It wasn’t at a conference on AI that I heard the truest sentence about AI in recent weeks. It was on a podcast about rest, listening to a coach talk about ten days on retreat in the Jura mountains.

She wasn’t talking about models, or productivity, or transformation. She was talking about everything that doing nothing gives us: rest, silence, the ideas that surface when the mind is finally allowed to settle. And along the way, in two sentences, she said what I’ve been defending for a long time about the place of AI.

In short

  • AI can relieve what weighs heaviest without ever being seen: the mental load, that endless list of things not to forget that occupies the mind at all times.
  • A mind that is relieved recovers what makes its value: its core strengths, its presence to others, its creative capacity. An AI that relieves rather than replaces: a tool that gives us back to ourselves.
  • Working better also means performing better: fewer errors, fewer departures, more loyal customers. Provided the use of the time given back is a choice, not an automatism.

A sentence heard where you’d least expect it

The episode is titled How not to feel guilty about resting. Angélique Robin, a personal-development coach, describes what ten days on retreat gave back to her. She unfolds the benefits of doing nothing: the mind at rest, the brain’s default mode working for us while we think we’re idle, the clarity and the ideas that return when we make room for emptiness.

And then, a little before the thirteenth minute, she mentions the arrangement that lets her hold this balance — leaning on (my translation from the French):

“[…] a digital ecosystem with AI to gain time for life on tasks that weigh on us, and to work far more often in our zone of genius.”

A few seconds later, she adds:

“AI and an ecosystem at our service let us build a life less full of all the truly heavy tasks that take up our time.”

I stopped on those two sentences. First because of what they say. Then because of who says them.

What these two sentences say

Read the words again — they are chosen with a precision that many technology speeches could envy.

Time for life. Not work time, not productivity, not efficiency. Time for life.

Tasks that weigh on us. Not useless tasks, nor even simple ones. Tasks that weigh — that is, that clutter the mind well beyond the time they take.

An AI at our service. The tool is in its place: at our service. Not at the controls, not in competition. Though we still have to ask whose service it is also in.

Working in your zone of genius. What each person does best, what they are there for, what looks like them.

In two sentences, it’s all there. AI isn’t there to save us time so we can do ever more. It’s there to let us devote more time to what genuinely deserves our attention.

This isn’t a naïve vision. It’s a humanist vision — not of the machine, but of the place we give it.

What I mean by a “humanist vision”

  • It doesn’t describe the machine. An AI isn’t humanist; it’s the choices of use and of organisation that can be.
  • It keeps the human at the centre of the decision. AI proposes, prepares, relieves. The human reviews, validates, decides — and answers for it.
  • It judges AI by what it gives back to people: time for life, attention, better conditions to exercise their core strengths.

And yet we must keep our eyes open. AI, in itself, is a tool. But to exist, it needs a technology, and technology is never neutral. It always belongs to someone, it generates a financial windfall, and therefore interests. In a geopolitically unstable world, where these tools are concentrated in a few hands, sovereignty is no detail.

For all that, we must not give up looking at AI through the lens of its uses and the services it renders. That is, in fact, the only way not to let others decide in our place what it should be.

These two sentences are not a study. They are a signal — the very one that led me to write this text. The signal that AI no longer belongs to tech alone.

That this vision comes from a world so far from the digital one is not incidental. Public debate on AI has been captured by two camps: those who sell it and those who fear it. Meanwhile, a coach who is anything but a tech enthusiast has made calm, pragmatic use of it, and speaks of it with an accuracy that ought to give us pause. She isn’t trying to be augmented. She’s trying to be relieved.

What really weighs: the mental load

The word “weigh” is worth pausing on, because it points to something precise.

In 1984, the French sociologist Monique Haicault coined the notion of charge mentale — the mental load — by observing women who juggled a job and the running of a household. Her finding: the weight of this double day doesn’t come only from the sum of the tasks. It comes from their simultaneity. At work, the mind stays occupied by what must be done at home; at home, by what awaits at the office. A constant, invisible tension that never truly stops.

Forty years on, the notion has broadened, but the mechanism is the same. What exhausts us isn’t so much doing as having to think about doing. The form to return before Friday. The follow-up email not yet written. The appointment to schedule. The letter we keep putting off because we don’t know how to word it. Each of these tasks is small. Together, they occupy the mind like a background noise that never switches off.

This is precisely where AI can be a real help. Not for the big decisions, not for what makes a person’s worth. For that background noise: untangling a procedure, drafting a first version, summarising an endless document, organising what’s scattered. Tasks that call for neither genius nor heart, but that take up an outsized place in our heads.

Let’s be clear-eyed: AI doesn’t remove the mental load, it shifts it. What it produces must be reviewed, checked, sometimes corrected. But this shift is favourable. Reviewing a well-structured quote is faster, and more interesting, than producing it line by line while wondering whether you’ve forgotten anything. You reach a finished result sooner, instead of saturating your mind with a thousand questions about whether you’re at the right level. Vigilance remains essential — an AI can omit a line or state an error with confidence — and it’s precisely this review that keeps the human in control.

Still, you have to choose what you delegate. Not everything that weighs is clutter: some tedious tasks are also the ones where judgement is exercised and kept sharp. Delegate the formatting, yes. Delegate the thinking, no.

What AI relieves that is most precious isn’t our time. It’s our attention.

What the mind recovers when it’s relieved

We often hear that AI threatens human creativity, now that it can produce texts, images, ideas. That mistakes the question.

The question isn’t whether the human or the machine is more creative. It’s the state a human mind needs to be in to create. And there, the answer is hardly in doubt: a saturated mind doesn’t create. It copes. It executes. It survives its list.

Angélique Robin mentions the brain’s default mode, and neuroscience sheds light on this intuition without settling it. Since Marcus Raichle’s work in the early 2000s, we’ve known that the brain at rest isn’t inactive: a set of regions activates precisely when attention detaches from external tasks, when the mind wanders. More recent research, notably by Roger Beaty and colleagues, links creative ability to the cooperation between this network of spontaneous thought and those of attention and control. This work doesn’t prove that a relieved mind becomes creative. It suggests that thought needs room to roam.

This is my conviction: a person freed from the recurring, tedious tasks that cluttered their mind doesn’t become creative by magic. They recover their creative capacity. They regain availability for what makes their singularity: their experience, their judgement, their sense of others, their know-how, their imagination.

AI doesn’t make us more human. It can give us back the time and the space to be.

Three faces

Let’s leave the office. These scenes concern neither architects nor consultants. They concern just about everyone.

She cares for her father. He’s ageing, he needs help, and she works full-time. Her mental load is exactly the one Monique Haicault described: at the office, she’s thinking about the home-care form to fill in, the pension-fund letter she doesn’t understand, the supporting document she’s been asked for a third time. In the evening, she’s thinking about tomorrow’s meeting. AI helps her untangle the forms, understand what’s being asked, prepare the letters she’d been putting off for weeks. She reviews them and validates them before sending. What she gained isn’t first of all time. It’s a less crowded head. The following Sunday, she spent the afternoon with her father without a phone in her hand, without a list in her head. Present, simply. That is time for life.

He’s a carpenter. He chose this trade for the wood, the gesture, the piece imagined with a client. He used to spend his evenings on quotes, invoices, follow-ups. AI now prepares the templates from his site notes; he reads every line, because a quote is his word. But the real change is elsewhere. The time he no longer spends on paperwork, he spends on his current jobs: he refines, tends to the finishing touches, picks up the detail he’d have let slip. And that quality, which his clients notice and talk about, brings him his next contracts. He didn’t choose to do more. He chose to do better — and that’s what makes his living.

She’s a nurse. In her ward, the time spent documenting rivals the time spent with patients, and the backlog of data entry weighs on the whole end of the shift. If AI lightens that load, it’s attention returned to the bedside: noticing the detail that’s off, hearing a worry, finding the right word for a family. She doesn’t just gain time. She gains better conditions to exercise her know-how — and that is precisely where she performs best.

Working better also means performing better

I’ll be told that all this is very nice, but that a company lives on what it produces. True. And that is exactly why the quality argument is not an argument of comfort.

Take the nurse again. The temptation is understandable: needs are immense, waiting lists real. Since she’s saving time, why not give her three more patients? Because the research is unambiguous about what overload produces. A landmark study, led by Linda Aiken across 168 Pennsylvania hospitals among more than 10,000 nurses, showed that each additional patient per nurse was associated with a 7% rise in the risk of death within thirty days, and a 23% rise in the risk of burnout. Overloading the caregiver doesn’t care for more people. It cares less well, and it wears her down.

The same mechanism is found far from the hospital. Researchers from Oxford, Erasmus and MIT followed nearly 1,800 British Telecom call-centre agents for six months, measuring their well-being each week and cross-referencing it with their actual results. One extra point of well-being, on a 0-to-10 scale, came with about 12% more productivity. And that gain didn’t come from extra hours: it came from the fact that employees closed more of the calls they made. They weren’t working more. They were working better — and that’s what made them more productive.

What remains is to make this “better” visible. Three indicators are enough, far more telling than a forty-line dashboard:

  • errors and rework — what has to be redone, corrected, caught up;
  • absenteeism and departures — the visible cost of saturation. The World Health Organization estimates that depression and anxiety cause the loss of 12 billion working days worldwide each year, some 1 trillion dollars in lost productivity;
  • customer or user loyalty — what comes back, what recommends, what renews.

There is, finally, an argument leaders would be wrong to overlook: the relationship to work has changed. In France, the share of working people for whom work is “very important” in their life fell from 60% in 1990 to 24% in 2021, according to Ifop. And for the first time in twenty-two years, Randstad’s international Workmonitor survey ranks work-life balance ahead of pay among employees’ motivations. People no longer want to spend fifty hours a week at work — and for many, even thirty-five weigh heavily. An organisation that can make work well done and a liveable life compatible isn’t being indulgent. It becomes attractive, and it keeps its talent.

And the one who decides?

In these three faces, two decide for themselves what to do with the time given back: the caregiver and the carpenter. The nurse, like many employees, doesn’t decide it alone. That’s where the difference between time for life and time for work plays out.

But the one who decides isn’t an adversary. The leader, the manager, the business owner carries their own mental load too — often heavier than one thinks. They too can free themselves from it. They too can choose to produce more, if they wish, or to recover time for life. What matters isn’t the answer. It’s that there be a choice, made consciously, rather than a time saved that dissolves by default into what already existed.

The same technology can relieve or intensify. It isn’t the technology that settles it — it’s the choices of those who organise it.

There remains the question of investment: who pays for the tool if the gain doesn’t first translate into volume? It deserves more than a paragraph. I’ve made it the thread of several other texts: the return on investment of AI plays out less in the tool than in what precedes it — reliable data, clear governance, cross-cutting skills.

What I won’t dodge

Three clear-eyed points accompany this vision.

The mental load isn’t only an individual matter. Monique Haicault showed it: it’s first a question of distribution, and therefore of organisation. No tool will fix, on its own, an organisation that fragments work and offloads complexity onto individuals — that’s what I described about Conway’s Law, when HR job titles fragment the system before the code does.

Some data can’t be entrusted to just any tool. Health data, in particular, has no business in a consumer tool. It demands an architecture designed for it — suitable hosting, minimised data, controlled access — which lets you benefit from it without taking risks.

AI has a very real physical footprint. It can also help us reduce our own: comparing options, estimating an impact, factoring in the carbon criterion where a human has no calculator in their head. But that help doesn’t make the tool frugal by magic. It’s only worth it if the use stays proportionate to what it relieves.

What we risk forgetting

The real risk may not be that AI replaces the human. It’s that we forget what makes our singularity, and let the time it gives back dissolve into “ever more” without even noticing.

Used well, AI can on the contrary let us find time again to think, create, learn, pass on, and tend the human relationships that give meaning to our work and our lives.

It took a podcast about rest for me to hear it put so simply. Perhaps that’s no accident: it’s when we stop that we see what counts. And perhaps that is the best use of AI — not to make us go faster, but to let us, at last, stop a little.

And you: what has AI already taken off your mind? And that time — what did you do with it?

Sources & further reading

The starting point

  • Angélique Robin, certified coach, podcast Une vie plus grande (“A Bigger Life”), episode Comment ne pas culpabiliser de te reposer (“How not to feel guilty about resting”), published 20 September 2026. Passages quoted at 12 min 53 and 13 min 04; translated from the French. → The episode on Deezer

The mental load

  • Monique Haicault, “La gestion ordinaire de la vie en deux,” Sociologie du travail, vol. 26, no. 3, 1984, pp. 268-277. The founding article on the notion of mental load (in French). → Text on HAL · DOI 10.3406/sotra.1984.2072
  • Monique Haicault, La charge mentale. Histoire d’une notion charnière (1976-2020), 2020 (in French). → HAL

The resting brain and creativity

  • Marcus E. Raichle et al., “A default mode of brain function,” Proceedings of the National Academy of Sciences, vol. 98, no. 2, 2001, pp. 676-682. → DOI 10.1073/pnas.98.2.676
  • Roger E. Beaty et al., “Robust prediction of individual creative ability from brain functional connectivity,” Proceedings of the National Academy of Sciences, vol. 115, no. 5, 2018, pp. 1087-1092. Creative ability linked to the cooperation between the default mode network and the attention and control networks. → DOI 10.1073/pnas.1713532115

Quality, performance and attractiveness

  • Linda H. Aiken, Sean P. Clarke, Douglas M. Sloane et al., “Hospital Nurse Staffing and Patient Mortality, Nurse Burnout, and Job Dissatisfaction,” JAMA, vol. 288, no. 16, 2002, pp. 1987-1993. 168 Pennsylvania hospitals, 10,184 nurses, 232,342 surgical patients: each additional patient per nurse is associated with a 7% rise in the risk of death within 30 days, a 23% rise in burnout and a 15% rise in job dissatisfaction. → PubMed
  • Clément S. Bellet, Jan-Emmanuel De Neve, George Ward, “Does Employee Happiness Have an Impact on Productivity?,” Management Science, vol. 70, no. 3, 2024 (online 2023), pp. 1656-1679. Nearly 1,800 British Telecom call-centre agents followed for six months; causal effect of well-being on productivity, driven mainly by a higher call-conversion rate rather than by increased working time. → DOI 10.1287/mnsc.2023.4766 · Wellbeing Research Centre, Oxford
  • World Health Organization, fact sheet Mental health at work: 12 billion working days lost each year worldwide to depression and anxiety, at a cost of around 1 trillion dollars in lost productivity. → who.int
  • Romain Bendavid and Flora Baumlin (Ifop), « Je t’aime, moi non plus » : les ambivalences du nouveau rapport au travail, Fondation Jean-Jaurès, 2023. Share of French people rating work “very important” in their life: 60% in 1990, 24% in 2021 (in French). → jean-jaures.org
  • Randstad, Workmonitor 2025, survey of 26,778 people in 35 countries: for the first time in 22 years, work-life balance ranks ahead of pay (83% vs 82%). → randstad.com
  • Nous venons du chaos, l’IA de la logique (2025, in French) → The book

Your reactions, disagreements and additions are welcome — and on this subject, especially your own experience of what AI has taken off your mind. If the mental load becomes too heavy, please don’t hesitate to talk to a healthcare professional. Write to me at laurence.poussard63@gmail.com.

#AI #MentalLoad #WellBeing #Creativity #FutureOfWork #TimeForLife