"AI Ascendant: Designing AIs to do the right thing" was a breakout session at DARPA's "Wait, What?" forum. It was hosted by DARPA's Information Innovation Office, and moderated by The New York Times' Steve Lohr. Participants included Trevor Darrell, Tom Dietterich, Yolanda Gil and Hadas Kress-Gazit.
Computer scientist John McCarthy summarized the central conjecture of artificial intelligence (AI) in a proposal for a two-month conference held at Dartmouth College in 1956: "[E]very aspect of learning or any other feature of intelligence can in principle be so precisely described that a machine can be made to simulate it." Sixty years of research has produced remarkable progress in every aspect of artificial intelligence: speech understanding, language translation, computer vision, machine learning, robotics, text mining, neuromorphic systems and much more. AI technologies pervade web search engines, advertising, recommendation services, social media, fraud detection and drug discovery. AI programs beat the world champions of chess and "Jeopardy!," but more importantly, AI technology is starting to be integrated into our critical infrastructure, our economy and our defense. This panel explored the issues that these changes raise. What new AI capabilities will be required? What safety and cybersecurity challenges must be addressed? What are the potential economic and strategic impacts?