19 April 2018

Pentagon, Intel Agencies Set Up New AI Joint Office

By PAUL MCLEARY

WASHINGTON: The Pentagon will submit a report to Congress this summer outlining plans for a new office to lead the military and intel agencies’ work developing and acquiring artificial intelligence tools, a high priority for the national security wing of the federal government alarmed at the huge leaps China is making in the field. Michael Griffin, the Pentagon’s research and engineering chief, told a crowd at the Hudson Institute on Friday that his office is still hammering out the details, but the Joint Artificial Intelligence Center (JAIC) will tie together the military’s efforts with those of the Intelligence Community, allowing them to combine efforts in a breakneck push to move government’s AI initiatives forward.

Griffin said that plans are still very much a work in progress, but he wants to make clear it won’t be a traditional joint office. For one, speed is critical. Once things are in place, the staff will need to move quickly to break down walls across the military and 17 intel shops to ensure American advantages in machine-learning outpace those of the Chinese.

But that might be a tall order. Many in Silicon Valley are balking at working with the government on military weapons or surveillance systems that they say could be used to kill people or intrude on personal privacy. That’s a problem the Chinese simply don’t have, as research institutes and universities are compelled to work with the authoritarian government when asked.

That difference was thrown into high relief earlier this month, when thousands of Google employees signed a petition demanding their company stop work with the Pentagon on Project Maven, a program that would use Google’s AI software to collect and analyze drone footage.

The Pentagon counters that Maven isn’t a weapons system, and makes use of open-source technology to speed up how human analysts sort through information. A DoD spokesperson told Breaking Defense that Maven “is fully governed by, and complies with” U.S. law and the laws of armed conflict and is “designed to ensure human involvement to the maximum extent possible in the employment of weapon systems.” 

Chinese officials have boasted that they plan to be the world leader in artificial intelligence by 2030, an effort underscored by massive investments that cut acrosscivilian and military applications. In 2017 alone, $12.5 billion in startup funding flowed into artificial intelligence companies, with Chinese startups receiving 48 percent of that money.

But Griffin isn’t only concerned about the Chinese advantages in AI. He’s also worried about the decline in the amount of hardware being built in the United States.

As much as 80 percent of the microelectronics being used in the United States today are produced in Taiwan, he said, and while the small island is an American ally, it sits “uncomfortably close to a nation that has in many ways declared itself to be an adversary to the United States.” He’s concerned that Taiwan’s microelectronic manufacturing could be taken out by cyberattack, crippling both the Pentagon’ and commercial sector’s ability to upgrade and replace its existing systems.

“If anoher nation can bring about the collapse of our civilian economy, in what sense can the Department of Defense defend the nation?” he asked. “If we cannot rely upon our software, if we cannot fully trust the software and electronics, then in what sense can we say we defend the nation?”

In other Pentagon technology news, the Pentagon’s head of acquisition and sustainment, Ellen Lord, told reporters at the Pentagon Friday that Jeff Boeing will be the building’s new special assistant for software acquisition. He’s currently the chief technology officer at Carnegie Mellon University’s Software Engineering Institute.

No comments: