In his article for VentureBeat, “The AI Research Agenda for the Next 20 years is Being Made Now,” (February 8, 2019), Khari Johnson discusses the recent meeting of the Association for the Advancement of Artificial Intelligence (AAAI), which convened in Hawai'i this February: “and among topics discussed was the roadmap for AI research in the United States for the next 20 years.” Johnson tells us that a report with recommendations for the future of AI is being prepared and “[t]he priorities set in the report could shape government policy and funding, national security, and people’s personal lives through health care, personalized education, and evidence-based social policy.”
The report has been commissioned by the Computing Community Consortium (CCC), and participants are coordinating with the National Science Foundation. Among the organizers of the initiative are Bart Selman (Computer Science, Cornell) and Fei-Fei Li (Computer Science, Stanford University; co-director of the Stanford Human-Centered AI Institute [HAI]; and former chief scientist for Google Cloud).
When asked about the notion of a national AI platform, Selman—who has recently been appointed President-elect of the AAAI—replied: “If you want to do common sense knowledge, if you want to do true natural language semantics, you need a good knowledge base; a good, large knowledge graph in a sense, but the knowledge graph, for example, that Google is developing is in house and not accessible to academic research. So we need a very large, shared resource that will be developed across the country, then shared via some institute or center that would manage that.”