Will Larson

Another great piece by Will Larson, where he argues that coding agents have largely solved the constraints of time and attention. The bottleneck has now shifted to judgment: knowing what to build and how to do it in a way that is scalable and maintainable.

🥉 The Claude C Compiler: What It Reveals About the Future of Software

12 min • by Chris Lattner

Chris Lattner, the creator of LLVM, reviews the AI-built C compiler from Anthropic. His verdict is that it’s a genuine milestone, calling it “a competent textbook implementation”. Worth reading to understand what Chris thinks AI can and can’t do today.

4) Block Cut 4,000 Jobs and Blamed AI

25 min • by Conor Bronsdon

Detailed write-up of the Block layoff story, including a lot of context about what possibly led to it, Jack Dorsey’s own track record, and benchmarks from other companies that are (and are not) going through similar paths. It’s a long read, 100% worth it!

And that’s it for today! If you find this list useful, please share it with your friends!

Share

You can also read the very same things I read every day via this free daily digest 📬

See you next week!

Sincerely 👋
Luca

Every week I share the top articles I have found and personally read on the internet. I only share the stuff I believe it’s worth your time.

100% signal, 0% slop.

I mainly read things from my public daily digest, which curates an ever growing list of sources. Here is this week’s best stuff 👇

🌀 The New Pyramid of Software Engineering

12 min • by Luca Rossi

Most AI advice feels either too volatile or too tactical to me. This week I published a piece that zooms out and proposes a three-layer pyramid — Developer Experience, AI, and Product Engineering — where each layer only works if the one below is solid. It also includes a self-assessment survey to rate your team on each layer.

🥇 My (Hypothetical) SRECon26 Keynote

8 min • by Charity Majors

One year after delivering the SRECon25 keynote, Charity reflects on how much her stance about AI has changed, going from grudgingly acknowledging it to fully embracing it. Superb post, extremely well written (as usual).

🥈 The Software Development Lifecycle Is Dead

11 min • by Boris Tane

Provokative post about how AI has compressed the SDLC to the point of being unrecognizable. The new lifecycle is a tight loop: intent → build → observe → repeat. The new big skill is context engineering, and the new safety net is observability.

🥉 The Stranger Secret: How to Talk to Anyone

8 min • by The Guardian

We’ve lost the habit of talking to strangers. Technology and social norms have shrunk everyday exchanges to near zero, and that’s a loss — for us and for society. This is a great reminder that small, unplanned conversations are one of the most enriching things we can do, and that the skill is worth actively maintaining.

4) Red/Green TDD — Agentic Engineering Patterns

2 min • by Simon Willison

“Use red/green TDD” might be the highest-leverage four-word prompt you can give a coding agent. These are Simon’s words, and I confirm! Write tests first, confirm they fail, then let the agent iterate until they pass. It prevents the two biggest agent risks: code that doesn’t work, and code that’s unnecessary. Part of Simon’s excellent new Agentic Engineering Patterns guide.

And that’s it for today! If you find this list useful, please share it with your friends!

Share

You can also read the very same things I read every day via this free daily digest 📬

See you next week!

Sincerely 👋
Luca

Every week I share the top articles I have found and personally read on the internet. I only share the stuff I believe it’s worth your time.

100% signal, 0% slop.

I mainly read things from my public daily digest, which curates an ever growing list of sources. Here is this week’s best stuff 👇

🌀 The Era of The Software Factory

14 min • by Luca Rossi & Rob Zuber

This week I have published some reflections following the latest State of Software Delivery report, and how it feels that the whole dev process is shifting into a new era. This is a thought piece grounded in a lot of data, which is hard to ignore at this point.

🥇 Outcome Engineering — The o16g Manifesto

15 min • by Cory Ondrejka

A manifesto for what comes after software engineering in the age of AI agents. I really liked the 16 principles and how they reframe our profession around outcomes instead of code. From killing the backlog, to verifying reality, to coordinating agents like a new kind of org. It all makes sense.

🥈 One List to Rule Them All

8 min • by James Stanier

Another great piece by James. The power of a clear, short, and single list of priorities can’t be overstated. I also love the many examples included in the article. Recommended!

🥉 In Defense of Not Reading the Code

10 min • by Ben Shoemaker

When AI writes the code, your focus should shift from reviewing the code line by line to investing in the harness — specs, test infra, observability. This is the type of controversial-but-grounded article we desperately need more of.

4) Agentic Email

4 min • by Martin Fowler

A short reflection by Martin Fowler about giving LLMs access to your email. FWIW, with OpenClaw I do exactly what Martin suggests: read-only access + drafts for human review.

And that’s it for today! If you find this list useful, please share it with your friends!

Share

You can also read the very same things I read every day via this free daily digest 📬

See you next week!

Sincerely 👋
Luca

Every week I share the top articles I have found and personally read on the internet. I only share the stuff I believe it’s worth your time.

100% signal, 0% slop.

I mainly read things from my public daily digest, which curates an ever growing list of sources. Here is this week’s best stuff 👇

🌀 My Experience with OpenClaw

14 min • by Luca Rossi

This week I published a full write-up of how I use OpenClaw at Refactoring and in my personal life. It’s been a transformational tool for me, and I believe a lot of critics are missing the point. In the article I cover it all: workflows, security, and some hot takes about the future.

🥇 The AI Vampire

13 min • by Steve Yegge

AI makes you wildly more productive, but where does all that extra value go? If the company captures it all, you just burn out faster. A lot of great thinking by Steve Yegge here. A sneaky risk of AI isn’t that it replaces you, it’s that it drains you while you keep shipping more.

🥈 How I Use Claude Code to Ship Like a Team of Five

9 min • by Kieran Klaassen

This is one of the best and most practical explainers about how to use Claude Code that I have encountered. Kieran goes through his exact workflows, and even includes video explainers!

🥉 Context Engineering for Coding Agents

8 min • by Birgitta Böckeler (I interviewed Birgitta here)

A thorough overview of how to configure what your coding agent sees: reusable prompts, context interfaces, MCP servers, slash commands. Build context gradually rather than dumping everything upfront, and iterate on what works. Birgitta is a top authority on this, and it shows.

4) Three Bad Managers

10 min • by Rands

Three archetypes of terrible managers (but good leaders!): The Artist who doesn’t value humans, The Dictator who dominates every conversation, and The Knife who is… well, just a mess. You’ve probably worked for at least one of these, the lessons are in what not to do.

And that’s it for today! If you find this list useful, please share it with your friends!

Share

You can also read the very same things I read every day via this free daily digest 📬

See you next week!

Sincerely 👋
Luca

Every week I share the top articles I have found and personally read on the internet. I only share the stuff I believe it’s worth your time.

100% signal, 0% slop.

I mainly read things from my public daily digest, which curates an ever growing list of sources. Here is this week’s best stuff 👇

🌀 CTO Archetypes

26 min • by Pat Kua

This week I am honored to host an original article by Pat on Refactoring! He proposes a framework to reflect on CTOs’ scope and responsibilities, based on 8 archetypes. It’s great for CTOs to understand where they fit best, for planning your career, and for improving alignment during hiring.

🥇 Automatic Programming

2 min • by antirez

Stop calling it vibe coding. When it’s engineers doing it, Antirez suggests the better (imo) term automatic programming. Automatic programming produces vastly different results depending on the human guiding the process with their intuition, design, and continuous steering. Vibe coding, instead, means generating software without being part of the process at all

🥈 Notes on Claude Coding

3 min • by Andrej Karpathy

When Andrej Karpathy speaks, we listen. In just a few weeks, he went from 80% manual coding to 80% agent coding. Great note on how his workflow is changing and his thoughts about it.

🥉 Shipping at Inference-Speed

13 min • by Peter Steinberger

Peter is the guy behind the sensational OpenClaw. He wrote this blog post about a month ago, and he explained in detail how he writes code with AI. It’s a refreshing take where you won’t find complex orchestrations or multi-agent setups. Mostly common sense stuff by a single human developer.

4) How to Write a Good Spec for AI Agents

27 min • by Addy Osmani

This is a lengthy post where Addy provides a lot of detailed advice on how to work with agents. With respect to Peter’s piece, there is a lot less chaos and more guardrails — it’s a more structured approach that works better for consolidated teams.

And that’s it for today! If you find this list useful, please share it with your friends!

Share

You can also read the very same things I read every day via this free daily digest 📬

See you next week!

Sincerely 👋
Luca

Every week I share the top articles I have found and personally read on the internet. I only share the stuff I believe it’s worth your time.

100% signal, 0% slop.

I mainly read things from my public daily digest, which curates an ever growing list of sources. Here is this week’s best stuff 👇

🌀 The Listening Tour

10 min • by Luca Rossi

Yep, that’s me! Many asked why I don’t include the new articles I write on Refactoring here, and yeah, I should probably do that! The listening tour is about how to find bottlenecks and friction, by interviewing your engineers and mapping problems to your dev process.

🥇 How I Use Claude Code

5 min • by Boris Cherny

Boris Cherny is the creator of Claude Code, and in this thread on X he reveals how he uses it. Spoiler: the setup is pretty vanilla, which is refreshing. He also describes what the whole team shares and how they keep agents improving all the time.

🥈 Compound Engineering: How Every Codes With Agents

15 min • by Dan Shipper & Kieran Klaassen

Warning: this is a paid article, but I kept finding it linked everywhere so I subscribed to Every just to read it. In traditional engineering, every feature generally makes the next harder to build. The idea behind compound engineering is that you can reverse this: each feature should make the next easier, by means of a loop through which AI agents get better and better. This was very interesting and well-worth the sub.

🥉 How I Estimate Work as a Staff Software Engineer

12 min • by Sean Goedecke

It is not possible to accurately estimate software projects — but why? This is possibly the best article I have ever read about estimates. It talks about knowns, unknowns, and reversing the script.

4) It’s better not to continue everything

4 min • by Benedikt Kantus

Continuation is our default mode of operation, but what if we flipped it? Make stopping the default: every quarter or year, everything pauses automatically. Free yourself from the sunken cost fallacy and only invest in what makes sense now. Great read!

5) Long Desperation: An Investment Thesis

15 min • by Cerebus

A significant portion of young people have become economically imprisoned: when housing prices have tripled against incomes, student debt has skyrocketed, and AI threatens career paths, traditional compound interest advice falls apart. Desperate people turn to gambling high-variance financial products looking for agency. The author suggests investing in platforms that help with that, but instead I found myself wondering how to reverse the trend.

And that’s it for today! If you find this list useful, please share it with your friends!

Share

You can also read the very same things I read every day via this free daily digest 📬

See you next week!

Sincerely 👋
Luca

Every week I share the top articles I have found and personally read on the internet. I only share the stuff I believe it’s worth your time.

100% signal, 0% slop.

I mainly read things from my public daily digest, which curates an ever growing list of sources. Here is this week’s best stuff 👇

🥇 Welcome to Gas Town

33 min • by Steve Yegge

This was the best—and wildest—reading of my week. Gas Town is like Kubernetes for AI coding agents, only with more chaos and Mad Max names. It’s an extreme take on what AI coding can be, which is useful to explore in any case: even if you are not ready for it (I am not!) it informs your intuition about what’s happening on the frontier.

🥈 Don't fall into the anti-AI hype

8 min • by Antirez

This article is the best example of the idea that “smart people can hold two opposite ideas in your head”. Salvatore “Antirez” Sanfilippo, the creator of Redis (who I interviewed here), is not a fan of AI and doesn’t want it to succeed, but he doesn’t let that interfere with his judgment as a software professional about what’s the best way to perform his job:

it is now clear that for most projects, writing the code yourself is no longer sensible, if not to have fun.

🥉 This Might Be Oversharing

6 min • by James O’Malley

A candid piece about how the author’s social life collapsed post-pandemic — and how he intentionally rebooted it. It gave me a lot to think about and provided great insights on building community. Might inspire you to organize dinners!

4) One Bottleneck at a Time

7 min • by James Stanier

Great piece that builds on the famous The Goal, by Eliyahu Goldratt. Systems have one constraint that matters most, and focusing elsewhere is a wasted effort. Find the bottleneck (PRs piling up? Slow deploys?), put your best people on it, fix it, then move on to the next one.

5) I Sell Onions on the Internet

12 min • by Peter Askew

This is an old article, and pretty famous too. I stumbled upon it again after a long time, and it’s just a delightful story that comes from what seems like a different era of the internet, while instead it’s just five years ago. I enjoyed reading it again.

And that’s it for today! If you find this list useful, please share it with your friends!

Share

You can also read the very same things I read every day via this free daily digest 📬

See you next week!

Sincerely 👋
Luca

Every week I share the top articles I have found and personally read on the internet. I only share the stuff I believe it’s worth your time.

100% signal, 0% slop.

I mainly read things from my public daily digest, which curates an ever growing list of sources. Here is this week’s best stuff 👇

🥇 21 Lessons From 14 Years at Google

9 min • by Addy Osmani

This is really a great article with a lot of non-trivial and counterintuitive lessons. It’s also well written and everything is explained in a crisp and memorable way.

🥈 The Unbearable Joy of Sitting Alone in a Cafe

5 min • by Candost Dagdeviren

A reflection on the power of disconnection and solitude. I find myself more and more drawn to these kinds of experiences (and articles). I’ll probably try to do the same myself — going to a café without a phone!

🥉 Forest and Desert

2 min • by Kent Beck & Beth Andres-Beck

This is not a new article, but I read it again last week to use it in a presentation. It’s a really great analogy that explores a problem that rarely gets discussed: different teams live different realities, so there is a lot (good) advice that not only doesn’t work for some — it becomes harmful.

The last two articles I include are interesting because they present opposite theses. 👇

4) LLMs Are Not Fun

2 min • by Ori Bernstein

Ori explains how LLMs sucked the fun out of his job, because they took away 1) the joy of solving programming problems, and 2) the joy of management and coaching people, as you now spend more time “managing” LLMs than managing humans.

I see where Ori is coming from, but I think his feelings are based on a fallacy re: what his job is about. I wrote kind of a response to this piece in the latest Refactoring article:

• Finding Yourself in the AI Era

5) Web Development Is Fun Again

3 min • by Mattias Geniar

This is literally the opposite article than the one above. Mattias found renewed joy in programming with LLMs because they made it easier to make ideas come true, stripping away a lot of the stupid complexity brought in by the modern web.

And that’s it for today! If you find this list useful, please share it with your friends!

Share

You can also read the very same things I read every day via this free daily digest 📬

See you next week!

Sincerely 👋
Luca

Every week I share the top articles I have found and personally read on the internet. I only share the stuff I believe it’s worth your time.

100% signal, 0% slop.

I mainly read things from my public daily digest, which curates an ever growing list of sources. Here is this week’s best stuff 👇

🥇 The Move Faster Manifesto

5 min • by Brian Guthrie

Speed is all that matters in software. I am a speed purist, meaning that I believe you can optimize for speed and basically everything else falls into place. This is a great manifesto that explains why it is the case, and what we should mean when we say “speed”.

🥈 Coding is the Least Interesting Part of the Work

8 min • by Obie Fernandez

This is another article I wish I had written myself. For engineers, the most fun, interesting, and ultimately real work is what happens before you write code. Which is why you shouldn’t be worried if you are touching less code because of coding agents. Not only that: chances are, by spending more time on the thinking rather than the typing, AI can make you grow faster at your job.

🥉 Maybe the Default Settings Are Too High

8 min • by David Cain

A meditation on slowing down consumption — reading, eating, everything. Modern life pushes us to consume fast, which paradoxically diminishes rewards. Slowing down, conversely, enhances both comprehension and enjoyment. This is just the perfect read to start the new year.

4) A Month of Chat-Oriented Programming

15 min • by Nick Radcliffe

An honest, no-BS account of pair programming with Claude Code for a full month. You will find that almost every week I post one of these in this digest, and I’ll keep doing that: things change fast, and it’s important to get fresh takes all the time. Here the author got stressed and furious at times, but also produced way more code than they would have alone.

5) Why Does Development Slow?

5 min • by Kent Beck

Software projects start fast, then slow to a crawl — and AI coding assistants make it worse. Kent argues it’s all about options: each new feature reduces future flexibility. The solution is alternating between adding features and restoring optionality through code cleanup.

Also made me think about how we defined milestones in our guide about planning software: create milestones that 1) maximize progress and 2) minimize design space reduction.

And that’s it for today! If you find this list useful, please share it with your friends!

Share

You can also read the very same things I read every day via this free daily digest 📬

See you next week!

Sincerely 👋
Luca

Every week I share the top articles I have found and personally read on the internet. I only share the stuff I believe it’s worth your time.

100% signal, 0% slop.

I mainly read things from my public daily digest, which curates an ever growing list of sources. Here is this week’s best stuff 👇

🥇 Thin Desires Are Eating Your Life

4 min • by JA Westenberg

This time of the year always sparks some reflection, and I find myself drawn towards philosophical articles more than technical ones. I love the simple mental model presented here: a thick desire is one that changes you in the process of pursuing it; a thin desire is one that doesn’t. Time to focus on what matters!

🥈 Logging Sucks

30 min • by Boris Tane

Amazing primer about how to move from a logging to an observability mindset. The solution is wide events: single, rich objects that capture everything about a request. This is the article you will keep linking to your friends who ask about observability.

🥉 The Depopulation Trade

8 min • by Stefano Bernardi

Stefano is a smart VC who put together a thorough thesis about what the future may look like. It’s a worthy read. In a nutshell, population collapse is happening faster than anyone predicted. With fewer people, we’ll need to either increase births, import workers, or reduce the need for workers altogether. Guess which option tech and AI are betting on — and what that means for the future of work.

4) Why People Really Quit

3 min • by Kim Scott

Kim Scott, author of Radical Candor, is back at it with a great piece that identifies two mgmt failure modes: micromanagement and absentee management. The sweet spot is being a thought partner — actively involved but empowering. The recap table is a gem and I am attaching it here 👇

5) Claude, Chrome, Cloudflare

2 min • by Simon Willison

Real-world example of using the brand new Claude browser extension to navigate Cloudflare’s maze of a dashboard. Simon’s blog is the #1 resource to learn about AI & coding these days, and this quick—but useful—post confirms it.

And that’s it for today! If you find this list useful, please share it with your friends!

Share

You can also read the very same things I read every day via this free daily digest 📬

See you next week!

Sincerely 👋
Luca

Every week I share the top articles I have found. I only share stuff I have read (or tried if it’s a tool) and I believe it’s worth your time. 100% signal, 0% slop.

I mainly read things from my public daily digest, which curates an ever growing list of sources. Here is this week’s best stuff 👇

👓 Readings

🥇 Facilitating AI Adoption at Imprint

13 min • by Will Larson

This is a side-quest (to say the least) for every engineering leader right now. Will shares his working notes on how he is driving AI adoption at Imprint, with a lot of practical ideas and lessons learned.

🥈 Using LLMs at Oxide

8 min • by Oxide

This is the best manifesto about team LLM adoption I have seen. It covers values, patterns, and anti-patterns, covering everything from first principles. Chances are you can take it and almost copy-paste it for your team. It’s also the perfect complement to Will’s article above.

🥉 Useful Engineering Management Artifacts

5 min • by Bjorn Roche

A practical toolkit that also works as a checklist for everything you may need as an engineering manager. Career development plans, title ladders, team charters, roadmaps, and more — all the key docs that help you scale management without losing your mind.

4) Would You Pay to Have Your Resume Read?

4 min • by Pawel Brodzinski

Great piece about how AI has broken tech hiring. It’s become a game of prompting one AI agent to pass through the filters of another AI agent. So, what if candidates paid a small fee to guarantee a human actually reads their resume?

5) Deliberately Shaping Engineering Orgs

3 min • by Aviv Ben-Yosef

Great list of four failure modes that happen when you are not intentionally shaping your engineering org. A reminder that you need to carve out time to deliberately design your team’s future state.

🔧 Tools

Skippr Finesse

A cool Chrome extension that gives product feedback about the website you are on! Simple, well-executed idea.

Aura Ink

A color e-paper frame that looks just amazing. I am a sucker for digital stuff that blends into your home, and this is the epitome of that. Too bad it costs an arm and a leg 😂 ($450)

Echo

Promising tool (still in early access) that turns your inbox into a doc with the main actionable items. Maybe it’s because I tend to write docs this way myself (simple headings + bullet points) but I think there is a lot of potential in products that take something messy (e.g email) and use AI to turn it into a simple doc. It got me thinking about what other domains could get this treatment.

And that’s it for today! If you find this list useful, please share it with your friends!

Share

You can also read the very same things I read every day via this free daily digest 📬

See you next week!

Sincerely 👋
Luca

Here are this week’s top items 👇

👓 Readings

🥇 Frictionless

Book • by Nicole Forsgren & Abi Noda

I just finished reading the new book by Nicole Forsgren (co-author of Accelerate) and Abi Noda (co-founder of DX) about removing barriers that slow engineering teams down. I loved it — feels like the perfect follow up to Accelerate.

🥈 Increasing Your Luck Surface Area

2 min • by Jason Roberts

I stumbled on this while reviewing Umberto’s article on Refactoring, and loved it. Your luck is not random, it’s a function of doing things + telling people about them. The more you do, and the more you share, the more serendipity finds you.

🥉 Invert, Always Invert

10 min • by James Stanier

Charlie Munger’s classic mental model applied to engineering leadership. Instead of asking “how do I succeed?”, ask “how would I guarantee failure?” Then avoid those things. Very practical article with a lot of examples.

4) How Good Engineers Write Bad Code at Big Companies

8 min • by Sean Goedecke

Great take that connects the dots between bad code, short tenure, and incentives that prize shipping rather than maintaining. That’s how you get great engineers to write bad code. It’s a systemic issue, not a talent one.

5) Revisiting Manager READMEs

4 min • by Camille Fournier

Years after her famous critique of Manager READMEs, Camille revisits the topic. She argues that READMEs are not 100% bad, but you can easily do better by decentralizing the instructions into the respective processes and ceremonies. Make instructions less about you and more about what needs to be done.

🛠️ Tools

Fizzy

Couldn’t I leave out a new tool from 37signals? Nope. Fizzy is a very opinionated take on Kanban. In a nutshell, it’s about keeping momentum, keeping lists fresh, and avoiding long backlogs. It’s peak 37signals philosophy, but the design is extremely polarizing and will not appeal to everyone!

PGlite

Full Postgres running in your browser via WASM. Just 3.7MB compressed. So cool! Useful for local-first apps, demos, or anywhere you need a real database without a server.

Pylar

Sensible middleware to give AI agents safe access to your structured data sources — i.e. databases. You can create governed SQL views, build MCP tools, and deploy to popular agent builder. For teams (legitimately) worried about letting LLMs loose on their dbs.

And that’s it for today! If you find this list useful, please share it with your friends!

Share

You can also read the very same things I read every day via this free daily digest 📬

See you next week!

Sincerely 👋
Luca

It’s still very experimental and every week I am trying new things. Here is what’s new today:

• 🏆 There is a ranking — I ranked articles in rough order of how much I liked them, so if you don’t have a lot of time you can choose to read only the best of the best.

• 📣 There is a sponsor — I am including one sponsored tool among the ones on the weekly list, but only if I can genuinely recommend it! Today’s one is Snyk.

Back to the list!

👓 Readings

🥇 Become the Consequence

10 min • by Rands

How to think about your role as a senior leader. You are not thinking in tasks anymore — you are thinking in strategy, sharing it repeatedly and enforcing the consequences of achieving / not achieving goals.

🥈 Tying Engineering Metrics to Business Metrics

4 min • by Iccha Sethi

How to tie leading engineering metrics like test coverage or PR size, to lagging business metrics like CAC or ARR. I consider myself a connoisseur of metrics articles and this is a really good one!

🥉 Writing a good CLAUDE.md

9 min • by Kyle Mistele

How to write an effective CLAUDE.md by thinking about strategy and first principles, rather than pure tactics. Loved many of the ideas, like the progressive disclosure, less is more, and the why/what/how.

4) Productivity Rain Dances

2 min • by Cal Newport

There are busy activities that feel productive but don’t lead to real results. Productivity nerds like me are often guilty of focusing more on inputs (how I do stuff), rather than actual outputs (what results they bring). Great reminder and memorable mental model.

5) Detailed Nano Banana Prompting

24 min • by Max Woolf

I think Nano Banana Pro didn’t get enough attention vs the wider Gemini 3 (and now Claude Opus 4.5) releases. For me it’s been an incredible step change in many of the things previous models struggled with, like text and diagrams. It can also be prompted with a lot of detail, and this is the best guide I have found about it.

🔧 Tools

1) Snyk (sponsor!)

Snyk is an awesome platform to handle all-things-security for your engineering team, and I have used it a lot in the past. Next week they are also hosting a live session about the new OWASP Top 10, to go through the new emerging threats, and practical measures you can adopt.

2) HOTO Tools

Ok this is not software! Inspired by this amazing video (check out the whole channel, really), I went on to a quest to upgrade some of the DYI tools I have at home, and I just love these by HOTO. I have one of their screwdrivers and am planning to buy more stuff!

3) Agenta.ai

This looks very good: a full open-source suite to handle all-things-LLMOps, including prompts, evals, observability, and more.

4) Once UI

I am a sucker for design systems / frameworks, but I also have a high standards. Once clears them: it’s open source, comprehensive, and includes both code (Next.js) and design (Figma).

And that’s it for today!

Sincerely 👋
Luca

Last week (actually two weeks ago! I took one off 🙏) I published a different email than usual, which instead of being an essay, was a short list of the best articles and tools I had found that week.

I did that because so many people ask me what to read these days, with AI and everything else going on.

So I made what I read public via this Mailbrew digest that you can clone / subscribe to by yourself, and condensed the very best items in that very email.

Then I asked readers what they thought about it, and the result was… astounding! 👇

So I got back to the drawing board, and thought deeper about how to turn this into a permanent, high-quality thing.

As most of you know, I am a full-time content creator, and other than Hybrid Hacker I also run another newsletter, Refactoring.

The way I think about my work is that of a curator of ideas. I read a lot of stuff, distill the best, come up with my own ideas, and combine them into long-form essays.

This work, which is largely individual and private, has also a public version attached, which is what makes me a creator, as opposed to just someone who likes to study things 👇

What’s interesting is that not all the things I do have such a public version. Namely, up until last week, both my reading list (sources) and the short list of things I have read and liked have always been 100% private.

So I realized that what I unintentionally did last week was attach a public version to these two remaining bits, in the form of the Mailbrew digest and the Hybrid Hacker newsletter.

I really like this idea, because it fixes two problems at once:

• It creates more incentives for me to do things well — as now it’s literally all public.

• It creates a clear separation between Hybrid Hacker and Refactoring, which have always had a lot of overlap otherwise.

So I am going to keep this going for a while and see how it turns out!

You may have questions, so feel free to reply to this email and I will get back to you. The most common one might be: what if you are a paid subscriber of Hybrid Hacker? I want these weekly digests to be free for everyone, but since you subscribed (and paid for) long-form essays, I am happy to convert your HH subscription into a Refactoring one at no additional cost (Refactoring costs $15 instead of $10, so consider this a Black Friday deal!). Refactoring, if you are not familiar with it, covers the same topic, with the same style, and also has a podcast, community, and large library attached to it!

If you prefer to unsubscribe instead, just reach out and I will issue a prorated discount. But I really encourage you to stay and see what we are up to!

So, if you missed last week’s email, you may wonder what these digests look like, so here is the one for this week! 👇

👓 Weekly Readings

1) Thoughtworks’ Technology Radar

30 min • by Thoughtworks

This bi-annual tech radar is simply a must-read every time it comes out. Get yourself a cup of your favorite hot beverage and go through at least the Adopt and Trial sections.

2) Understanding Spec-Driven Development

13 min • by Birgitta Bockeler (↗ our interview with her)

SDD is one of the latest buzzwords in working with LLMs, and Birgitta did an awesome job at demystifying it from first principles, plus trying a bunch of tools for it. More resources:

↗ SDD: The Waterfall Strikes Back — a counter-take

↗ Kiro • Spec-kit • Tessl — the tools tried by Birgitta

3) How to Write a Great AGENTS.md

7 min • by Matt Nigh (Github)

This article was super useful to me to get practical examples of which agents you should create, how to set their boundaries, and how to structure the agent files.

4) Will Larson is back to coding at work after a decade away

5 min • by Will Larson (Imprint)

Will is a legend and this is the best article I have read about how managers can get back to coding thanks to AI, and whether they should. It is a balanced take that includes opportunities, risks, and how to do it well.

5) Maybe You Are Not Actually Trying

6 mins • by Cate Hall (Astera)

Amazing article about agency. Sometimes we stop trying in one area of life because we got stuck once and never revisited. That frozen response can look like effort, but it isn’t so. It gave me a lot to think about.

🔧 Weekly Tools

1) Nano Banana Pro

Mind-blowing new model by Google for image generation. It is especially great at diagrams and infographics, most often one-shotting them with perfect text — and in 4K!

2) Google Antigravity

Google was on fire this week and also released its very first IDE. It’s a fork of VS Code with a more agentic spin. I tried on a few projects and it feels different than Cursor, will probably write a longer take at some point. I appreciate it allows for multiple models, but I guess it’s still not 100% optimized because I keep hitting quota limits, apparently for concurrent requests.

3) Kiro / Spec-kit / Tessl

After reading the SDD article above I went into a rabbit hole and tried a bunch of tools for it!

4) LLM Council

Nice open source project by Karpathy where your questions are sent to multiple LLMs and each model reviews and ranks the others’ answers anonymously. Similar to Pewdiepie’s experiment, which I also enjoyed a lot.

And that’s it for today!

Sincerely 👋
Luca

Today I am trying something different than usual, and rather than giving you an essay about some specific topic, I am going to list the best articles I have read this week, plus the most interesting tools I spotted.

Why?!?

Because so many people ask me what I am reading these days. Engineers are hungrier than ever for good reading, because of everything that it’s going on.

So I started by doing a simple thing: I made what I read public. I created this Mailbrew digest that I receive every day, which includes a combination of community listings (Hacker News, Slashdot, Product Hunt), magazines, and amazing individual blogs.

Mailbrew it’s free and you can clone the digest or simply subscribe to it to read literally the same things I read every day 👇

Read the same things I read

But that’s not enough! I still read probably only 10% of what I receive in the digest, and, out of that, very few articles were actually worth my (and your) time.

So let’s go one step forward, and just list the best things I actually read, and why.

Also, mind you, the fact I read those this week doesn’t mean they went out this week (even if it’s often the case). I could have saved them for later and caught up after days, weeks, or even months!

But if you accommodate for some delay, the promise here is immaculate signal — only the very best comes through.

So here they are for this week:

👓 Weekly Readings

1) Positraction

2 min • by Boz (Meta)

Boz managed to put into words why diversity matters, better than I ever could. An evergreen piece, in only 200 words.

2) The Profitable Startup

4 min • by Karri Saarinen (Linear)

Linear is spear-heading a culture shift in which even top-tier, VC-backed startups can focus on quality and profitability rather than growth at all costs. Awesome article with many timeless mental models.

3) Collaboration sucks

4 min • by Charles Cook (PostHog)

Refreshing and provocative piece on why solo work is often preferable to collaboration. This is one of the most important mental models to wrap our heads around, especially now in the age of AI and product engineering.

4) Reflections on Palantir

24 min • by Nabeel Qureshi (Palantir)

One of the best pieces I read the whole year. It covers so much ground: the case for working for morally ambiguous companies, forward-deployed engineering (which is hot right now), company culture, and more.

5) The Sound of Inevitability

2 min • by Tom Renner

Inevitabilism is the belief that certain developments are impossible to avoid. This is most often false, and we should fight for the future we want, rather than take others’ futures for granted.

I also have a short list of tools I found this week which I saved for later.

I can’t fully endorse them because I barely managed to try them, but that’s not the point: when something looks interesting, I save it in my tool database on Notion, and possibly fetch it in the future when needed.

Consider this a way to keep the pulse on new things that get released:

🔧 Weekly Tools

1) Jam.dev

Auto-captures all the info engineers need to reproduce a bug. Looks pretty neat!

2) Vibecode

A mobile app to vibe-code other mobile apps! It builds everything in React Native + Expo, and you can test it live on your phone. Nice and well-made!

3) Fire your QA

The name was enough to look into it 😂 looks pretty clever but hard to figure out how well it will work on real-world scenarios.

4) Chatter

Search for stock tickers and fetch ideas from Reddit. This was way better made than I expected! It gives plenty of context about the stock and even displays the hit rate of the authors of posts.

5) Cyrus

A Claude Code agent that attaches to Linear and works autonomously when you attach tasks to it. I got pitched Cyrus many times and only recently got to test it, and it’s really cool (see the video demo for more detail).

And that’s it! How did you like this email? Since it’s a new format, I would love to get feedback from you 👇

And that’s it for today!

Sincerely
Luca

It’s an exclusive club with other illustrious members, such as writing tests, estimating tasks, or updating JIRA.

The natural restraint towards such tasks is often dismissed as laziness. But laziness carries meaning.

When devs don’t want to do something, what they are really saying is:

• It’s a hassle — which means: high cost

• It’s not worth it — which means: low value.

People have a small ROI calculator in their head, which is saying that the ROI is negative.

This version of laziness is useful and has driven important conversations in the past:

• Tests — many teams today write less unit tests in favor of more integration tests. We went from the testing pyramid to the testing trophy.

• Estimates — we spend less effort on estimates than in the past. T-shirt sizes and story points often replace man-days or hours, being both lighter and more manageable.

The goal of this article is to have the same conversation on tech docs to make them 1) less of a hassle, and 2) more worthwhile.

This is what we are going to cover:

• 📖 Types of tech docs — the three main ones.

• ✏️ What are design docs — and why you should write them.

• 📋 Structure of a design doc — my own template based on three areas: essentials, disclaimers, and reminders.

• ❓ FAQs — maintaining docs, involving stakeholders, and deciding when docs are not needed at all.

• 📚 Resources and case studies — how design docs are managed at companies like Google, Plaid, and others.

Let’s dive in 👇

📖 Types of tech docs

This is the first source of confusion. Many times you hear people asking “do we have the docs?” — but what docs?

It is useful to organize them into three main types:

1. Functional spec — the product requirements. To communicate what should be achieved from the user perspective. It includes UI/UX specs.

2. Design doc — the tech design and implementation strategy. To record decisions, communicate trade-offs, and converge on a solution.

3. Tech spec — API specs, database schema, and the overall description of what’s live.

In my experience, everybody understands the difference between the product and the tech stuff — that is, between the first type and the other two. It is easy because those are usually created by different people: PMs and designers on one side; engineers on the other side.

On the contrary, the difference between design docs and tech specs is tricky.

In my opinion, 80% of the time what you really need is a design doc rather than a tech spec. That is, you need to record your decisions rather than describe what’s live.

To understand this better, let’s cover what design docs are and why you should write them.

✏️ What are design docs

A design doc illustrates the tech design and implementation strategy for a given initiative.

It is meant to be created before you start writing code, and to be shared with other stakeholders to converge on the solution.

You write it for several reasons:

1. 💭 Help your reasoning — writing is thinking. Going through the process of writing a formal document helps solidify your reasoning and come up with a better solution.

2. 🔨 Improve the design — find the best solution through the contribution of others. Like a code review, but for the design.

3. 🤝 Create consensus — align with stakeholders about what should be built, make sure important decisions are participated, and make people feel invested in them.

4. 📓 Record the decision — for posterity and documentation. That includes the trade-offs you evaluated, alternative solutions you discarded, open questions, and more.

In that respect, a design doc is useful throughout the whole lifecycle of a project.

• Before release — it drives the process that makes people converge on a good solution.

• After release — it acts as a decision record, that is often more useful than pure tech specs.

So, what goes into a design doc? 👇

📋 Structure of a design doc

Read more

Back then I published my review on Refactoring, and I recently picked it up again because I believe many of the topics it covers have become especially relevant today, with all the AI craziness.

Some context first: Kahneman was a psychologist and economist. In 2002 he was awarded the Nobel Prize in Economic Sciences for his work in applying psychological insights to economic theory, particularly in the area of judgment under uncertainty. He passed away last year — so reading and reviewing his work felt like a fitting tribute.

So, “Thinking, Fast and Slow” is not just another pop psychology book. It’s a world-renowned cornerstone work in understanding human decision-making and cognitive biases, backed by decades of rigorous research.

And it shows.

In fact, if you get the physical edition, the first thing you will probably notice is that this is a hefty book. It clocks in at nearly 500 pages, with a 35-page appendix of notes and references. It’s the work of a lifetime: the lifetime of one of the most brilliant and prolific scientists of his generation.

“Thinking, Fast and Slow” is organized following a top-down approach: it starts with big ideas and gradually drills down into specifics. So, while it contains an inordinate amount of ideas, it never feels daunting or overwhelming, thanks to the clear structure, plenty of examples, and the easy writing style. We will follow its structure pretty much verbatim in this review.

Also, while it wasn’t written specifically for people in tech, I have found its insights to be profoundly relevant to our job.

As engineering leaders, we’re constantly making decisions: we estimate projects, assess risks, allocate resources, and navigate complex human dynamics. Most of these decisions are made under uncertainty, and Kahneman’s work shows just how bad we are at these. He explains, with scientific precision, why we make irrational decisions, even when we believe we’re being perfectly logical.

Ultimately, the book challenges us to think about thinking, by revealing invisible forces that shape our judgment. Understanding these forces helps us make better decisions, in work and life.

So let’s explore the key concepts from the book, and discuss how they apply to our work. Here’s our agenda for today:

• 🧠 Two Systems — understanding our dual thinking processes

• 🎭 Biases and Heuristics — recognizing our mental shortcuts

• 📊 Prospect Theory — rethinking how we assess value and risk

• 🤝 The Two Selves — balancing experience and memory in design

Let’s dive in!

🧠 A Tale of Two Systems

Kahneman’s central thesis revolves around two modes of thought: System 1 and System 2. This concept is fundamental to understanding how we make decisions:

• 1️⃣ System 1 — is fast, intuitive, and emotional. It operates automatically and quickly, with little or no effort and no sense of voluntary control.

• 2️⃣ System 2 — on the other hand, is slower, more deliberative, and more logical. It allocates attention to mental activities that demand it, including complex computations.

What’s surprising — and somewhat unsettling — is how much we rely on System 1, even when we think we’re being logical and methodical. We often pride ourselves on our analytical abilities (especially in engineering 🫠), but Kahneman shows that we’re far more inclined to make quick, intuitive judgments.

The main problem is that we don’t have a reliable way to figure out when to engage System 2 — the analytical side — vs. accepting the quick answer provided by System 1.

This is perfectly displayed in the famous bat and ball problem:

“A bat and ball cost $1.10. The bat costs one dollar more than the ball. How much does the ball cost?”

Most people answer 10 cents, which is System 1 stuff. The correct answer (5 cents) would require engaging System 2, but we usually accept the quick answer because it looks good.

This reminded me of chess. During lunch I often watch Hikaru videos, where he often talks about the problem of knowing when to spend your time. Modern chess has shifted more and more towards shorter time controls (e.g. 10-minute games), which means players spend little time, on average, on every move. What’s interesting, especially in grandmaster games, is that players do not spend a similar amount of time on every move — they blitz most of them (i.e. few seconds) and spend long chunks of several minutes on a few crucial ones.

Blitz moves are System 1 — quick judgment and pattern-matching — while the long ones are when players engage their analytical brains.

Knowing when it is worth spending more time on a move — i.e. when to engage System 2 vs when to trust intuition — is a crucial quality that separates outstanding players from the good ones.

Ok, but chess is time-constrained: in real life, couldn’t we deliberately skew towards System 2 most of the time? No sir, because System 2 is 1) slow, and 2) extremely costly, while System 1 is basically free.

Using System 2 quickly leads to cognitive strain, so we can rely on it sparingly.

About this, Kahneman described an experiment where people were asked to solve math problems while simultaneously memorizing a 7-digit number. As their System 2 was busy with the memorization, they made impulsive (System 1) decisions on the math problems, leading to comical mistakes.

But what mistakes? Time to talk about biases 👇

🎭 Biases and Heuristics

In the second chapter, Kahneman explores a wide range of cognitive biases and heuristics. These shortcuts often serve us well, but they can also lead us astray, especially in complex scenarios.

The book provides a long list, and you may already know many of them. Here are the ones that resonated the most with my own experience:

Read more