The following is by CCC Director Ann Drobnis. 

The Thirty-Third Association for the Advancement of Artificial Intelligence (AAAI) conference is taking place this week in Honolulu, Hawaii. The purpose of this conference is to promote research in artificial intelligence (AI) and scientific exchange among AI researchers, practitioners, scientists, and engineers in affiliated disciplines, making it an ideal place to present the AI Roadmap to the community, seeking input and feedback as the Roadmap is being written.

Yolanda Gil and Bart Selman, Co-Chairs of the AI Roadmap were accompanied by workshop co-chairs Marie desJardins, Ken Forbus (Integrated Intelligence Workshop), Dan Weld (Interaction Workshop), and Tom Dietterich (Self Aware Learning Workshop) to present a 20-Year Roadmap for AI Research and gather community feedback at a TownHall meeting. The entire presentation and discussion can be viewed here (approximately 50-minute presentation followed by 50-minute Q&A).

The workshops identified Societal Drivers for AI Research:

• Boost Health / Quality of Life
• Lifelong Learning
• Reinventing Business Models
• Accelerate Scientific Discovery
• Social Justice
• Cyber defense and Security

Rising out of the three community workshops, the feeling is that we as a community need a new era of Audacious AI Research that will engage the community through large-scale shared resources to tackle broad goals. Recommendations to achieve this new era are:

• An Open National AI Platform that offers a shared ecosystem infrastructure for AI research, including hardware, data, software, services, and people
• New Funding Programs and Mechanisms at large scale, sustained, and collaborative
• Broadening AI Education
• Promotion of AI Policy and Ethics

The feedback and discussion was extremely positive about the need for such a roadmap to help drive future AI research directions. One of the key comments came from Ed Feigenbaum of Stanford, stating that we need “an army of AI stars, not just a platoon,” indicating that in the current conditions, it is going to be very difficult to train enough people to truly tackle the research challenges facing us in AI. We need to ensure that ‘this army’ is diverse so that it can tackle the challenges we don’t yet know exist.

A draft of the roadmap will be released in February for comment, with a final version out in April. Please stay tuned to the CCC blog for the release of the draft and an opportunity to provide comment.