Collected empirical evidence for misalignment and power-seeking behavior in AI

Published 24 October, 2023

Rose Hadshar compiled a database of empirical evidence for the claims that AI poses a risk of being misaligned and seeking power.


The database of empirical evidence can be viewed here or below.

Primary author: Rose Hadshar

arguments_for_ai_risk/is_ai_an_existential_threat_to_humanity/database_of_empirical_evidence_about_ai_risk.txt · Last modified: 2023/10/24 23:14 by harlanstewart