The focus system I borrowed from an engineer's blog + 18 prompts to actually move the dials

When people ask how I “get so much done,” they usually assume I have more discipline, better hacks, or some kind of magical AI setup.

I don’t.

What I do have is a mental model I borrowed from an engineer named Can Duruk, plus a handful of AI workflows that help me quietly adjust the conditions of my day instead of just grinding through it.

Duruk’s insight is simple: your ability to focus isn’t really a character trait. It’s a function of three things—how often you’re interrupted, how long it takes your brain to reload after each interruption, and how much unbroken time your work actually requires. Once you see those three variables, you can stop treating bad days as personal failures and start treating them as something you can actually reason about.

The math is a little uncomfortable. Research shows that many knowledge workers get interrupted every few minutes. If your brain needs ten or fifteen minutes to reload each time, the numbers don’t work out. No amount of willpower fixes arithmetic.

But the same math is also kind of hopeful. Small changes to those parameters can produce surprisingly large effects. One fewer interruption per hour might dramatically change how many deep work blocks you get in a week. That’s where AI starts to feel useful to me—not as a productivity boost, but as a way to nudge those variables.

Here’s what’s inside:

• The three-variable model in plain language. How interruption rate, recovery time, and block size interact to shape your day before it even starts.

• What the research actually says. The numbers from real workplaces, and why they explain a lot of the exhaustion people feel.

• How I’ve been using AI on each variable. Some workflows I’ve found helpful for reducing interruptions, reloading context faster, and reshaping work to fit fragmented time.

• 18 prompts you can use. Organized by what you’re trying to do:

  • Diagnosis & Baseline — Figure out your own λ, Δ, and θ, or audit your calendar to see what your schedule actually allows.

  • Reducing Interruptions (λ) — Build a triage system for email and messages, convert meetings to async, or audit your own self-interruption habits.

  • Faster Context Reload (Δ) — Capture your state at the end of a work block, synthesize scattered sources into a single status snapshot, design a personal reload ritual, or extract decisions from messy handwritten notes.

  • Reshaping Work (θ) — Chunk a project to fit your fragmented schedule, decompose a single overwhelming task, or match today’s tasks to your actual calendar.

  • Team-Level Focus — Audit focus capacity across a team, draft focus norms that aren’t preachy, or calculate the true cost of a meeting in focus blocks destroyed.

  • Weekly Practice — Set up a one-week experiment, run a weekly review through the λ/Δ/θ lens, or troubleshoot what went wrong on a bad day.

The full article breaks down the model, walks through how I use AI on each variable, and includes the complete prompt library. If you’ve ever gotten to 5pm and realized you spent the whole day responding to things—Slack pings, “quick questions,” meetings that could have been emails—without ever touching the work that actually matters to you, this might help explain what’s happening. And more importantly, give you something to do about it.