Nick Bostrom (2014) Oxford University Press. URL: https://archive.org/details/superintelligenc00unse Abstract. A systematic examination of the possible routes to machine superintelligence and the existential risks such systems might pose. Popularised the ANI/AGI/ASI trichotomy and framed the modern debate about AI safety. Tags: ai-safety agi philosophy url-only Cited in:Chapter 16: Ethics & Safety PreviousNicholas Carlini & Andreas Terzis (2022) NextNicolas Carion, Francisco Massa, Gabriel Synnaeve, Nicolas Usunier, Alexander Kirillov, & Sergey Zagoruyko (2020)