Decide the night before, and use one fixed rule to break ties: pick the task that moves your most important project forward, not the one that feels most urgent. That single pre-commitment kills the morning debate that quietly eats your first, sharpest hour.
Here's the trap. You sit down ready to focus, open your list, and spend fifteen or twenty minutes weighing options. Which task? Is now the right time? Should I clear the small stuff first? By the time you pick, your energy has leaked out through a hundred tiny comparisons. The problem was never focus. It was the choosing.
Why choosing is the real bottleneck
Deep work fails less from distraction and more from decision fatigue right at the threshold. Every unmade decision you carry into your focus block gets made during the block, when it costs the most. You're not lazy. You're spending willpower on selection instead of execution.
So move the decision earlier, make it smaller, make it repeatable. When the choice is already made, starting is almost boring. And boring is exactly what you want at the start of deep work.
The playbook: a repeatable system for picking one task
This is the loop to run most days. Five minutes the evening before, zero in the morning.
Step 1: Close the day by naming tomorrow's one task
Before you shut your laptop, write down the single task that gets your first deep work block tomorrow. Not three tasks. One. If you can't name it in a sentence, it's not defined enough to do. It's a project pretending to be a task.
Too vague: "Work on the proposal." Decided: "Draft the pricing section of the Acme proposal, three options." The second one you can start without thinking.
Step 2: Use one tie-breaker rule, always the same
When two tasks compete, don't re-argue the merits every time. Apply a rule you picked in advance. One example: the task that moves your most important project, even if it's uncomfortable, beats the task that's merely due soon.
That rule doesn't have to be your rule. The point is that any consistent rule ends the debate. You stop negotiating with yourself every morning.
Step 3: Shrink the entry point
Define the first physical action, not the whole task. "Open the doc and write the section header" is a starting move. "Finish the proposal" is a mountain. Decision fatigue loves mountains. Give yourself a doorway instead.
Step 4: Let AI do the pre-sorting, not the deciding
Here's where AI earns its place, and where most people misuse it. Don't ask it to "plan my day" and hope it decides for you. Ask it to cut your options down to a shortlist so you can make one clean call.
The approach: paste tomorrow's candidate tasks and ask the model to group them, flag which one advances your priority project, and name the smallest first action for the top pick. You're using AI as a filter, not an oracle. The final choice stays yours. The deliberation is what you're handing off.
The trick is asking the same way every time, so the output is consistent and you're not designing a new prompt at 9pm. That repeatability is the whole game. If you'd rather skip the trial and error of building and refining those prompts yourself, the Deep Work Day Designer AI Prompt Pack includes tested prompts for exactly this kind of pre-sorting and one-task selection, so you can run the routine from day one.
Step 5: Protect the decision from your morning self
Your morning self is optimistic, distractible, and quick to reopen settled questions. Don't let it. Write the chosen task somewhere you'll see it before your inbox. A sticky note, the top of your doc, a calendar block title. When you sit down, the decision greets you already made.
What to do when the plan breaks
It will break. A fire lands, a meeting moves, your energy tanks. The rule: you can swap the task, but you can't re-run the whole selection. Pick the next item on last night's shortlist. One substitution, no fresh debate. That keeps a bad morning from turning into a paralyzed one.
When you genuinely have no priority
Sometimes nothing feels important. That's usually a sign your projects aren't defined, not that you're out of work. On those days, do the smallest task that removes future decisions. Clean up your project list, write clearer next-actions. You're not doing deep work. You're clearing tomorrow's fog.
Why this beats a longer, smarter system
Every extra step you add to "choose my task" is a new place to stall. Aim for fewer choices, not a slicker dashboard. A five-minute nightly close, one tie-breaker rule, and a shrunk entry point will beat any elaborate framework you have to re-learn each week. Simple survives contact with a tired brain. Clever usually doesn't.
FAQ
How long should choosing my deep work task actually take?
Decide the night before and the morning cost is zero. The evening decision itself should take about five minutes. If it runs longer, your tasks are too vague. Tighten them into single, startable actions.
Can't I just ask AI to pick the task for me?
You can, but it's worth steering away from that. Hand the final call to a model and it chips away at your sense of ownership, and you'll second-guess it anyway. Use AI to shortlist and pre-sort. Keep the one-line decision human. That split is what actually cuts the fatigue.
What if my priorities change mid-week?
Update the shortlist during your evening close, not mid-morning. Keep one weekly review to reset priorities and let daily decisions stay small. Changing direction is fine. Changing it live, every day, is where the fatigue creeps back.
Do I need special tools to do this?
No. A notebook and a consistent habit cover most of it. AI helps when your candidate list is long or messy and you want the sorting done fast, but the system works on paper alone. Tools should reduce decisions, never add them.