This document describes a scientific workshop that took place from February 24-26, 2017, focusing on challenges of artificial intelligence and envisioning and addressing adverse outcomes. The workshop was organized by the Origins Project and had multiple sessions with teams discussing topics such as AI in financial markets, democracy, information, and identity, war & peace, cybersecurity, unintended consequences of AI systems, deep long-term societal influences, and agency and disempowerment. The workshop aimed to identify key challenges ahead, including very costly outcomes, and to come up with potential solutions such as best practices, mechanism design, monitoring and regulatory activity. For instance, in the session on financial markets, the discussion revolved around the potential for AI-powered market manipulation and its consequences on market efficiency. The participants explored scenarios that might lead to costly outcomes and considered what could be done to counter this direction and maintain efficient functioning of financial markets. In summary, the document presents a scientific exploration of the challenges posed by artificial intelligence, with a focus on understanding potential adverse outcomes and identifying solutions to mitigate these risks.