American A.I. Initiative: Will Trump's Executive Order Have an Impact?

The president is gearing up to announce a new initiative to expand the country’s competitiveness in terms of artificial intelligence on Monday. But since the new “American A.I. Initiative” doesn’t create any new funding for projects related to A.I., according to a Sunday press call with reporters, some experts are worried that the initiative will lack its desired impact.

The American A.I. initiative will have five key components, according to MIT Technological Review: Re-directing existing funding to prioritize A.I.; share federal computing power and data with more A.I. researchers; worker re-training; creating new standards; and emphasizing international collaboration. Jason Furman, who authored a report on the impact from A.I. for the previous administration, said the initiative had “all the right elements” but also that it was “aspirational with no details.”

In an op-ed for Wired about the proposal, Michael Kratsios, a special adviser to the president on science and technology policy, highlighted the potential for the U.S. to remain a leader in emerging fields “like autonomous cars, industrial robots, algorithms for disease diagnosis, and more.”

“This initiative will focus federal government resources to develop AI. Our approach will increase our prosperity, enhance our national and economic security, and improve quality of life for the American people,” Kratsios’s op-ed reads. “To turn these ideas into reality, we need infrastructure. For AI, that means data, models, and computational resources.”

Is the U.S. Falling Behind in A.I.?

If the set of policies seem kind of broad, it also seems fair to point out that so are the challenges and threats posed by A.I., which range from national security concerns about how to protect ourselves from increasingly sophisticated cyberattacks to economic ones, for example how to ensure that rising automated industries create as many jobs as they displace. Other countries have similar wide-reaching initiatives, last March, French president Emmanuel Macron told Wired he decided to outline his government’s priorities for A.I., in part, after seeing the innovations on display at CES from Israeli and American firms.

But the country with the most audacious and cohesive vision for fomenting A.I. innovation is probably China, who in 2017 released a three part roadmap to become the world leader. Its roadmap promised a “new generation” of A.I. by 2020, and a “major breakthrough” by 2025. While its similarly vague platform was mocked at the time (including by us), Chinese A.I. startups have since boomed.

In 2014, only four of the startups that currently boast “unicorn” status of being worth a billion dollars or were originated in China, according to data from CB Insights. 17 of the unicorns on CB Insights’ list who joined in 2014 were from the United States. In the last three years, however, China has rapidly closed the gap, by 2017 achieving relative parity with the United States, according to The New Yorker’s Evan Osnos. By his count, 17 of the new unicorns were American, and 15 were Chinese.

Many of those startups are in fields where less robust civil liberties protections gives Chinese firms an advantage. In the topic of facial recognition, for example, papers by Chinese authors are already cited more often than American-authored papers. But there’s another important distinction, too: China’s central government is able to support these industries both politically, and financially. To remain as competitive, America will probably, at some point, have to do the same.