What’s the Price Tag for Preventing an AI Apocalypse?
What’s the Price Tag for Preventing an AI Apocalypse?
https://www.gsb.stanford.edu/insights/whats-price-tag-preventing-ai-apocalypse
Publish Date: 2026-01-23 13:30:00
Source Domain: www.gsb.stanford.edu
-
Existential Risks from AI: Charles Jones highlights two potential worst-case scenarios from superhuman AI: misuse of AI-engineered superviruses or an AI acting unmaliciously but catastrophically, wiping out humans.
-
Call for Safety Research Investment: Jones calls for diverting funds from the competitive AI race towards safety research to mitigate existential risks, contending that spending should be substantial but unclear.
-
Massive Potential Investment Needed: Jones models suggest an optimum expenditure of at least 1% of the U.S. GDP, equating to over $310 billion annually for AI risk mitigation, with potential for much higher funding.
-
Economic Justifiability: Jones compares the value of human life with economic actions taken during COVID, drawing parallels to argue that society is willing to spend significant resources to avert deadly outcomes.
-
Large-Scale Risk Mitigation Difficulty: Jones suggests aligning AI risk mitigation with nuclear weapon control, emphasizing the complex task of preventing AI misuse, much like controlling nuclear proliferation.
