Hey, Luca here, welcome to a weekly edition of the💡Monday Ideas💡 from Refactoring! To access all our articles, library, and community, subscribe to the full version:
Warp is the AI terminal built for devs who want to use agents in every step of their workflow. It’s the top overall coding agent, and it’s #1 on Terminal-Bench, ahead of Codex and Claude Code.
It includes:
✅ Long-running commands — something no other tool supports.
✅ Agent multi-threading — run multiple agents in parallel, all under your control.
✅ Full dev lifecycle support — setup → coding → deployment.
However, 90% of it looks like the classic engineering tradeoff: doing a bit more work now (e.g. good code, good testing, good docs) in order to reap the benefits (i.e. higher velocity) later.
But the thing is: AI is tipping the scale, on both sides:
Doing good work is easier — on one side, AI can lift a lot of the grunt work off our shoulders: it is good at writing tests, good at writing docs, removes the cognitive load from remembering syntaxes and DSLs, and so on.
Benefits of good work are larger — on the other side, doing things right now doesn’t only benefit us humans, it also benefits AI itself. So it’s double the upside.
In other words, AI skyrockets the ROI of good engineering by simultaneously 1) reducing the effort, and 2) increasing the benefits.
We explored what this means (plus we took stock of where we are right now) in this recent article 👇
The piece did a great job at covering the what and the how, so I am here to add a small bit about the when.
There’s a saying that goes: there is no such thing as over-communicating — well, there definitely is. Managers are busy, so there’s a fine line between being informative and creating noise.
So just like you have a tech stack, I find it useful to define an update stack — a combination of tools and cadences that work well together. For example:
💬 Async messages (daily) — for significant progress or blockers.
📋 Written reports (weekly) — for structured updates about key results and next steps.
🌱 1:1s (weekly or biweekly) — for deeper conversations about growth, wellbeing, and strategy.
Match information to the right format: blockers need immediate attention, status updates can be periodic, and 1:1s should focus on personal growth rather than status.
And as Anna said, to make sure your manager keeps caring, only say stuff that matters.
If you don’t have good answers, consider not saying anything at all.
Finally, updates shouldn’t just flow upward — they should create discussion. Ask feedback about how you are doing: what’s working? What feels like noise? What’s missing?
You can find Anna’s piece here, and you may also enjoy the one below about managing up 👇
Exactly one year ago we reviewed Thinking, Fast and Slow, the seminal work by Nobel Prize winner Daniel Kahneman, on behavioral science. One of the parts that stuck the most with me is about cognitive biases and heuristics: shortcuts that often serve us well, but can also lead us astray.
When it comes to bad biases, three resonated the most with my experience:
1️⃣ Anchoring bias
Our tendency to rely too heavily on the first piece of information offered.
In an experiment, participants spun a random wheel that landed on either 10 or 65, then were asked to estimate the percentage of African countries in the UN. Those who saw 10 guessed 25% on average, while those who saw 65 guessed 45%. Even though the wheel was random and participants knew it, the anchor significantly influenced their estimates.
2️⃣ Availability bias
We overestimate the probability of events that are easily remembered.
E.g. people overestimate plane crash likelihood compared to car accidents because plane crashes are more dramatic and widely reported. In tech, this leads us to overemphasize recent failures when making strategic decisions.
3️⃣ Confirmation bias
Our inclination to interpret information in ways that confirm preexisting beliefs.
In studies about capital punishment effectiveness, supporters found pro-deterrence data more convincing while opponents found anti-deterrence data more persuasive—both groups became more entrenched despite seeing the same information.
So, how do you fight biases? Kahneman says you can’t fight them in the moment: you need to fight them upfront. For scenarios requiring accurate judgment, you must design systems and procedures that counteract biases and ensure good decision-making.
When left to ourselves, improvement is possible but limited. If we want to enforce quality judgment, it’s all about process.
And that’s it for today! If you are finding this newsletter valuable, consider doing any of these:
1) 🔒 Subscribe to the full version — if you aren’t already, consider becoming a paid subscriber. 1700+ engineers and managers have joined already! Learn more about the benefits of the paid plan here.
2) 📣 Advertise with us — we are always looking for great products that we can recommend to our readers. If you are interested in reaching an audience of tech executives, decision-makers, and engineers, you may want to advertise with us 👇
