The study aims to help society, including regulators considering AI safety regulations and organizations considering adopting AI, understand the societal implications of ever-smarter machines.