Making an engine more efficient is how you end up burning more coal
A Victorian economist noticed something about steam engines in 1865 that still wrecks every clean-tech optimism today.
The intuitive fix for a resource running low is to use it more efficiently. William Stanley Jevons, writing in 1865 as Britain worried about exhausting its coal seams, showed the intuition backwards.
James Watt's improved steam engine burned far less coal per unit of work than earlier designs. Jevons's observation was that this didn't reduce coal consumption — it expanded it, because cheaper, more efficient power made steam profitable for jobs it hadn't been worth using it for before. Factories that couldn't justify a coal-hungry engine suddenly could justify an efficient one. Total coal use climbed. What's now called the Jevons paradox is exactly this: efficiency gains lower the cost of using a resource, and lower costs tend to summon more total use, not less.
The pattern keeps resurfacing somewhere new every generation. More fuel-efficient cars correlate with people driving more miles, not fewer. Cheaper LED lighting led to more lighting installed, in more places, left on longer, rather than a straightforward drop in electricity for lighting. Efficiency didn't fail in these cases. It worked exactly as designed and simply changed what became affordable to do more of.
Jevons wasn't arguing against efficiency. He was arguing against the comfort of assuming it settles the question by itself.