Trump enters US into global AI arms-race with detail-light federal investment plan
US President Donald Trump has signed an executive order to commit certain federal government resources to the research and development of artificial intelligence (AI) technologies, tools, and services.
Trump stopped short of putting a financial value on the investment; instead, he promised wider availability of federal resources including “data, models, and computing resources” to America’s AI communities.
The White House said its AI package would work as a stimulus to the nation’s academic and enterprise sectors, and its economy and security. However, the initiative was light on detail. Instead, it addresses, in general terms, the findings of the last US government’s investigation into the field of AI, completed just before Trump took office.
AI is considered by governments, as well as industries, to be a new economic weapon in a global arms race. But the US government is late to the field. This is, in part, because of the inherent strength in the tech field of US enterprises. Federal interventionism is rare, especially in an industry that is increasingly propping up the wider economy.
Barack Obama’s administration published at least two key papers in late 2016 – a review of AI and automation, and a strategic plan – that describe a loose plan to establish the country as the undisputed king of industrial AI, and to hoist up and shore up its economy at the same time.
Crucially, they state the US workforce at large is unprepared – and that a serious education programme, through online courses and in-house schemes, will be required, alongside the aggressive import of rarefied software engineering talent from abroad. The report notes as well a major injection of federal funds will be needed to complement the giant leaps by the private sector.
The plan also points to the rising debate around regulation, and the urgent role of national governments to ensure compliance and also stimulate innovation.
Trump’s early proposal, called the American AI Initiative, sets out five key areas, which address the points made by the Obama investigation. These include continued investment in research and development, new governance for the AI in technology and industry, a push on science and technology training, and cooperation with other nations in the field.
A fifth area of focus, around “unleashing” the federal government’s AI resources, read as follows:
“The initiative directs agencies to make federal data, models, and computing resources more available to America’s AI R&D experts, researchers, and industries to foster public trust and increase the value of these resources to AI R&D experts, while maintaining the safety, security, civil liberties, privacy, and confidentiality protections we all expect.
“This action will drive our top-notch AI research toward new technological breakthroughs and promote scientific discovery, economic competitiveness, and national security.”
The lack of financial commitment was widely picked up on, and contrasted heavily with the recent AI policy of other nations, notably in Europe. France has committed €1.5 billion to AI research in the period through to 2022, the UK has announced a £1 billion joint investment to stimulate the country’s AI industry, and the European Commission is seeking to raise €20 billion from European governments and private enterprises to stimulate the AI industry in the region.
Obama’s plan on AI acknowledged China is now leading the way in terms of published and quoted research papers on AI, over-taking the US in 2014.
The county’s strategic plan on AI, from the middle of 2017, makes a schedule for its progress with next-generation AI technologies, including big data intelligence, swarm intelligence and autonomous intelligence systems. It wants “iconic advances” in products, software and services by 2020 – and it wants the world to know, of course.
Its plan states that AI will be established as the driving force for China’s industrial and economic transformation by 2025, and the country will be established as the world’s primary AI innovation centre by 2030. There’s a lot of bombast in there, but there’s considerably more detail than the US government has so far mustered.