You show up. You work your hours. You go home.
An employer purchases a certain amount of our time and productive capacity. In return, we receive compensation and, theoretically, enough time outside of work to have a life.
The unintended consequence of change is that Agentic AI may blow apart that bargain.
Because if AI agents can perform in minutes work that once consumed hours—researching, analyzing, scheduling, producing reports, preparing presentations, writing code, responding to customers, managing workflows—then something very important happens:
The relationship between time and output begins to break down.
And that leaves us with a question much bigger than whether the 40-hour workweek survives.
Imagine a professional whose job today requires roughly 40 hours of work each week.
Now imagine that, through the effective use of AI agents, that person and their organization can produce the same amount of useful output in 30 hours.
What happens to the other ten?
One possibility is extraordinary.
The company says: We are accomplishing what we need to accomplish. Go live your life.
Maybe the workweek becomes 32 hours.
Maybe Friday disappears.
Maybe people work intensely for portions of the week and have much greater flexibility during the rest.
Maybe parents get more time with their children. Caregivers get more breathing room. People exercise, volunteer, learn, create, rest or simply spend more time doing things that have nothing whatsoever to do with increasing shareholder value.
That could be one of the great dividends of artificial intelligence.
It could also help millions of people achieve something closer to their Work Fit: work that fits who they are, how they want to live and the contribution they want to make.
The employer says:
Great. You finished that work in 30 hours.
Here’s ten more hours of work.
This gets us to a question I raised earlier this week and can’t stop thinking about.
If technology makes a worker twice as productive, who gets the benefit?
The customer?
The shareholder?
The employer?
The employee?
Or some combination of all four?
That question may become one of the defining workplace debates of the AI era.
Those are not quite the same thing.
Productivity tells us that we can accomplish more with fewer inputs.
Productive capacity asks what we choose to do with the capability we have created.
And that is a leadership decision.
This is where the discussion gets uncomfortable.
Imagine two publicly traded competitors.
Both adopt agentic AI.
Both discover that their employees can accomplish substantially more in substantially less time.
If our people can produce 25 percent more, why settle for the same output?
Why not increase sales targets?
Why not increase client loads?
Why not reduce headcount and distribute the remaining work among fewer people?
Why not demand faster product cycles?
Why not turn every hour AI saves into another hour of human productive capacity?
The market may reward the organization that extracts the greatest immediate economic value from AI—not necessarily the organization that distributes some of that value back to the people creating it.
A CEO who tells investors, “AI allowed us to give thousands of employees Friday afternoons back” may have created enormous human value.
But another CEO may tell investors:
“AI allowed us to increase output 30 percent while reducing our workforce 15 percent.”
But we should be asking the question.
Technologically, the argument may become easier.
If organizations can produce the same output in dramatically less human time, maintaining a five-day, 40-hour schedule becomes increasingly difficult to justify on productivity grounds alone.
But economically?
The opposite pressure could emerge.
That would give us a strange outcome.
We would have created technologies capable of liberating people from enormous amounts of routine work—and used them instead to make people work harder.
Forty hours of work in the AI age could feel very different than we are use to working.
More decisions.
More projects.
More clients.
More responsiveness.
More measurement.
More expectations…and
Less recovery time between tasks because the technology eliminated the natural pauses that once existed within work.
The case for sharing the AI dividend with workers doesn’t have to rest entirely on benevolence.
People are not machines.
There is a limit to the number of consequential decisions we can make well.
There is a limit to how much complexity we can absorb.
There is a limit to how many clients, projects, conversations and problems we can manage simultaneously before the quality of our judgment begins to decline.
And that matters even more in an AI-supported workplace.
Those capabilities do not necessarily improve when every available minute is filled.
Quite the opposite.
A company that gives people some of their AI-created time back may discover that it hasn’t sacrificed productive capacity at all.
It may have strengthened the human capabilities that become most valuable when machines do everything else faster.
I don’t think the answer is simply legislating a 32-hour workweek.
Nor do I think every company or every kind of work can operate on the same schedule.
Healthcare isn’t advertising.
Manufacturing isn’t consulting.
Hospitality isn’t software development.
And a boot-strapped startup fighting for survival operates differently from a mature enterprise.
But we should stop treating the workweek as an immutable law of nature.
It isn’t.
It is an arrangement.
And arrangements can change when the conditions underlying them change.
Perhaps employees receive some of the productivity dividend as higher compensation.
Perhaps customers receive some through lower prices or better service.
Perhaps shareholders receive some through increased profitability.
And perhaps workers receive something that has rarely appeared on a corporate balance sheet: time.
Every company adopting agentic AI is going to make this decision, whether its leaders recognize it or not.
What happens when AI saves an employee an hour?
Do we immediately refill it?
What happens when it saves a day?
What happens when an entire team can accomplish its previous workload by Thursday afternoon?
Do we celebrate the achievement and send them home?
Or, do we address the backlog?
Does it see AI as a way to improve work—or primarily as a way to extract more work from humans?
I believe that distinction will matter enormously.
Companies that simply turn every AI efficiency into another demand on their people may achieve spectacular productivity improvements.
For a while.
But they may also create exhausted organizations in which employees discover that every new technology promising to make their lives easier ultimately makes their workloads heavier.
And then many of them may run hair on fire for the exits.
Once people understand that bargain, why would they embrace the next productivity tool?
I don’t know.
And that’s exactly why I think we need to start having this discussion now.
Agentic AI is giving organizations an extraordinary new productive capacity.
We can use that capacity to produce more.
We can use some of it to work less.
Or we can attempt to do both.
But we shouldn’t allow the answer to emerge accidentally from quarterly earnings pressure, utilization targets and the assumption that every efficiency must immediately become additional output.
Because underneath the debate about the four-day week is a much more consequential question about the future of work:
If technology allows us to accomplish dramatically more with dramatically less human effort, should humans actually benefit from that progress?
Or will the reward for becoming more productive simply be more work?
I’d especially like to hear from leaders, employees, investors and board members:
That may be one of the defining Work Fit questions of the next decade.
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.
No posts
Comments
Nothing yet. Say the first thing.
Sign in to join the conversation.