Harmonically

AI drafts your email in ten seconds. It summarises the forty-page report before your coffee cools. It writes the first version of the proposal while you’re still in a meeting.

By any reasonable measure, you should have more time than you did two years ago.

So why are you more tired?

If that question lands, you’re not imagining it. And you’re not doing AI wrong. You’re doing what almost everyone does with time that’s handed back to them: filling it straight back up with more output.

This post is about the one decision that determines whether AI makes your life better or just makes it faster — what you do with the time it saves.

The Promise: Work Smart, Not Hard

The core message of the AI era is genuinely good news. You no longer need to grind through every task by hand. Research on generative AI in professional writing tasks has found meaningful reductions in time spent, often without any drop in quality. The hard, repetitive parts of knowledge work are becoming cheaper by the month.

That means the old equation — more hours equals more results — is breaking down. Working smart with AI beats working hard without it.

But “working smart” has a second half that almost nobody talks about. Saving the time is the easy part. Keeping it is the hard part.

What Actually Happens to AI-Saved Time

In February 2026, researchers from UC Berkeley’s Haas School of Business published findings in Harvard Business Review from an eight-month study of around 200 employees at a US tech company. Their headline: AI didn’t reduce work. It intensified it.

What makes the finding striking is that AI use at the company was voluntary. Nobody piled on extra work. Employees filled the freed-up capacity themselves — working faster, taking on a broader range of tasks, and letting work spill into more hours of the day, often without being asked.

The researchers described a new rhythm: managing several threads at once, checking AI outputs constantly, reviving long-deferred tasks because the AI could “handle them,” and ending up with more open loops than before. It felt like momentum. It was really continuous attention switching.

That’s the trap. AI hands you a time dividend — and your default setting reinvests it, automatically, in more work.

Why This Is a Nervous System Problem, Not a Discipline Problem

It’s tempting to read this as a willpower issue: just stop taking on more. But the mechanism runs deeper than discipline.

1. AI shifts your work from doing to judging. When AI does the drafting, your job becomes reviewing, deciding and course-correcting. That’s the most cognitively expensive work there is. You’ve traded low-effort production for high-effort evaluation — and evaluation is exactly the kind of work that keeps your stress response switched on.

2. Open loops block detachment. Sabine Sonnentag’s research on recovery shows that psychological detachment — genuinely switching off from work — is one of the strongest predictors of recovering well. Every AI thread you leave running in the background is a loop your brain keeps monitoring. More threads, less detachment.

3. Tired brains make worse calls. Sleep research from UC Berkeley found that sleep-deprived participants showed markedly amplified amygdala reactivity with weaker regulation from the prefrontal cortex. In plain terms: under-recovered brains are more reactive and less discerning. That’s a serious problem when your main job has become judging AI output.

4. More hours still don’t equal more results. John Pencavel’s research on working hours found that output per hour falls sharply beyond roughly 50 hours a week. AI doesn’t repeal that curve — it just gets you to the flat part faster.

Put together: in the AI era, the scarce resource isn’t your typing speed. It’s the quality of your attention and judgment. And those are built by recovery.

The Reframe: Not Resting Is the Least Productive Thing You Can Do

Here’s the belief that keeps high achievers stuck: “If AI saved me an hour, the smart thing is to use that hour to get ahead.”

It sounds rational. But it treats your brain like a machine with unlimited uptime. No factory runs that way. Machines get scheduled maintenance precisely because running them flat out, without downtime, ends in expensive breakdown. Burnout is that breakdown.

So let’s flip it. Getting enough rest isn’t the opposite of working smart. It’s the part of working smart that AI can’t do for you. AI can draft, summarise and calculate. It cannot recover your nervous system, restore your focus, or bring back your creativity. Only rest does that.

That’s the real opportunity of this moment. For the first time, many high performers have time appearing in their day that didn’t exist before. The question is whether it goes back into the machine — or back into you.

How to Reinvest Your AI Time Dividend in Rest

You don’t need to find new time. AI is already creating it. You just need a system so it doesn’t quietly disappear.

1. Notice the dividend. For one week, jot down every time AI noticeably speeds up a task and roughly how much time it saved. Most people are surprised — both by how much time appears, and by how little of it they can account for afterwards.

2. Decide the default before the time appears. Time without a plan gets absorbed by whatever is loudest, and work is always loudest. Set a simple rule in advance: “When AI finishes something early, I will…” — and fill in a specific recovery action.

3. Ring-fence a share of it. You don’t have to give all of it back to yourself. Decide on a share — a third, half — and treat it as non-negotiable, like planned maintenance on the calendar.

4. Match the rest to the size of the dividend.

• 3 minutes: a few slow exhales, eyes off the screen, looking at something far away.
• 15 minutes: a walk outside without your phone.
• An hour: something that genuinely nurtures you — time with someone you love, a hobby with no goal, being in nature.
• Half a day or more: a proper zero-output block, where nothing is expected of you. (If you haven’t yet, read [how to take a rest day without guilt].)

5. Don’t spend “AI waiting time” on more screens. While an agent runs or a long output generates, the reflex is to open another tab. Instead, treat it as a micro-gap for recovery — stand up, stretch, breathe. Watching a progress bar while switching between three other threads is not rest.

6. Cap your open loops. Decide how many AI threads you’ll run in parallel, and close them before you end your day. A short shutdown ritual — writing down what’s still open and where you’ll pick it up — tells your brain it’s allowed to let go.

Beyond Recovery: Using the Dividend to Enrich Your Life

Rest isn’t only about getting back to baseline. Used well, the time AI gives back can do something bigger: make your life feel like yours again.

Research by Ryan, Bernstein and Brown found that people feel better on weekends largely because they feel more autonomous and more connected to others — not simply because they’re away from work. That’s a clue for where your dividend creates the most value: in time you choose freely, and time spent with people who matter to you.

And when you do get those moments, stay with them. Psychologists Fred Bryant and Joseph Veroff showed that savouring — deliberately noticing and lingering on good experiences — amplifies their positive effect. A reinvested hour you actually savour is worth far more than a “free” hour you spend half-checking your inbox.

A Quick Word of Caution

If you’re already exhausted, sleeping badly or noticing physical symptoms that don’t improve with rest, please talk to a doctor. Reinvesting time in rest helps a lot, but it doesn’t replace proper care.

Frequently Asked Questions

Is AI causing burnout? Not on its own. But early research suggests that without boundaries, AI can intensify work — more tasks, more multitasking, longer hours. The burnout risk comes from how the saved time gets used.

Shouldn’t I use the time AI saves to get ahead? Some of it, sure. But if all of it goes back into output, your judgment and focus — the skills AI makes most valuable — will quietly degrade. Recovery is how you stay ahead.

How much AI-saved time should go to rest? There’s no magic number. Start with a share you can protect consistently, such as a third or half, and adjust from there.

What if I can’t tell how much time AI is saving me? Track it for one week. Even rough estimates make the dividend visible — and what’s visible can be protected.

Your Next Step

Working smart with AI is only half the equation. The other half is making sure the time it frees actually comes back to you.

The AI Time Dividend workbook assists with a one-week time audit, establishes a reinvestment rule, presents a rest menu according to available time, includes guidelines to avoid AI overload, and features a four-week tracker. Get the workbook here.