Against cluelessness: pockets of predictability

Abstract

Longtermism, the ethical principle of prioritizing the well-being of future generations, is often criticized on the grounds that we are “clueless” about the long-term effects of our interventions. This article argues that this argument fails because there are some interventions whose long-term effects we can plausibly predict. Examples include interventions aimed at near-term existential risks and artificial intelligence lock-in scenarios, as well as interventions aimed at increasing the effective altruist community’s capacity. Although our overall epistemic position regarding the long-term future is impoverished, there are “pockets of predictability” where we can reliably affect the long-term future. – AI-generated abstract.