AI May Not Actually Reduce Workloads

chrisgrayson · x · 2026-07-13

This repost criticizes Amodei's tendency to heavily promote the narrative that "AI will take over everything" around funding milestones, viewing it as a standard PR pitch.

The core argument hinges on the Jevons Paradox: making a resource more efficient technologically often expands demand rather than reducing total consumption. The author uses Hinton's famous 2016 prediction about no longer training radiologists as a counter-example. After AI made image analysis faster and cheaper, clinical requests for imaging surged, actually increasing radiologist headcounts and salaries. This demonstrates that the linear narrative of "AI replacing jobs" is fundamentally flawed.

Related event: Jevons Paradox in the AI Era: Efficiency Expands Demand Instead of Eliminating Jobs(6 posts)→

Original post →

More from AGI Musings

AGI Musings channel →