Some questions about the President’s American Artificial Intelligence Initiative

President Trump announced the “American AI initiative” with an executive order. Like many such presidential pronouncements (and not just from Trump), it’s somehow both blustery and mushy. Let’s take it apart and see if it makes sense.

In what follows, commentary is mine. Passages in bold italic are passive voice, weasel words, or other notably questionable content. (Original is here.)

President Donald J. Trump Is Accelerating America’s Leadership in Artificial Intelligence

“Continued American leadership in Artificial Intelligence is of paramount importance to maintaining the economic and national security of the United States.” — President Donald J. Trump

Commentary: The sentiment is valid — America does need to lead in AI, and the government can help. This announcement is a little Trump-heavy, given that everyone else in academia, corporate America, and government will actually be doing the work, but hey, it’s his announcement.

ADVANCING TECHNOLOGY: President Trump is directing the Federal Government to prioritize research and development of America’s artificial intelligence (AI) capabilities.

• Today, President Trump launched his American AI Initiative that directs Federal agencies to prioritize investments in research and development of AI.
• The Initiative focuses Federal Government resources toward developing AI technology and ensuring that the next great AI inventions are made in the United States. Highlights include:
  • Unleashing innovation in AI by supporting long-term research and increasing access to Federal data and computing resources for America’s AI researchers;
  • Promoting a responsible approach to AI by encouraging transformative applications of AI, while continually improving outcomes for workers and users alike;
  • Preparing America’s workforce for the jobs of the future by directing Federal agencies to help Americans gain AI-relevant skills and promoting Science, Technology, Engineering and Mathematics (STEM) education;
  • Directing regulatory agencies to establish guidance for AI development and use across technologies and industrial sectors;
  • Calling upon the National Institute of Standards and Technology to lead the development of appropriate technical and safety standards for AI systems; and
  • Protecting America’s advantage in AI by promoting an international environment that supports American AI research and industries while securing critical AI technology.

Commentary: From a writing perspective, these bullets are short, clear, and free of unnecessary jargon. (OK, there’s no need to capitalize Federal, Science, or Technology, but I guess Trump couldn’t resist.) These items are vague: there is neither money nor concrete goals. But directionally, they all make sense except the last bullet.

Based on my AI research I would favor research, access to federal data, a “responsible” approach, STEM education, regulatory guidance (I thought we were reducing regulations?), and standards for certain types of AI systems. “Continually improving outcomes for workers and users alike” is impossible — there’s no escaping the fact that AI will kill many jobs. The last bullet is a hash — I’m not sure that “America’s advantage in AI” is real or measurable compared to the Chinese, and I’m not sure how you could boost American AI and secure it at the same time.

DRIVING BREAKTHROUGHS: Developing America’s ability to leverage AI is critical to increasing prosperity, enhancing our national and economic security, and protecting our values.

• AI promises great benefits for American workers, with the potential to improve safety and increase productivity.
  • Investment in AI is critical to creating the industries of the future like autonomous cars, industrial robots, algorithms for disease diagnosis, and more.
• Additionally, maintaining our Nation’s global leadership in AI will ensure that the technology is developed in a manner consistent with our Nation’s values, policies, and priorities.
• However, our Nation urgently needs workers and businesses skilled in AI and capable of leading our country’s development and application of AI into the future.
• AI must also be developed in a way that does not compromise our American values, civil liberties, or freedoms.
• Rapid technological innovation can lead to radical improvements in society and quality of life.

Commentary: Now we’re starting to shade into bullshit writing, with “leverage” and “and more.” And we have the first appearances of passive-voice wishful thinking with “is developed” and “must also be developed,” written that way to avoid mentioning the responsibility or regulations aimed at developers.

Regarding the content, like all technologies, AI will help some workers, generate some new jobs, and kill others. AI is subject to biases, but sadly, bias is “consistent with our Nation’s values, policies, and priorities.” As for the last two bullets, “developed in a way that does not compromise American values, civil liberties, or freedoms” sounds like a blanket justification for regulating anything that verges on AI — which at this point, is just about any algorithm at all. And “Rapid technological innovation can lead to radical improvements in society and quality of life” is a vacuous truism — why bother saying this?

VISIONARY LEADERSHIP: President Trump is committed to enacting policies that promote and protect American leadership in technological development and innovation.

• President Trump has directed his Administration to take action to promote technical education and apprenticeships for American workers.
• The President has taken numerous steps to promote high quality STEM and computer science education in our schools and universities, with a particular focus on women and girls.
• The President is also protecting our National Security Innovation Base against those who illicitly acquire America’s hard-earned intellectual property.

Commentary: More rah-rah stuff. Trump has supported vocational education and railed against IP theft, especially by Chinese companies.

Why does this pronouncement exist?

From a substantive point of view, there’s not much here. I’m sure that’s typical of presidential pronouncements in general. There are few questionable statements here, no threats, and no specifics. It’s rational, which is certainly not typical of Trump’s statements in general. So why does it exist?

I can think of three reasons:

1. Trump’s administration wants you to know that it is aware of the existence and importance of AI.
2. It’s the first step in preparing for policies that will rein in IP theft, classify AI techniques, or recommend legislation to regulate AI.
3. Trump actually wants to direct federal policy and decision-makers in a way that is aware of and promotes AI, and cabinet members will act on these priorities.

In a normal presidency, you’d give the President the benefit of the doubt and assume reason No. 3. In this case, I wonder if reasons No. 1 and No. 2 are more important.

My final question is this: AI will kill blue collar jobs in retail, warehouses, and customer service, for example. It will create opportunities as well, but for more skilled workers. Trump says one major justification for his immigration policies is that immigrants take Americans’ jobs. If he wants to stop immigrants, is he going to try to stop robots who are doing the same thing?