Before you delegate

Before you delegate
7 minutes by Wes Kao

Delegating is hard when explaining a task feels slower than just doing it yourself. Wes shares a simple checklist of six questions that makes it easier: understand what the person already knows, share the why, gather needed resources, show what good looks like, set a clear timeline, and flag likely risks. These questions help build shared context fast and improve the quality of work handed off.

[Webinar] Measuring AI developer productivity: Can you prove AI is working?
sponsored by Unblocked

AI is in your engineering workflow. While the token spend shows it, the throughput doesn't. The human is very much still in the loop, and that's a context problem. Join live on Aug 19 (FREE) to learn how to track AI productivity, how context maturity affects where teams stall, and what it takes to get real value from your agents.

Overpromising
6 minutes by Stay SaaSy

Promising career rewards to keep team members happy often backfires badly. The initial boost fades fast, and if the promise falls through, trust is shattered and can take years to rebuild. Managers should only make promises they can fully guarantee, and caveat anything uncertain with honest disclaimers. Focus on what you can control, not what you hope will happen.

Engineering management after the cost of code collapsed
13 minutes by Karim Jedda

Engineering management rules aren't uniformly obsolete or valid in the AI era. The key is checking what assumption each practice rests on. If a rule depends on the cost of writing code, revisit it. If it depends on how humans coordinate, verify, or accept risk, it still holds. Code generation got cheap, but specification, judgment, and ownership did not. Whoever has to sign a decision still sets the pace.

When nothing happens
12 minutes by Mike Fisher

Organizations don't punish bad behavior — they punish bad behavior when results turn. Research shows that whether a harsh leader is seen as inspiring or abusive depends mostly on whether the team is winning. Cases like Steve Jobs, Linda Wachner, and Bobby Knight all follow the same pattern: the behavior stays constant, but the moment results drop, tolerance evaporates. The real question for any leader is how much cover strong performance is currently buying for someone on their team.

The lights were never off
14 minutes by Ilko Kacharov

The idea of the dark factory promises software development with little or no human involvement. But the teams closest to this work tell a more nuanced story. Humans still review, verify, and guide AI-generated code. The real challenge is not removing people, but replacing the work reviewers do with strong tests, simulations, automation, and safeguards.

And the most popular article from the last issue was:

• A practical guide to fast onboarding