Why there's no AI spam flood yet: 100x the cost of mail merge for only 2-10x the CTR

paul_cal · x · 2026-09-15

Responding to a question about why the predicted flood of AI-generated spam never materialized, paulcal lays out the economics: LLM inference is cheap in absolute terms but roughly 100x more expensive than traditional mail merge, while delivering only 2-10x better click-through rates. Given how ineffective spam already is per send—especially after filtering—scaling it with LLMs doesn't pencil out, which explains the absence of a spam deluge.

Original post →

More from AGI Musings

AGI Musings channel →