Donald Trump to sign an executive order to prioritize AI issues

Better late than never.

President Donald Trump will sign an executive order to form the American AI Initiative, the White House Office of Science and Technology (WHOST) announced Monday. The order tasks federal agencies to cooperate on advancing research and development in AI, as well as considering legislative frameworks for its governance. As Reuters points out, however, the order does not specify a source of funding for the initiative.

Besides a one-day listening session on AI that the WHOST held in May 2018, the last word the administration gave on AI came from Treasury Secretary Steve Mnuchin, who said in 2017 that the prospect of AI automation threatening jobs was “not even on our radar screen.”

Michael Kratsios, the Deputy U.S. Chief Technology Officer and Deputy Assistant to the President at the White House Office of Science and Technology Policy, trumpeted the initiative through an op-ed in Wired entitled Why the US Needs a Strategy for AI.

“President Trump is taking action to ensure that AI continues to be fueled by American ingenuity, reflects American values, and is applied for the benefit of the American people,” Kratsios wrote.

The order, according to Kratsios’ op-ed, will be three-fold. Similar to the May summit, it emphasizes the potential for economic advancement through AI. But it also acknowledges the need to explore legislation to promote privacy and civil liberties.

First, it will order federal agencies to promote research and development. That includes infrastructure, such as the databases that inform AI (because a government AI database doesn’t sound scary at all).

Next, it will direct the National Council for the American Worker (a body the administration created in July 2018), along with an AI select committee, to create job training opportunities for an “AI R&D workforce.”

Finally, Kratsios writes that several agencies will be directed to work together to form regulatory (“and non-regulatory”) guidance on AI.

“As part of the American AI Initiative, federal agencies will also work to build public trust by exploring regulatory and non-regulatory approaches to govern new AI applications,” Kratsios writes. “To this end, the White House Office of Science and Technology Policy, Domestic Policy Council, and National Economic Council will work with regulatory agencies and other stakeholders to craft guidance for AI technologies that will promote innovation while respecting privacy, civil liberties, and American values.”

The forthcoming executive order is a welcome change from the administration’s past stance on AI, or lack thereof. As previously mentioned, Steve Mnuchin gave the administration’s first official word on AI when he was asked whether the administration was worried about the potential of AI to threaten manual labor jobs. Mnuchin responded that they were “not worried at all” and said that the possibility of AI replacing human held jobs was “50 to 100 more years” in the future — a statement that experts have soundly refuted (that future is already here).

Additionally, Trump’s determination to wipe the slate clean of Obama’s legacy put the administration behind when it came to science and technology. Trump began his tenure by gutting the WHOST, including the committees and employees the Obama administration had put in place specifically to advise on AI. 

Enter Michael Kratsios, who served as the de-facto head of the WHOST while Trump took more than a year to find a chief science adviser. As Vox pointed out, Kratsios came with little technical experience, and no scientific training or credentials. Instead, the then-31-year-old had previously served as chief of staff to the infamously conservative early Facebook investor Peter Thiel.

But Kratsios appears to be beating the AI drum. He was behind the May AI summit, and expounded on coming actions on AI in December. He also took Trump’s passing reference to “innovation” in the state of the union as a chance to trumpet the need for more investment in AI.

In an era full of smart assistant mishaps, facial recognition legislative hot potato, and big, big data, a unified American framework and legislation could help chart the course of AI in a positive direction. This is something 18 other countries have already undertaken, according to a Canadian report.

Is the right person to guide AI’s future Katsios, a disciple of the deregulation-happy godfather of Facebook? At this point, we don’t have a choice. At least AI is finally, in Mnuchin’s words, on the “radar screen.”