Safety Experts Warn: Controlling AGI Risk Remains Bleak Even With Known Measures

GarrisonLovely · x · 2026-08-08

Buck clarified that if AI developers competently implement safety measures we know about, risk from sub-ASI misalignment will be way lower, though these techniques probably fail for superintelligence. Garrison Lovely expresses doubt about our ability to control AGI even if we did everything right, noting the additional risks of power concentration, gradual disempowerment, and destabilization.

Original post →

More from AGI Musings

AGI Musings channel →