Artificial intelligence is becoming cheaper, faster, and easier to deploy. Yet instead of shrinking workloads, it may be doing the opposite. That tension sits at the center of a growing debate across Silicon Valley and academic economics, and it has been articulated most clearly by Aaron Levie, CEO of Box, who recently connected modern AI adoption to Jevons Paradox, a 19th‑century economic theory that warns efficiency gains can increase overall consumption.
Levie’s argument lands at a moment when businesses are racing to automate knowledge work. As AI model costs fall and capabilities expand, the promise of productivity is colliding with an uncomfortable reality: lower costs often encourage more usage, not restraint.
Aaron Levie on AI and Jevons Paradox
In a widely circulated post, Aaron Levie framed today’s AI boom through the lens of Jevons Paradox, the idea that efficiency gains can drive higher total demand.
“Jevons paradox is coming to knowledge work,” said Levie
He explains how falling costs for AI make it possible to tackle more work than ever before,
“By making it far cheaper to take on any type of task that we can possibly imagine, we’re ultimately going to be doing far more,”
Levie also highlights the future scope of AI-driven work, noting that most AI usage will extend beyond current human tasks,
“The vast majority of AI tokens in the future will be used on things we don’t even do today as workers: they will be used on the software projects that wouldn’t have been started, the contracts that wouldn’t have been reviewed, the medical research that wouldn’t have been discovered, and the marketing campaign that wouldn’t have been launched otherwise.”
Taken together, Levie’s observations frame AI not just as a cost-saving tool, but as a demand-expanding force that reshapes the scale, scope, and nature of knowledge work across organizations. The paradox he identifies suggests that as AI becomes cheaper, it will enable more work rather than less, illustrating the modern implications of efficiency-driven demand.
The Economist Behind the Theory
The paradox Levie references traces back to William Stanley Jevons, a British economist writing during the Industrial Revolution. Observing coal consumption in 19th‑century England, Jevons noticed that efficiency improvements did not reduce usage.
“It is a confusion of ideas to suppose that the economical use of fuel is equivalent to diminished consumption. The very contrary is the truth.”
Jevons argued that efficiency makes a resource more attractive, encouraging broader adoption. What coal was to steam engines, compute may now be to artificial intelligence.
Economists See a Familiar Pattern in AI
Modern economists have increasingly applied Jevons Paradox to digital technologies. Erik Brynjolfsson, a Stanford University economist and one of the world’s leading researchers on productivity and technology, has described how AI efficiency can expand rather than contract economic activity.
In an interview reported by NPR’s Planet Money, Brynjolfsson explained that when powerful technologies become cheaper, organizations invent new tasks and processes to absorb the surplus capacity. The result is not less work, but different work, often at a larger scale.
Brynjolfsson’s research has long emphasized that productivity gains rarely translate directly into reduced labor. Instead, they tend to reshape workflows, create new demand, and shift where human effort is applied.
Beyond Academia: Industry Leaders Echo the Concern
The Jevons Paradox framing is not limited to economists. Technology leaders have publicly acknowledged the same dynamic as AI models become more efficient and widespread.
Satya Nadella, CEO of Microsoft, captured the logic succinctly in his post on X,
“Jevons paradox strikes again! As AI gets more efficient and accessible, we will see its use skyrocket, turning it into a commodity we just can’t get enough of.”
Nadella’s statement underscores a critical point: as AI becomes cheaper and more capable, organizations and individuals are likely to use it far more extensively, not less. Efficiency does not automatically translate into reduced workload; instead, it creates new opportunities, expands the scope of tasks, and drives greater adoption across enterprises and consumers alike.
This aligns with the concerns voiced by economists and AI practitioners, illustrating that the paradox of efficiency driving demand is now a reality in the corporate world, not just an academic concept.
Policy and Labor Implications
Economists studying AI’s long-term impact see broader consequences beyond simple efficiency gains. Anton Korinek, Professor at the University of Virginia, Department of Economics and Darden School of Business, writes in his article The Economics of Transformative AI on nber.org,
“AI systems advance toward mastering all forms of cognitive work that can be performed by humans, including new tasks that don’t even exist yet.” said Anton in
Korinek’s research emphasizes that rapid AI adoption can outpace institutional and labor market adjustments, creating both opportunities and challenges for workers and policymakers. Rather than merely automating existing tasks, AI expands the scope of work itself, generating new forms of cognitive labor that organizations may seek to exploit.
This perspective aligns with Jevons Paradox: as AI becomes cheaper and more capable, total demand for knowledge work is likely to increase rather than decrease. Policymakers and businesses must consider not only efficiency gains but also the broader economic and social consequences of an ever-expanding digital workforce.
Implications for Business and Society
The convergence of Levie’s industry perspective and economists’ warnings arrives as enterprises roll out AI at scale. From customer support to legal review, the question is no longer whether AI saves time, but what organizations do with the time it saves.
If Jevons Paradox holds, efficiency gains will not lead to shorter workdays or reduced workloads. Instead, they will fuel higher expectations, faster cycles, and more output per worker.
Levie’s framing captures the dilemma facing modern businesses. AI may be the most powerful productivity tool ever created, but productivity itself can become a trap. As intelligence gets cheaper, the demand for it may become effectively limitless.
In that sense, Jevons Paradox is not a historical curiosity. It is a warning, resurfacing in the age of artificial intelligence, that efficiency alone does not determine how much work gets done. It only changes how much work we decide is worth doing.
Why AI Won’t Take Jobs, But Create More
As debates about automation and employment intensify, Aaron Levie, in another LinkedIn post, offered a grounded perspective on how artificial intelligence will actually reshape work, not simply eliminate it.
“AI will largely automate tasks, not jobs,”
He explains that when AI agents automate parts of a job, the definition of that job expands rather than disappears:
“Even as AI agents get better at automating more of what we define as a job in a particular field, we will just raise the bar for what the job is.”
Levie argues that digital technologies have historically transformed work by automating discrete tasks and creating new expectations for roles, for example turning telephone operators, typists, and clerks into obsolete tasks integrated into broader careers. Applied to AI, the same dynamic suggests that as tools become more capable, the nature of work evolves, requiring higher‑level judgment, creativity, and complex decision‑making rather than simply eliminating roles.
This perspective resonates with economic trends across multiple industries, where automation has historically augmented human labor rather than made it redundant. The implication for business leaders is clear: AI adoption should be framed around how it expands the scope and value of human work, not just how it cuts costs.
