Partnership on AI (PAI) unites rearch institutions, companies, civil society organizations, and more to discuss best practices in AI research and to foster public dialogue.
Research in artificial intelligence (AI) has taken place without much ethical oversight. However, the growing risk of harmful impacts from AI applications indicates a need for the AI community to reflect in a systematic way on the downstream consequences of the research it is undertaking.
A new white paper, "Managing the Risks of AI Research," calls on authors to provide a transparent discussion of motivation and contribution, and recommends including a statement on downstream consequences that should be as detailed as needed.
From Nature Machine Intelligence
View Full Article