AGI safety and losing electricity/industry resilience cost-effectiveness

Abstract

This paper analyzes the cost-effectiveness of interventions for a global or regional loss of electricity and industry, comparing it to the cost-effectiveness of funding artificial general intelligence (AGI) safety. The paper analyzes two cost-effectiveness models for interventions for a loss of industry, one based on a poll of Effective Altruism participants and the other based on calculations by Anders Sandberg. Both models are based on Monte Carlo simulations. While both AGI safety and interventions for a loss of industry are highly uncertain, the paper concludes that spending money on interventions for a loss of industry is more cost-effective than additional funding for AGI safety beyond the expected $3 billion. The paper further concludes that both cause areas save lives cheaply in the present generation and that funding for interventions for a loss of industry is particularly urgent. – AI-generated abstract.