Policies that create new funding for basic research and facilitate the inflow of skills from overseas would be the more intelligent path
_(2).jpg)
By Robert Cyran and Gina Chon
NEW YORK/WASHINGTON (Reuters Breakingviews) - The U.S. government’s artificial intelligence push is logical, but it could use a bit more smarts. President Donald Trump on Monday signed an executive order designating AI research a priority, though without allotting new funds. This is in line with some preoccupations of the current administration, and jarringly at odds with others.
One big motivator of the new AI drive is China. The U.S. government worries about China’s ambitions in technology, and an ongoing push to rein in trade abuses is in part an attempt to ensure America doesn’t lose its competitive edge. In November, the Commerce Department asked for public comment on a proposal to curb exports of certain technologies, including AI. Such rules would restrict U.S. businesses from sending some products overseas, including China.
Marking AI as a national priority may actually be helpful. Even if no new money is injected, more of what government agencies spend will presumably go toward developing smart systems. That in turn provides assurance to scientists looking for funding and firms seeking potential customers. Limiting technology transfer could also help preserve a U.S lead. Industrial espionage is nothing new – purloined English know-how fueled New England’s textile mills’ development in the 1700s – but China has been particularly aggressive.
Still, Trump is fighting the tide. Basic research, which covers investigations into how the world works and can result in great leaps forward in medicine, new materials and energy, tends to rely on government spending. Total R&D outlays by the government have fallen as a percentage of the federal budget over the past six decades, according to the American Association for the Advancement of Science. Spending on research by companies has risen over that time, but the private sector tends to be less comfortable with investigations that may not pay off for decades, if ever.
Encouraging immigration would help too. Nearly 40 per cent of America’s winners of Nobel Prizes in chemistry, medicine and physics from 2000 to 2017 were immigrants, according to the National Foundation for American Policy. Foreign-born people are less than 15 per cent of the population as a whole. If the administration is really serious about AI, policies that create new funding for basic research and facilitate the inflow of skills from overseas would be the more intelligent path.
