What will more intelligence actually do for us?
from Aarva Hosted by Noah Smith
If machines can outsmart mathematicians, why haven't they replaced truck drivers? In an essay published on 26 July 2026, the economics blog Noahpinion notices a quiet contradiction in the artificial intelligence story. Machines are now solving math problems that stump the world's best human experts, yet our morning commutes and daily jobs look exactly the same. The promised economic explosion has not arrived. The piece asks a simple question: what if raw brainpower has diminishing returns? It suggests we are expecting computers to act like solitary human geniuses, which blinds us to the completely different kind of industrial revolution they will actually trigger. An argument that the economic value of artificial intelligence will come not from limitless superintelligence, but from combining human reasoning with the physical scale of computers. The text outlines three ways this combination could drive productivity, even if individual cognitive gains face diminishing returns. It details how AI can multiply cognitive labor through data centers, extract hidden manufacturing techniques from sensor data, and identify scientific laws that humans cannot perceive. Read at source: Noahpinion For more features and details, visit aarva.app
Embed this episode
No transcript yet.