Revised rules of engineering leadership

Revised rules of engineering leadership
10 minutes by Will Larson

AI tools have made individual engineers far more productive, letting small teams complete in weeks what used to take years. But good tooling alone isn't enough. Strong development practices, durable teams with deep domain knowledge, and fast executive decision-making still determine how much of that speed translates into real progress. The bottlenecks have shifted, but the fundamentals of good engineering leadership have not.

Your teams may slog through context rot. Your agents do not.
sponsored by Falconer

Your teams know which docs, Slack threads, and old decisions to take with a grain of salt. Agents don't have that ability. Falconer gives you a live Knowledge Health score, showing what's trustworthy or drifting, and helps fix the context your teams and agents rely on. Get your free report.

AI demands more engineering discipline. Not less
18 minutes by Charity Majors

Charity argues that AI makes code generation easier, but increases the need for engineering discipline. As code becomes cheap and disposable, teams must focus more on testing, validation, observability, and system reliability. The future of software engineering depends on understanding and managing systems, not just writing code.

Here’s the rub: We don’t believe you
10 minutes by Michael Lopp

New senior leaders should quickly build communication structures: regular one-on-ones with key people, a staff meeting, an extended staff meeting including all managers, and an all-hands presentation. These help uncover hidden problems the team cannot see. Expect resistance when sharing findings, as those closest to the work often cannot recognize what is broken. Fresh perspective is a short-lived advantage, so use it fast.

The reliable team trap
16 minutes by Kevin Goldsmith

Doing good work reliably builds trust, but that trust only earns you more work, not more influence. Organizations reward execution and strategic standing through separate mechanisms, and being great at delivery can actually lock you out of bigger decisions. To shift how you are seen, Kevin suggests to make your judgment visible, build leaders under you to free up capacity, connect your team's output to real business outcomes, and set your own direction before your boss does it for you. Work alone does not speak for itself.

AI made the boring work visible. Cut the work, not the people.
8 minutes by Jeff Gothelf

AI is automating millions of jobs, and companies like Coinbase, Meta, and Cisco are cutting staff as a result. But IKEA shows a smarter path: when AI handled 47% of customer calls, they studied the rest, found design conversations humans handled best, retrained 8,500 workers as remote interior design consultants, and made €1.3 billion. Layoffs are costly and teach remaining staff to hide productivity gains. The better question is where freed-up people can do work AI cannot.

And the most popular article from the last issue was:

• Doing nothing at work