Will AI Efficiency Actually Amplify Demand?

sanjaykalra · x · 2026-07-14

This is an in-depth opinion piece centered around the **Jevons Paradox**, asking whether the tools we use to "save the world" might ultimately drive global consumption even higher. The author first reviews the history of the steam engine and coal: improved efficiency did not reduce coal usage but rather unlocked new demand. Applying this logic to AI, they present two opposing views: - **The Moderation Camp** believes that cheaper tokens and more efficient models will lead demand into a gentle plateau, where growth is ultimately suppressed. - **The Expansion Camp** argues that market demand isn't a fixed pool but is highly elastic and constantly generating new use cases; cheaper cognitive costs will spawn more agents, deeper task chains, and even turn "cognition itself" into a new, deep-seated demand. The author ultimately frames the question as a judgment call: will the supply-demand relationship in the AI era look more like a "saturable body" or an "infinitely expanding market"?

Related event: AI Efficiency Gains May Amplify Demand(2 posts)→

Original post →

More from AGI Musings

AGI Musings channel →