Advancing ‘machine learning’ Pentagon goal for AI in 2019

Advancing ‘machine learning’ Pentagon goal for AI in 2019

Published
The Pentagon is increasing work on artificial intelligence in 2019 i part to catch up with adversaries then find areas to leapfrog ahead (DARPA illustration)

WASHINGTON — The boundaries of artificial intelligence soared in 2018, largely riding on the strides of machine learning that teaches computers to perform tasks that human beings do naturally on a daily basis.

Now the question is: Can machine learning continue a torrid pace of growth and sophistication in 2019 — helping both private and public sectors keep pace with developments in other nations.

The Pentagon plans a two-pronged approach for snaring the vastness of artificial intelligence and channeling machine learning into a usable and ideally competitive force, officials told TMN in recent interviews.

“The thing that will be hitting us over and over again will be data,” Dana Deasy, chief information officer at the Pentagon, told the House Armed Services Committee in December. “Do we really understand where the sources of our data come from? How do you ingest it, what are its formats, do we have duplicative data, and how do we bring it together?”

This week DARPA, the Pentagon’s research and innovation wing, announced it is launching an effort to use artificial intelligence (AI) to help uncover complex and superficially unrelated events found throughout information sources, then collate them for the user to take proactive measures.

In June, the Pentagon created the Joint Artificial Intelligence Center as one way to boost U.S. efforts to catch up in the AI race with China and Russia. The JAIC will have oversight over most defense agency and related AI initiatives.

Machine learning is given much importance because it helps in predicting behavior and recognizing patterns that humans, with their limited capacity, cannot predict, analysts say. It is a branch of artificial intelligence based on the idea that systems can learn from data, identify patterns and make decisions with minimal human intervention, they say.

Machine learning not only demands endless data — which there is plenty available — as well as harnessing that data in useful manners. The latter is the challenge, Pentagon officials have said.

Currently, different parts of the Pentagon have independent databases with unique standards that are not compatible with similar databases within the defense department. A first step would to get them along the same page, officials have said.

Next would be making compatible the reams of data that exists from universities, think tanks, private reattach facilities and other entities, they said.

The explosion of interest in artificial intelligence comes as many U.S. officials acknowledge the nation has slipped behind competitors in developing AI and machine learning advances.

“Our overall competitive advantage has reduced,” Gen. Joe Dunford, chair of the Joint Chiefs of Staff, told Congress last year. “Whoever has a competitive advantage in artificial intelligence and can field system (could) very well have an overall competitive advantage.”

  • Subscribe to Talk Media News


  • NO COMMENTS

    Leave a Reply

    Your email address will not be published. Required fields are marked *

    This site uses Akismet to reduce spam. Learn how your comment data is processed.