There’s something almost beautiful about the audacity of it.
We’re told that artificial intelligence has made us ten times more productive, that the old constraints on human output are falling away, that we’re entering an age of abundance, where work that once took days can be accomplished in hours, and work that once took hours can be accomplished in minutes.
We’re told to embrace the opportunity. To reinvent ourselves. To accelerate.
And then, before we’ve had so much as a moment to enjoy this miraculous new prosperity, we’re told we need to accelerate some more.
Faster. More efficiently. With fewer resources, if possible. Ideally without complaining about it.
But hey, I’ll grant us a quick breath to ask a question on our behalf: where the hell is our pie?
I’m not talking about a slice of the pie, no, nor some modest increase in the percentage of company profits allocated to the people whose labor produces them. We’re not after a pizza party, an employee appreciation luncheon, or an additional $3.50 in annual bonuses if we somehow clear another year’s ever-rising expectations.
I’m talking about a whole damn pie.
If the gains are really as extraordinary as we’re being told, there should be pies everywhere. Pies stacked in the break room. Pies lining the hallways. Pies being wheeled into the office on carts because there are simply too many pies to carry.
Instead, all we seem to be getting are demands to bake them faster without daring to have ourselves a taste along the way.
The Miracle of Doing More for Less
To me it’s clear that AI has transformed how we work. I see daily what these tools can do. I use them to accomplish things that would have been prohibitively time-consuming only a few years ago, so I’m not going to pretend the technology is useless just because I have concerns about how it’s being deployed.
What irks me—and what should irk all of us—is that there’s a difference between a technology making certain tasks dramatically easier and an entire workforce becoming ten times more productive. Why? Because the evidence, as it turns out, is considerably less extravagant than the rhetoric.
A study of more than 5,000 customer support agents found that access to a generative AI assistant improved productivity by about 14% on average, with substantially larger gains among less experienced workers. That’s meaningful. It’s also a long way from tenfold.1
More recently, researchers studying the adoption of AI chatbots in Denmark found that workers reported productivity benefits, but those benefits had not translated into measurable increases in earnings or hours worked two years after ChatGPT’s introduction.2
Then there’s a 2026 review from the International Labour Organization that concluded that productivity improvements are real but uneven, frequently unverified, and have yet to translate consistently into higher measured output or earnings.3
It seems to me the pies are increasing in size, even if not tenfold. And yet, here we are years later, with only unevenly distributed gains to show for it.
In other words, when businesses want employees to embrace AI, its productivity benefits are practically supernatural, but when it comes time to compensate employees for those benefits, the gains become rather difficult to locate.
Now, I’ll allow that maybe the promised gains haven’t materialized yet. Maybe they’re being consumed by implementation costs, quality issues, and the enormous expense of making these systems work reliably. Maybe companies are taking a long-term view of an investment whose returns won’t arrive for years.
Fine. Those are legitimate possibilities.
But if that’s the case, perhaps we could dispense with the breathless declarations about how much more productive everyone has supposedly become; we shouldn’t declare a productivity revolution when setting expectations and then plead uncertainty when discussing how its gains should be distributed.
To put it plainly: Pick a lane.
The Productivity Dividend
There’s a historical dimension to all of this that’s worth understanding.
The Economic Policy Institute has documented a divergence between the growth of productivity and the compensation of typical American workers since 1979. While both measures rose at roughly comparable rates in the decades following World War II, productivity subsequently began growing much faster than typical workers’ pay.4
That doesn’t mean, however, that workers have received none of the benefits of technological progress. Real compensation has risen, and the relationship between productivity and wages depends in part on how each is measured. But the gains haven’t been distributed evenly, and productivity growth has not guaranteed comparable improvements for the typical worker.
The lesson should be obvious: Greater productivity doesn’t automatically produce greater prosperity for the people doing the work, and that distribution is a choice.
And now we’re being required to take part in another technological transformation without much discussion of whether the people making it possible will share proportionately in its rewards.
Consider what happens when a worker figures out how to accomplish eight hours of work in four. There are several possibilities.
Their employer could pay them the same salary for a shorter workday, recognizing that the value of the work hasn’t changed simply because it takes less time to produce.
The employer could maintain the eight-hour day but increase compensation to reflect the additional value being created.
The employer could share a portion of the savings through bonuses, profit-sharing, or other benefits.
Or the employer could congratulate the worker on their ingenuity, assign them another four hours of work, and begin planning next year’s productivity targets around the assumption that they’ll maintain their new pace.
Now, you tell me: Which of these feels most familiar?
Granted, that last option isn’t unreasonable from a certain perspective. Businesses exist to create value, and there’s nothing wrong with using new technology to serve more customers, develop better products, or pursue opportunities that weren’t previously economical. The problem arises, however, when the additional value flows exclusively in one direction.
When every improvement in efficiency becomes the new minimum expectation. When yesterday’s extraordinary performance becomes today’s baseline. When an employee who discovers a better way to do something is rewarded primarily with more things to do.
Eventually, innovation begins to look less like an opportunity and more like a trap, and why wouldn’t it? If becoming more efficient only earns you more work, what incentive do you have to become more efficient?
We Were Supposed to Be Buying Back Our Time
For all the talk about AI transforming the workplace, there’s remarkably little imagination being applied to what a transformed workplace might actually look like.
We seem perfectly capable of imagining software that writes code, drafts contracts, summarizes meetings, analyzes data, and handles increasingly complicated administrative work, but the moment someone suggests that employees might work fewer hours as a result, imagination deserts us.
Forty hours is forty hours, apparently—a number as immutable as gravity.
Never mind that it’s a convention rather than a law of nature. Never mind that the entire premise of a productivity-enhancing technology is that we can accomplish more with less effort. Apparently, the one thing we’re not allowed to accomplish with less effort is our jobs. Infuriating, to say the least.
And we’re not merely talking about money here. We’re talking about human lives.
We’re talking about the hours people spend away from their families. The hobbies they abandon because they’re too exhausted to pursue them. The books they don’t write, the gardens they don’t tend, the communities they don’t participate in because work has consumed so much of their available time and attention.
We’re talking about the parent who could pick up their child from school. The caregiver who could be home for another afternoon. The person who could spend a Friday hiking instead of answering emails.
If AI genuinely allows us to accomplish in thirty hours what previously required forty, why is a thirty-hour workweek treated as a radical proposition? Why is the assumed outcome another ten hours of productivity rather than ten hours of life?
And before someone starts explaining the importance of competition, let me acknowledge the obvious: Yes, businesses compete. Yes, reducing costs and increasing output are essential to remaining viable. Yes, sometimes productivity gains must be reinvested just to stay in business.
I’m not arguing that every minute saved should automatically become a minute of paid vacation. I’m arguing that the possibility—or something like it—should be on the table. I’m arguing that the people whose labor and expertise turn technology into actual business value deserve a voice in determining what happens to that value. I’m arguing that the conversation about productivity should involve more than how much additional output can be extracted from an employee before they collapse.
Because if the purpose of technological progress isn’t to improve human life, what the hell are we progressing toward?
Who Actually Made the Pie?
There’s another peculiar feature of the AI productivity narrative that bothers me, namely that we’re regularly encouraged to think of the technology itself as the source of all this newfound value.
AI wrote the code. AI drafted the proposal. AI found the answer.
But who identified the problem? Who knew what the customer needed? Who recognized that the generated code was wrong? Who understood the organization’s constraints, made the judgment calls, corrected the mistakes, and took responsibility when the system produced something that looked convincing but wasn’t remotely fit for purpose?
Who learned to use the tools in the first place?
Who figured out which workflows could be improved, which ones shouldn’t be automated, and how to integrate these capabilities into systems that already existed?
We did.
Businesses are investing in AI—and those investments deserve returns—but buying access to a tool isn’t the same thing as creating value with it. An employer can purchase the most extraordinary technology in the world and still accomplish very little without people who know what they’re doing, yet the story we’re increasingly invited to believe is that the technology deserves credit for the gains while the workers fortunate enough to use it should simply be grateful they remain employed.
What a convenient arrangement.
Capital provides the tool. Labor figures out how to make it useful. Capital claims the returns. Labor inherits a new baseline.
And if labor objects, someone will invariably point out that the technology makes their work easier—as though easier work were inherently less valuable, as though the amount of suffering required to produce something ought to determine how much its creator deserves to be paid.
What an absurd conception of value.
If someone invents a more efficient way to manufacture a product, we celebrate the innovation. We don’t insist that the product must now be worthless because producing it requires less effort.
Why, then, should an employee’s contribution become less deserving of compensation simply because they’ve learned to produce it more efficiently?
The relevant question, then, has little to do with how difficult the work has become and everything to do with how much value the work creates and how that value ought to be distributed among the people and institutions responsible for creating it.
Show Us the Math
This is where I’d like corporate leaders to start being considerably more specific.
If AI is making your employees dramatically more productive, quantify it. Show us the costs. Show us the savings. Show us the increased output, the improved quality, the additional revenue, and the return on the investment.
Then show us where the gains are going.
Are they funding additional hiring? Better products? Lower prices for customers? Higher wages? Shorter workweeks? Increased retirement contributions?
Or are they flowing into executive compensation, shareholder distributions, and financial targets that seem to become more ambitious every time they’re achieved?
Some mixture of these outcomes may be perfectly defensible. Shareholders take risks. Companies need capital. Customers should benefit from innovation, too.
But workers aren’t merely an expense to be minimized in pursuit of those other interests. We’re participants in the enterprise. Our knowledge, judgment, adaptability, and effort are often the very things converting a technological possibility into an economic reality, and we deserve more than a congratulatory email about how much we’ve accomplished.
I want to see companies establish meaningful mechanisms for sharing productivity gains. Profit-sharing tied to measurable improvements. Compensation that reflects expanding responsibilities. Paid time returned to employees. Clear opportunities for frontline workers to participate in decisions about automation.
I want to see this not because every company has unlimited resources, but because productivity should be a source of shared prosperity rather than an endlessly escalating obligation.
And if the gains aren’t sufficient to support any of those things? Then stop telling us we’re ten times more productive. Stop using the language of abundance to justify the expectations of scarcity.
The Whole Damn Pie
What frustrates me most is that I actually believe in the promise of this technology.
I believe there are extraordinary possibilities in giving people tools that allow them to create more, learn faster, and overcome limitations that once constrained their ambitions.
I believe there’s something genuinely exciting about making complex work more accessible and giving people capabilities they didn’t previously possess.
But I don’t believe that excitement obligates us to surrender the benefits of our ingenuity to the organizations that employ us, and I certainly don’t believe we should be expected to respond to every technological advance by making ourselves available for even more work.
At some point, the question has to change.
Not: How much faster can we go?
Not: How much more can we produce?
Not: How many additional responsibilities can we fit into the same number of hours?
But: When do we get to enjoy the benefits of what we’ve already accomplished?
Because if we really are producing twice as much, or three times as much, or ten times as much value as we were before, there ought to be a corresponding improvement in our lives.
Maybe that means better compensation. Maybe it means more time with our families. Maybe it means some combination of financial security, professional autonomy, and the freedom to pursue things that have absolutely nothing to do with generating returns for somebody else.
Workers deserve to participate in deciding how rewards are distributed, and until that happens, I’m not particularly interested in hearing another inspirational speech about how AI is going to make us all superhuman.
If these are the greatest productivity gains in generations, if we’re truly entering an unprecedented age of technological abundance, then there ought to be enough prosperity to go around.
I’m not asking for a bigger slice of the same old pie.
According to the people announcing this revolution, we should have enough ingredients, equipment, and newfound capacity to open a whole damn bakery.
So quit telling us how much faster we need to work.
Show us the pies.
Tell Me in the Comments
When a tool makes your work faster, what actually happens to the time it saves?
If your workplace became substantially more productive, would you rather see the gains in pay, shorter hours, or something else? Who should help decide?
What’s the strongest case for putting those gains back into the business instead of returning them to workers? Does it persuade you?