causal
Unregulated AI arms race has wrong incentives
An unregulated arms race among all parties is not the right system of incentives for finding a safety solution; the 'more good AIs than bad AIs' defense fails because in conflict the advantage may lie with the attacker, as with cyberattacks (defender must plug all holes, attacker needs one) or bioweapons (attacker works for months then releases globally while cures take months).
causalpending
Speaker
Joshua BengioEvidence Quote
“it's not because you have more like good people controlling the eyes than bad ones that uh the world is protected it just doesn't work like this”
Created: 6/14/2026, 2:02:08 AM
Your notes
Loading notes...