Your Fridge Door Is Open. Stop Pushing It.

Last updated: 2026-09-15

You know that fridge that keeps drifting open? You walk by, give it a shove, and it drifts open again. Most people treat that as a willpower problem. Push harder. Check more often. Get annoyed at yourself.

I know. A fridge, on a productivity site. Stay with me.

The seal is worn. Replace it and the door stays shut. You stop thinking about it.

I keep coming back to that image when I look at work and life. When something keeps going wrong, the instinct is to try harder. Work later. Be more disciplined. That’s a behavioral patch. The thing that actually works is finding the broken seal in the system.

Brooks and I frame productivity through three layers: personal systems for showing up well, work systems for getting things done, and strategic systems for staying aligned with the right goals over time. If one layer is leaking, you feel it in the others.

Make the good version of you repeatable

A personal system is about making the best version of yourself more repeatable. Sleep. Exercise. Food. Relationships. Starting the day well.

I treat my morning routine as a personal operating system. I control the first few hours with choices like a high-protein breakfast, then I let the rest of the day handle surprises more flexibly. I don’t try to script everything. I just want a launchpad for energy and focus.

I even turned breakfast into a feedback system. Morning meals that left me alert and then crashing were easy to miss until I started paying attention to the pattern. I’m no doctor. That’s my observation. It still changed breakfast, because the pattern showed up clearly instead of a vague “I feel off.”

Brooks made exercise the first thing he does after waking up, with the workout determined by the day. No morning negotiation. The trigger is waking up. The action is already decided.

Does this sound familiar? You reopen the same debate every morning. Remove one daily decision. Pre-decide the trigger and the action.

Shrink the list, then ask “how can I AI this?”

My current question for every work system is: “How can I AI this?” I’ve been experimenting with AI for email triage, weekly reviews, to-do list prioritization, meeting notes, and follow-up drafts.

I limit my daily to-do list to three priorities. That forces real prioritization. Otherwise the task list becomes a disguised wish list, and you spend the day feeling behind on work that was never going to happen today. Plan tomorrow with three must-do items.

Brooks starts with a morning review. He looks at yesterday, checks the calendar, decides what matters today, flags tasks in OmniFocus, then works from a focused list instead of the full task database. Clarity first. Then execution.

Email is one of the highest-leverage places for AI because so much of inbox management is rule-based. Classify the message. Decide whether it needs a response. Draft or route it. Archive what does not matter. I’ve been experimenting with Lindy on this. Design the workflow first. If you can write the rules a human already follows, you can usually hand part of that loop to AI.

Meetings are messier.

I had an AI meeting workflow where a notetaker attended a meeting I wasn’t even on, then sent a Slack summary with decisions, sentiment, next steps, and a drafted follow-up email ready in Gmail. You stop reconstructing the meeting from memory.

Brooks predicts AI notetakers may become the next workplace etiquette battleground. Useful for you. Annoying when every meeting suddenly has several bots asking to join. Privacy, consent, and convenience don’t always agree.

My current workaround for bot-heavy meetings is experimenting with local notetakers like Granola, which record from the computer instead of joining as another visible participant. Still get consent. Accuracy is imperfect. Multi-speaker identification is still messy.

I’ve seen teams scale review feedback by turning their criteria into a custom GPT that people could use before submitting work. Less repetitive coaching. Faster final reviews.

Brooks saw product specs getting rejected by development until product managers ran them through an AI checker against the same criteria before handoff. Issues got caught earlier.

That’s his systems-thinking model: input, process, output. When something is not working, don’t randomly tweak the process. First ask whether the real problem is what is entering the system, what happens inside it, or what result it produces.

The systems that don’t scream

The weekly review is probably the single most impactful system for most people. It creates regular clarity before problems snowball.

Strategic systems are the slower, less urgent ones that keep life and work pointed in the right direction: weekly reviews, thinking time, monthly planning, quarterly planning, reflection, and goal setting. They rarely feel urgent. That’s why they disappear.

Brooks’s warning is that life may not fall apart immediately. You can spend a year paddling hard without actually moving forward.

Are you living in alignment, or just paddling?

My preferred goal-setting cadence is 90 days. Long enough to make meaningful progress. Short enough to create urgency and keep the action plan connected to reality.

If you’ve been busy and still feel directionless, restart the weekly review before you add another tool.

One tweak. This week.

Don’t overhaul all three layers at once.

Choose one personal, work, or strategic system and make one small tweak this week. One tweak a week compounds into massive productivity gains.

Maybe it’s three priorities tomorrow. Maybe it’s asking “how can I AI this?” of one inbox rule, one meeting note, or one handoff checklist. Maybe it’s putting the weekly review back on the calendar.

Which seal are you going to fix?