Jevons Paradox, Our Work Hours, and AI

I recently came across a concept called Jevons Paradox while watching one of my favorite YouTubers, Hank Green, andI thought it was worth sharing. It's an idea from economics that describes something counterintuitive: making a resource more efficient to use can actually increase its overall consumption, not decrease it.

At first glance, it seems obvious that efficiency should lower consumption. Take coffee, for example. If your caffeine intake became more "efficient" say condensed into a single espresso shot instead of a full mug, you'd assume the total volume of coffee consumed would drop. And for a lot of people, that holds true.

But Jevons Paradox tells us this isn't a universal law. Sometimes, efficiency does the opposite of what we expect. It says that when a system becomes more efficient, the cost of achieving the same output drops. Because it's now cheaper (in time, money, or effort) to get the same result, demand increases. And as demand rises, we end up circling back to consuming the same amount of the resource or even more than before.

A good example always discussed is the coal consumption during the Industrial Revolution. As steam engines became more efficient at converting coal into energy, many assumed coal usage would fall. Instead, the opposite happened, cheaper and more efficient engines made coal-powered machinery viable for far more industries, and total coal consumption skyrocketed.

AI Seems to Be Following the Same Pattern

We're watching a similar story unfold with AI today. As AI models become more efficient, smaller, faster, cheaper to run per query, the natural assumption is that overall energy consumption should decrease. But instead, easier and cheaper access to AI is driving explosive adoption across industries, devices, and daily habits. More efficient models haven't slowed energy demand from AI but they've fueled a race to build ever-larger data centers and deploy AI everywhere, pushing total energy consumption upward.

What Intrigues Me Most: Modern Life

This is where it gets personal. We live in an era with more tools, automation, and access to knowledge than any generation before us. We've built machines, apps, and systems specifically designed to make our lives more efficient. But has that actually reduced the volume of hours we put into work?

Or are we, in true Jevons Paradox fashion, just using that newfound efficiency to take on more and more tasks, more responsibilities, more "productivity" until we're right back to working just as many hours, if not more?

I wonder if we're sometimes too narrowly focused on optimizing individual tasks, writing faster, automating a process, cutting a meeting short and without stepping back to ask a more holistic question: is all this efficiency actually buying us back time, or just quietly raising the bar for how much we're expected to do? Worth a discussion, right?

A Question for You

Has your work efficiency gone up thanks to better tools and AI? And if so, have your actual work hours gone down? Or has the bar just moved?

I find this genuinely fascinating. I'll be reading more about this and sharing my thoughts in a follow-up post. Stay tuned.