Gary Marcus amplifies satire on regulating open-source rivals under the banner of AI safety

GaryMarcus · x · 2026-09-23

Gary Marcus retweeted a satirical jab at the AI-safety discourse: the claim that models are "so incredibly dangerous" we must regulate their open-source competitors and implement a slowdown — while the slowdown's real motivation has more to do with overinvestment and LLMs hitting the plateau of an S-shaped curve than with genuine danger.

Original post →

More from AGI Musings

AGI Musings channel →