More inference than training machines means writing will soon beat reading
GregoryDiamos · x · 2026-09-22
An intriguing observation: AI slop is already too much for machines to read.
The argument: there are far more inference machines than training machines, and inference is cheaper than training. Therefore we'll soon live in a world where writing is cheaper than reading — even for AIs, making filtering and attention the scarce resource.
More from AGI Musings
- Everyone draws a blue/white collar line: the psychology of automation double standards — RexDouglass · 2026-09-22
- AI-vs-mathematicians drama erupts: 1M+ follower academic tells AI to 'wipe them out' — burny_tech · 2026-09-22
- Why doesn't anyone romanticize coders? AI era's double standard on protecting work — RexDouglass · 2026-09-22
- Conference fixes for AI paper flood: per-author submission caps and faster fake-profile detection — sarahookr · 2026-09-22
- repligate on EA's power trap: optimizing for legible status eventually bites you — repligate · 2026-09-22
- Even Dath Ilan couldn't innovate that fast: the Dyson-sphere debate continues — austinc3301 · 2026-09-22