The White House released its “National Artificial Intelligence Research and Development Strategic Plan” [PDF] this month. One of the chief concerns in the report is the fact that of the two countries doing the most research on artificial intelligence, more specifically on deep learning, the US has fallen into second place. The below graphs were published on page 13 of the report (page 21 if you’re looking at the PDF) and are supposed to be cause for concern. According to the graphs, China has outpaced the US in the publication of papers regarding “deep learning,” and, more to the point, has also had more such papers cited by other researchers.
People have been quick to point out that these metrics are not necessarily meaningful. A paper can be cited for multiple reasons, including to point out its inaccuracies. The mere fact that Chinese researchers have been cited more, and that China has overall published more papers (with a three-fold population), does not necessarily mean the US is falling behind in terms of deep learning research. After all, one of China’s biggest companies, Baidu, was recently pushing to hire artificial intelligence researchers for its Silicon Valley lab. Its competition for talent is the likes of Google, Apple, and Microsoft, all of whom already have viable deep-learning projects deployed in the form of Google Now, Siri, and Cortana.
The report speaks to areas of AI research that have been less pursued or are most likely to be neglected by industry, making them most important for Federal investment dollars. It notes that most progress in AI has involved specific applications, and says on page 19:
Using these systems on a wider range of problems requires a significant re-engineering effort. In contrast, the long-term goal of general AI is to create systems that exhibit the flexibility and versatility of human intelligence in a broad range of cognitive domains, including learning, language, perception, reasoning, creativity, and planning.
Broad learning capabilities would provide general AI systems the ability to transfer knowledge from one domain to another and to interactively learn from experience and from humans. General AI has been an ambition of researchers since the advent of AI, but current systems are still far from achieving this goal.
The report also suggests that it would be valuable to invest in research on “human-like AI” so systems can “explain themselves in ways that people can understand.”
The future definitely involves a lot more artificial intelligence. Some believe that most or all manual labor will be conducted by robots one day, and that people at all levels of society will benefit with the emergence of a universal basic income. Others still fear the changes that will come due to “the rise of the machines.” Whatever the future brings, the Obama administration wants to see the US leading the way rather than China or any of the other countries currently investing in its ongoing development.
Images from Shutterstock. Charts courtesy of the White House report.