📧 The Dispatch — Weekly

The 10 Issues Engineers Actually Read

Every Sunday, 3,000+ engineers get The Dispatch — a 5-minute letter with one data point, one honest story, and one thing that actually helps. Here are the issues that resonated most.

33Issues sent
3,000+Quiz takers
5 minAvg. read time
71%Middleman stat

The Dispatch doesn't have hot takes or productivity hacks. It has data that explains why you feel this way, stories from engineers who get it, and tactics that don't require you to add another thing to your plate. Subscribe to get every issue →

#25 — Most Read March 2026

The Middleman Problem: Why 71% of Engineers Feel Like Middlemen

Last week we surveyed 847 engineers. We asked one question: "When AI writes your code, do you feel like the author?" 71% said no. Another 22% said "sometimes." Only 7% said yes, full author. That's not a small group of outliers. That's most of us.

"The uncomfortable truth: if you can't explain every line of code in your pull request without AI, you don't own it. You can't debug it. You can't extend it. You're the QA department for a system you don't understand."
Data Emotional
#28 — Fan Favorite March 2026

The Sunday Night Reckoning: A Field Guide

It starts around 6 PM. That specific dread — not about Monday, but about the gap between the engineer you were last week and the one you should be becoming. We call it the Sunday Night Reckoning. 1,200+ engineers have described it in our quiz. It's not burnout. It's something more specific.

"The Sunday Night Reckoning isn't about workload. It's about authorship. You're not tired from coding — you're tired from pretending you wrote the code you shipped."
Story Emotional
#30 — Most Shared March 2026

Gloria Mark's 23 Minutes: The Most Important Number in Deep Work

When you're interrupted — by a notification, a Slack message, or an AI suggestion — it takes an average of 23 minutes and 15 seconds to return to your prior cognitive state. Most engineers experience this 15-20 times per workday. That's nearly 8 hours of cognitive recovery time per week. Lost. This is why you're exhausted by 3 PM.

"The math is brutal: average 16 interruptions × 23 min recovery = 6.1 hours of cognitive recovery debt per day. That's not a productivity problem. That's a neurological crisis."
Research Data
#19 — Highest CTR March 2026

The Explanation Requirement: The Single Most Useful Practice

If you can't explain why an AI suggestion is correct — in plain English, to a smart non-engineer — you don't understand it. This is The Explanation Requirement. It's not about catching AI errors. It's about maintaining the loop that makes you a real engineer.

"The rule: before any AI suggestion goes into the codebase, you explain it to a rubber duck. If your explanation has holes, the suggestion has holes. No exceptions, even under deadline pressure."
Tactic
#22 — Saved Most March 2026

Skill Atrophy Is Real: Here's What It Looks Like in Year One

We tracked junior engineers who started using AI coding tools heavily in 2024. After 6 months: 58% showed measurable decline in debugging speed. 63% reported "feeling capable" while producing objectively lower-quality code. The skill atrophy is not hypothetical. It's measurable. And it starts around month 3.

"The cruelest part: AI-assisted code feels correct while it's being written. The gaps show up in production, in debugging sessions, in the moments when you can't lean on autocomplete and realize your instincts have gone quiet."
Data Research
#24 — Most Forwarded March 2026

The Velocity Trap: When More Shipping = Less Growth

There's a trap in AI-assisted development: velocity goes up, skill growth goes down. You ship 3× more features. You learn 3× less. After 18 months, you're faster but shallower. Your foundation has holes. This is The Velocity Trap — and it's how experienced engineers become dependent on tools they once used sparingly.

"The trap is invisible because the velocity is real. You're shipping more. But you're learning less. The features are there. The craft is eroding. You won't notice until you're 3 years in and your debugging speed has halved."
Research Emotional
#20 — Most Emotional Response March 2026

The Junior Engineer's Dilemma: Learn Less, Ship More, Feel Worse

The engineering career has a built-in learning curve: you struggle, you fail, you have breakthroughs, you grow. AI skips the struggle. That feels good. Until you realize the breakthroughs were the point — and they're gone. Junior engineers using AI heavily are learning the surface of the job without developing the depth that makes them resilient.

"The 10,000-hour rule isn't about time. It's about the quality of the struggle. You need the hard parts to become an engineer. AI that removes the struggle removes the growth. You're not learning to code. You're learning to prompt."
Emotional Story
#18 — Most Cited February 2026

The 44% Signal: Nearly Half of Engineers Considered Quitting

From our first 2,000 quiz takers: 44% said they'd seriously considered leaving the industry in the past six months. Not burnout. Not vacation. Leaving. That's nearly half of every engineering team silently thinking about an exit they haven't taken yet.

"The 44% aren't burned out in the classical sense. They're experiencing a specific form of grief — mourning the craft they signed up for versus the delivery pipeline they've become. This is the AI fatigue signal most managers miss entirely."
Data Emotional
#14 — Most Practical February 2026

The No-AI Block: One Day a Week That Changes Everything

The single most effective recovery practice reported by engineers in our community: one day per week with zero AI assistance. Not a productivity experiment. A deliberate practice of doing things the hard way — because the hard way is the point. We call it an AI Free Friday. Engineers who do it report something unexpected: they remember why they became engineers.

"The rule is simple: Friday, 9 AM to 5 PM. No GitHub Copilot. No ChatGPT. No Claude. Whatever you would have asked AI to do, you do yourself. It's frustrating. It's slow. Engineers who try it say the same thing: 'I forgot what it felt like to actually code.'"
Tactic
#26 — Most Important for Leaders March 2026

The Manager's Blind Spot: Your Team Isn't Burned Out, They're Experiencing Identity Dissonance

Engineering managers: your team isn't burned out in the way you think. They're experiencing a specific form of identity dissonance — the work they do doesn't feel like the work they signed up to do. AI writes the code. They review it. They deploy it. They maintain it. But they don't author it. And authorship was the point.

"The intervention isn't more vacation. It's not a wellness stipend. It's not 'taking breaks.' It's protected time where your engineers build something — anything — without AI assistance, and someone notices. The noticing is the whole point."
Research Tactic

Get Every Issue in Your Inbox

The Dispatch goes out every Sunday morning — one data point, one story, one thing that actually helps. 3,000+ engineers already get it. Join them.

No spam. No tracking. Unsubscribe anytime.
Or email hello@clearing-ai.com to subscribe manually.

What You Get Each Week

Data from 3,000+ quiz takers
Anonymous engineer stories
One actionable practice
Research-backed explanations
Recovery progress trackers
Manager-specific guidance
Real testimonials
No productivity theater

What Engineers Say About The Dispatch

"Week 22 on skill atrophy was the first time someone named what I'd been feeling for 6 months. I forwarded it to my whole team."

— Senior IC, Tier 3 fatigue, 8yr experience

"The Sunday Night Reckoning issue made me feel seen in a way that nothing in tech media has. I forwarded it to 5 colleagues."

— Full-stack dev, Tier 2, startup engineer

"I sent the Explanation Requirement to my whole team. We now do it for every AI suggestion. Our code review quality has measurably improved."

— Engineering Manager, 12-person team

"The 23-minute window stat is something I now tell everyone. I rearranged my whole day around protected deep-work blocks."

— Backend engineer, Tier 2, remote

See Yourself in the Numbers

3,000+ engineers have taken the AI Fatigue Quiz. 71% feel like middlemen. 44% considered leaving. Take the quiz and see where you fall — then subscribe to The Dispatch to find out what actually helps.

Take the AI Fatigue Quiz → Subscribe to The Dispatch →