The 40-Hour Workweek Is About to Become Much Harder to Defend

You showed up. You worked your hours. You went home.

Five days.

Eight hours a day.

Forty hours.

The arrangement made a certain amount of sense in an industrial economy. If you were assembling automobiles, processing insurance claims, typing correspondence, or managing paper records, output was closely connected to the number of human hours available to produce it.

But what happens when that connection begins to break?

Agentic AI may be about to force that question upon us.

Because if I can accomplish by Wednesday afternoon what once took me until Friday at 5 p.m., an uncomfortable question follows:

We tend to talk about the 40-hour workweek as though it were a law of nature.

It isn’t.

It is a negotiated artifact of industrialization.

The great labor battles of the 19th and early 20th centuries were, in large measure, battles over time. Workers fought against brutally long days and weeks. The eight-hour movement eventually helped establish a new social bargain: employers could purchase a substantial portion of a worker’s productive time, but not all of it.

That was progress.

And eventually the 40-hour week became deeply embedded in American economic life.

But notice what remained largely unquestioned.

Even as manufacturing gave way to services, and services gave way to knowledge work, we kept the industrial clock.

Accountants worked 40 hours.

Marketers worked 40 hours.

Software developers worked 40 hours.

Human resources professionals worked 40 hours.

Managers worked 40 hours—and often many more.

The nature of the output changed dramatically.

The unit used to measure the worker’s contribution did not.

For years, technology improved workplace productivity incrementally.

Email accelerated correspondence.

Spreadsheets accelerated calculation.

Search engines accelerated research.

Collaboration platforms accelerated communication.

But humans remained firmly in the workflow.

Research.

Analysis.

Scheduling.

Summarization.

Drafting.

Reporting.

Data retrieval.

Presentation development.

Workflow management.

And increasingly, coordination among other AI agents.

That means the productive capacity of an individual knowledge worker could expand substantially without a corresponding increase in human working time.

Imagine that a marketing manager in 2024 required five days to complete a week’s responsibilities.

By 2027, aided by sophisticated AI agents, she can accomplish comparable work by Wednesday afternoon.

What happens next?

There are at least two very different possibilities.

This is probably the outcome many workers already expect.

You finished your old workload in three days?

Wonderful.

Here’s more work.

Instead of managing five accounts, manage eight.

Instead of producing two reports, produce six.

Instead of overseeing one project, oversee three.

Instead of responding to 50 customers, respond to 150.

The workweek remains 40 hours—or 45 or 50—but considerably more productive activity is squeezed into it.

From the perspective of a quarterly earnings report, the logic is seductive.

If technology can make each employee dramatically more productive, then the organization can generate more revenue with the same headcount—or perhaps generate the same revenue with fewer people.

Shareholders benefit.

Margins improve.

Executives celebrate productivity gains.

But the human being may experience something very different.

Instead, it becomes the mechanism through which more drudgery can be assigned.

We should be very careful about calling that progress.

There is another possibility.

What if we decide that some portion of the productivity dividend created by AI should return to the person whose skills, judgment, experience, and oversight make that productivity possible?

Perhaps the employee who once needed 40 hours to produce a certain outcome now needs 32.

Or 30.

Or 25.

And perhaps we stop treating the remaining hours as automatically belonging to the employer.

That doesn’t necessarily mean everyone heads to the golf course at noon on Wednesday.

It could mean more time for learning.

More time for mentoring younger colleagues.

More time for experimentation.

More time for creative thinking.

More time for family.

More time for community.

More time for exercise and health.

Or, sometimes, simply more time that belongs to you.

That last possibility may be the most radical.

This may become one of the defining workplace questions of the next decade.

Forty hours?

Or valuable outcomes?

We have often blurred the two because, historically, they were difficult to separate.

If a project required 40 hours of human effort, then purchasing 40 hours and purchasing the outcome were essentially the same thing.

But if AI helps someone produce that outcome in 25 hours, they become very different propositions.

And that creates an extraordinary tension.

The employer may say:

We pay you for 40 hours. If you finish one assignment early, we’ll give you another.

The employee may reasonably respond:

But if my experience, judgment, and effective use of AI allow me to produce twice as much value in the same amount of time, shouldn’t I participate in that productivity gain?

Both sides can make an argument.

Which is precisely why we need to start having this conversation now.

I am particularly interested in what happens inside publicly traded companies.

The pressure to capture the AI productivity dividend will be enormous.

Imagine telling investors that your company has deployed AI systems that increase employee productive capacity by 30 percent.

What happens next?

Does the company announce that employees can now work four-day weeks?

Perhaps.

But Wall Street may ask another question:

Why aren’t you using that additional capacity to grow earnings?

That pressure matters.

One says AI should allow humans to work better.

The other says AI should allow humans to produce more.

Those are not necessarily the same thing.

And the difference between them may shape the workplace for a generation.

There is another problem with simply filling every hour AI frees.

Human beings are not machines with unused processing capacity.

We need cognitive white space.

We need conversations that don’t have immediate deliverables attached to them.

We need time to think.

We need time to notice things.

We need time to build relationships.

We need time to develop younger colleagues.

We need time to question whether what we’re doing makes sense at all.

Judgment.

Discernment.

Curiosity.

Creativity.

Empathy.

Mentorship.

Strategic thinking.

Those capabilities do not necessarily flourish when every minute of the day is optimized for measurable output.

For me, this discussion leads directly back to Work Fit.

Work Fit isn’t about avoiding hard work.

And it certainly isn’t about expecting employers to pay people for doing nothing.

For more than a century, 40 hours became our default answer.

But perhaps it was never the answer.

Perhaps it was simply the best compromise available at the time.

Technology has changed.

Work has changed.

Productivity has changed.

And now AI may dramatically change the amount of human time required to create economic value.

If that happens, our assumptions about the workweek should change with it.

Maybe the future is a four-day week.

Maybe it is a 30-hour week.

Maybe different professions develop completely different arrangements.

Maybe we move away from measuring professional work primarily through time and toward something much closer to outcomes.

There probably won’t be one answer.

But there is one question I think every employer—and every employee—should begin asking:

Because the answer cannot automatically be:

The employer.

If AI allows Tuesday’s work to be finished Monday, and Friday’s work to be finished Wednesday, we have created something enormously valuable.

Not merely productivity.

Time.

The question is whether companies will treat that dividend as another resource to extract—or something their people deserve to share.

Perhaps the 40-hour workweek survives the age of agentic AI.

But for the first time in generations, employers may have to make a much stronger case for why it should.

Our best days lie ahead.

Note: image in this post rendered on ChatGPT based on our prompts.

I’m Dan Smolen. As host and executive producer of What’s Your Work Fit? I help you make your work and workplace decisions result in better and more satisfying professional experiences and outcomes. I am also a Founding Member of The Future of Work Alliance.

Join What's Your Work Fit? on LinkedIn

Join What's Your Work Fit? on YouTube