causal
Ethics Will Not Emerge Spontaneously With Intelligence
The claim that any machine intelligent enough to cause trouble will automatically have appropriate altruistic objectives is not really an argument but a hope that 'magic happens'; there is nothing special about humans that would make superintelligent machines align with us any more than with cockroaches, so beneficial behavior must be deliberately built in through solving the alignment problem, not assumed to arrive with intelligence.
causalpending
Speaker
Stuart RussellEvidence Quote
“there's nothing special about humans in the universe why why are these super intelligent machines not aligning themselves with the cockroaches”
Created: 6/14/2026, 2:27:33 AM
Your notes
Loading notes...