The two-minute rule in the age of AI

What happens when AI makes ten-minute tasks take ten seconds?

One of the most useful ideas in productivity is almost embarrassingly simple. If a task takes less than two minutes, do it now. Don't capture it, don't defer it, don't add it to a list. Just do it. The overhead of tracking it will cost more than the task itself.

David Allen called this the two-minute rule in Getting Things Done, and people who apply it consistently report something close to magic. Within a few hours of ruthlessly doing every tiny task the moment it arises, the ambient weight of all those little obligations starts to lift. The inbox gets lighter. The to-do list gets shorter. The background hum of "I should really..." quiets down.

The logic is economic. Every task you defer has a carrying cost: capturing it, filing it, reviewing it, remembering it, and eventually doing it anyway. For anything that takes less than two minutes, the carrying cost exceeds the cost of just doing the thing. So you do it.

But here's what's interesting. Allen set the threshold at two minutes in 2001. Back then, the time certain tasks took was relatively fixed. Finding a specific file on your computer: five minutes minimum, depending on how messy your folders were. Summarising a document: twenty minutes if you were fast. Organising research across multiple sources: an hour, easily. Writing a follow-up email after a meeting: fifteen minutes to compose something decent.

Those time estimates have changed. Dramatically.

The shifting threshold

AI hasn't just made some tasks faster. It's moved entire categories of work from one side of the two-minute line to the other.

Summarising a ten-page document used to be a twenty-minute task. Now you paste it into an AI assistant and get a summary in seconds. That's not a faster version of the same task. It's a different task entirely, one that now falls well below the two-minute threshold, which means under Allen's own logic you should do it immediately rather than defer it.

Finding a specific piece of information across your files, notes, and saved articles used to require remembering where you put it, navigating folder hierarchies, and scanning through documents. Five to fifteen minutes, sometimes more. With semantic search that understands what you're looking for rather than matching exact keywords, it takes seconds.

Drafting a first pass at an email, a meeting summary, a brief: these used to be tasks you'd put on a list and get to later because they required enough thought and composition time to feel like real work. Now an AI draft takes seconds to generate, and your job is to edit and approve rather than write from scratch. The creative and strategic thinking still takes as long as it ever did. But the mechanical parts, the assembly of words, the formatting, the initial structure, those have collapsed.

The result is that the two-minute rule now applies to a much larger portion of your task list than Allen ever imagined.

The decision tax

This matters more than it might seem, because of how decisions work in the brain.

Roy Baumeister's research on what he called ego depletion found that every act of self-regulation, including every decision you make, draws from the same limited pool of mental energy. A small decision and a big decision deplete the same resource. Choosing what to have for lunch and choosing whether to accept a job offer both cost you something. The small decisions cost less individually, but there are hundreds of them per day, and they add up.

Every task that sits on your to-do list represents at least one future decision: when will I do this? Some represent several: when will I do this, how will I do this, do I even still need to do this? The longer a task sits there, the more decisions it generates. You see it in your morning review and decide "not now." You see it again at lunch and decide "later." You see it the next day and decide "tomorrow." Each of those micro-decisions costs something, even though you feel like you didn't do anything.

This is why people who adopt the two-minute rule feel lighter almost immediately. It's not just that they've completed tasks. They've eliminated decisions. Every tiny task you do immediately is a task you'll never have to decide about again.

Now extend that logic to AI. If AI can handle the filing, the tagging, the categorising, the summarising, the searching, then each of those is a decision you no longer make. You don't decide where to file a note because AI auto-tags it. You don't decide when to summarise a document because it takes five seconds so you do it now. You don't decide when to search for something because searching is effortless. The decisions evaporate, and with them, a significant chunk of daily cognitive load.

Voltage preservation

There's a connection here to something Fernando Borretti describes in his writing about ADHD, which he calls the voltage model of energy. His argument is that mental energy isn't like a battery that drains evenly. It's like voltage, where different tasks require different thresholds, and once you drop below a threshold, certain tasks become impossible no matter how much total energy you have left.

Small administrative tasks, the filing, the tagging, the "where does this go" decisions, are individually low-voltage. But collectively they erode your voltage throughout the day. Every small decision is a micro-drain. By the time you get to the creative work, the strategic thinking, the hard problems that actually need your full capacity, you've already lost ground to a hundred tiny obligations.

If AI handles those micro-tasks automatically, the voltage stays higher for longer. You arrive at the hard problems with more capacity. The filing happened in the background. The summaries were generated on capture. The search takes seconds instead of minutes. None of these individually feel like a big deal, but the cumulative effect of not making dozens of small decisions per day is significant.

What this actually looks like

In practice, this plays out across your whole workday in ways that are easy to miss because each individual time saving seems trivial.

You save an article from the web. Previously: bookmark it, decide which folder, tag it, maybe add a note about why you saved it. Now: clip it, done. AI tags and categorises it. Time saved: two minutes, plus one decision eliminated.

You finish a meeting. Previously: review your notes, write up a summary, identify action items, file the notes somewhere. Now: the recording is transcribed automatically, AI generates a summary and action items. You review and edit. Time saved: ten to fifteen minutes, plus several decisions eliminated.

You need to find something you saved weeks ago but can't remember where. Previously: open three apps, search by keyword, scan through results, hope you tagged it correctly. Now: describe what you're looking for in natural language, get the result. Time saved: five minutes, plus the frustration tax of not finding it.

None of these is transformative on its own. But across a full day, you might save an hour of fragmented time and eliminate dozens of micro-decisions. That's an hour of cognitive capacity returned to you for the work that actually needs a human brain.

The new rule

Allen's two-minute rule was about doing small things immediately because the overhead of deferring them costs more than doing them. That logic hasn't changed. What's changed is what counts as small.

Tasks that used to take ten or twenty minutes now take seconds. Decisions that used to require your judgment can now be handled automatically. The two-minute line has shifted, and a huge amount of work that used to live on the "to process later" pile now falls below it.

The new version of the rule might be something like: if AI can do it in seconds, let it happen now. If AI can do it automatically, let it happen in the background. Save your decisions, your judgment, and your creative energy for the things that genuinely need them.

Allen was right that your brain is for having ideas, not holding them. The update is that your brain is also for making important decisions, not trivial ones. Every trivial decision you can eliminate or automate is energy preserved for the decisions that matter.

Inspired by David Allen's Getting Things Done.