More Worried About Capable Stupidity Than Superintelligence
mrjonfinger · x · 2026-09-11
A tech commentator offers a contrarian take on AI capability: rather than fearing superintelligent AGI, we should worry about "very capable stupidity." He compares Google's Astra to "a marmoset brain trained on all of human knowledge," noting humans are themselves narrow due to cognitive limits, and that malice and apathy are both ways of thinking less. In his view, fears of AI wiping out humanity through indifference stem from misreading this novel, repeatedly impressive capable stupidity as intelligence.
Related event: Blogger warns of "capable stupidity" over superintelligence(2 posts)→
More from AGI Musings
- 66% of companies cut entry-level tech roles due to AI automation — DavidLinthicum · 2026-09-11
- If inference gets cheap enough, 'monitoring' may quietly become AI's main use case — Disastrous_Plate114 · 2026-09-11
- Jensen vs Coxon: Why the lab researcher who gave up equity to doompost beats the GPU salesman — Linahuaa · 2026-09-11
- Can humanity ever agree on ASI safeguards? US-China distrust makes it near-impossible — AIandDesign · 2026-09-11
- Boaz Barak backs AI slowdown stance, drawing flak over newcomer credentials — deanwball · 2026-09-11
- AGI as task time horizon vs meetings — and why fabs should train their own models — jwt0625 · 2026-09-11