If your team is happy, are you doing a good job?

If your team is happy, are you doing a good job?
11 minutes by Charity Majors

Teams work best when people feel safe enough to take risks and learn together, but somewhere along the way "psychological safety" got confused with "never feeling bad." Good leadership means challenging people and saying no, not just keeping them happy. We tend to blame or praise individuals for outcomes that are really shaped by team structures and systems. Learning is uncomfortable, slow, and hard to measure, but that is exactly what makes it valuable.

Stop guessing whether your AI works. Train your team to build AI evaluations in two weeks.
sponsored by AE Studio

AE Studio collaborates with Anthropic on AI research. We advise companies including Samsung, Walmart and Berkshire Hathaway. In two weeks, your team learns to define success, uncover failures, and guide development with AI evaluations. We work on your own system. You keep every suite we build. See how it works.

The second messenger
10 minutes by Mike Fisher

Ignaz Semmelweis had overwhelming evidence that unwashed hands spread deadly infection, yet died ignored. Joseph Lister had weaker data but transformed medicine. Mike argues the difference was craft: Lister published quickly, explained the mechanism behind his findings, demonstrated his method publicly, recruited allies, and admitted mistakes openly. Good evidence alone rarely wins. To change minds, Mike suggests you need to make your findings reproducible, find people who will carry the idea, and give others a way to be part of the solution rather than the problem.

Speeding up code review with "small-change"
7 minutes by Will Larson

Engineering teams often slow down as PR volume grows, since more code means longer review times. To tackle this, a "small-change" system was built to auto-approve low-risk PRs that meet strict size, quality, and author-experience checks. Will points out that early results show about 10% of PRs qualify, getting fast approvals without human bottlenecks. Success depends on expanding that percentage while strengthening automated verification over time.

Irrational decisions
9 minutes by David Pereira

According to David, safe decisions keep you where you are. Taking chances others call crazy or stupid is often what leads to real growth. After quitting a well-paid CEO role with nothing lined up, David opened up a new path that brought more fulfillment and income than before. The difference between stupid and crazy is simple: stupid means committing to something without trying to figure it out; crazy means committing and then figuring it out.

Nine questions I ask in interviews that I wish I'd asked years ago
8 minutes by Louise Deason

Most candidates treat job interviews as one-sided exams, but the question period at the end is a powerful and overlooked tool. Asking things like how promotions actually happen, what the team is genuinely bad at, or why the role is open can reveal what companies never put on their careers page. Pay attention to hesitations and vague answers as much as the words themselves. A company that resists being questioned has already answered your most important question.

industry

• Google cloud outage takes down HSBC, Revolut, Bank of Scotland, Lloyds, Equifax, more
• 30 months on the exponential (in charts)
• Anthropic secured highest junk credit rating in June
• How the Claude Code team does code review

security

• Patch the Planet: global response coordination as agents uncover critical vulnerabilities
• PHP2Shell — AI uncovers critical vulnerabilities in the backbone of the web
• Eight Bits of Trust; eMudhra's CA keys cracked in milliseconds
• How abliterated models can get you pwned
• France's supermarket fridges were vulnerable for months

And the most popular article from the last issue was:

• How to be more ambitious