Artificial General Intelligence (AGI) represents a potential existential threat as companies race to build machines capable of performing any human cognitive task. Experts in the field, including Adam Gleave and Connor Leahy, express alarm over the rapid pace of development, with many assigning high "PDoom" scores—the probability of doom—to the likelihood of catastrophic outcomes. The transition from AI as a passive tool to an autonomous agent capable of recursive self-improvement creates a scenario where machines could surpass human intelligence and agency within years. Research from organizations like METR indicates that AI autonomy is doubling in capability every few months, leaving little time for safety alignment. If these systems develop goals misaligned with human survival, the risk of total social collapse or extinction becomes a plausible reality that current regulatory frameworks are ill-equipped to manage.
Sign in to continue reading, translating and more.
Open full episode in Podwise
