Every efficiency wave for a century promised to give us time and delivered more work. AI is the largest such wave yet, and by default it will go the same way. The five moves are how you change the default. They are tactics, but underneath them is one stance: the point of becoming more efficient was never to produce more. It was to become more free.

The order matters as much as the moves, so take them in sequence.

Move one: See

You cannot reclaim time you cannot see. Your memory of your week is unreliable in specific, predictable directions: you underestimate how long things take and overestimate how much you worked on what mattered.1 So the first move is measurement, an honest audit of where the hours actually go. Not where you think they go. Where they go. Everything else optimises whatever this reveals, which is why guessing here corrupts every move that follows.

Move two: Shed

The most natural thing in the world is to point your new AI tools at your pile of low-value work and start doing it faster. Resist that with everything you have. Before you automate anything, ask what it is actually for, and if the honest answer is “nothing,” delete it. Eliminate, then automate, never the other way around, because automating a task is the most expensive way to discover you should have removed it.2 Your week is a monument to addition; this move is the subtraction your mind keeps skipping.

Move three: Shift

Now, and only now, hand the right survivors to the machine. The skill is not using AI but delegating to it: directing it like a capable new hire, giving it context and examples, and checking its work before it goes out under your name.3 Delegate the verifiable, the shallow, and the supporting work. Keep your core craft, the differentiated work you are actually paid for and that keeps your skills sharp.

Move four: Shield

This is the move the whole method turns on. Every previous step creates a vacuum, and a vacuum in your working life does not stay empty. Parkinson’s Law, the Jevons rebound, and the always-on culture will refill it automatically unless you defend it with structure.4 Willpower fails under pressure. Commitment devices do not: hard time boundaries, an output cap, reactive work herded into named windows. A block others can freely schedule over is not a boundary, it is a suggestion.

Move five: Spend

Reclaimed time is worth nothing until you decide what it is for, and if you do not decide, it will be decided for you, by guilt and the pull of the inbox. The strongest evidence in the whole method points one way: relationships first, then active engagement that produces flow, over passive consumption that hands the time back hollowed out.5 This is the reason all the other moves exist.

Why this works when willpower does not

Each move is structural, not heroic. See replaces guesswork with data. Shed removes the work rather than speeding it up. Shift offloads without abdicating. Shield builds walls that hold when you are tired. Spend gives the time a destination so it stops being captured by default. Run them in order and the time AI saves finally reaches you, instead of leaking back into the work, like every windfall before it.

Frequently asked questions

What are the five moves to reclaim your time?

See (audit your real week), Shed (kill the work that should not exist), Shift (delegate the right tasks to AI), Shield (defend the reclaimed time with structure), and Spend (reinvest the hours deliberately rather than letting them refill with work).

Why does the order of the five moves matter?

Because eliminating must come before automating, and defending must come before the time can be spent. Automating work you should have deleted is the most expensive mistake, and reclaimed time that is not defended refills automatically.

Can you use the five moves without a four-day week policy?

Yes. The method is designed for individuals. It does not require schedule control or company permission, though those make it easier. Each move is something you can start on Monday.

About the author

Tom Goodwin

Tom Goodwin is the author of Don’t Work Harder, a book about taking the time AI gives back as time rather than more work. He is a co-founder of GAMEPLAN and writes on productivity, technology, and the economics of the working week.

Footnotes

Footnotes

  1. On the planning fallacy and the overestimation of hours worked, see Daniel Kahneman, Thinking, Fast and Slow (2011), and time-diary research summarised in Chapter 5 of Don’t Work Harder. ↩

  2. Peter Drucker, “Managing for Business Effectiveness,” Harvard Business Review (May 1963); and “the algorithm” in Walter Isaacson, Elon Musk (2023), which places automation last. ↩

  3. Ethan Mollick, Co-Intelligence: Living and Working with AI (2024): “delegate, don’t abdicate.” ↩

  4. On commitment devices, see Jon Elster, Ulysses and the Sirens (1979), and Richard Thaler and Cass Sunstein, Nudge (2008). ↩

  5. Harvard Study of Adult Development; Robert Waldinger and Marc Schulz, The Good Life (2023): relationship quality is the strongest predictor of long-term health and happiness. ↩