US Department of Defense to Invest $2 Billion in AI Research

The Defense Advanced Research Projects Agency said the money would go toward a strategy that is seeking contextual reasoning in artificial intelligence systems.
Mario L. Major

For a large part of the twentieth century, there was growing competition among countries to build up an arsenal of nuclear arms, part of what was referred to as the Nuclear Arms Race.


In this century, thankfully, the defense strategy focuses on new and dynamic aspects of defense technology, many of them centered around enhancing virtual reality (VR) and artificial intelligence (AI) capabilities in various ways, with the military in several countries locking horns to establish the top spot in AI research and development efforts.

A big development in this area came with the announcement of the Defense Advanced Research Projects Agency's (DARPA) plans to invest $2 billion in AI research over the next five years. Part of the agency's "AI Next" campaign, the initiative centers around three core concepts:

--> Handcrafted Knowledge

--> Statistical Learning

--> Contextual Reasoning

The plan was unveiled yesterday in a speech given by DARPA Director Steven Walker at the agency's D60 Symposium, held Wednesday to Friday of this week, at the Gaylord Resort and Convention Center in National Harbor, Maryland. It seems fitting, as this year's meeting also coincides with DARPA's 60th anniversary, serving as a great occasion for setting a new direction for the agency. 

The "third wave" approach to AI

At the foundation of DARPA's research efforts for AI is the "third wave" philosophy, which is part of the agency's Artificial Intelligence Exploration program (AIE) announced in July of this year, which consists of a more practical and fast-tracking approach to research:

“DARPA has established a streamlined process to push the state of the art in AI through regular and relatively short-term technology development projects," said DARPA Deputy Director Peter Highnam.

Most Popular

"The intent is to get researchers on contract quickly to test the value and feasibility of innovative concepts. Where we’re successful, individual projects could lead to larger research and development programs spurring major AI breakthroughs," he adds. 

A race to AI defense supremacy?

For the largest part of this year, there had been discussions in various sectors about the race to build up AI defense systems capabilities between the US and China. A report released last year by the Center for a New American Security (CNAS) said China was winning that race. 

For this reason, many view DARPA's recent series of announcements as a counter move by the US government. IE consulted with AI experts on the news who showed both excitement over the initiative's impact on the field and concern about its potential contribution to a future AI arm race. 

"Just like DARPA helped launch the information revolution with ARPANET, which lead to the birth of the internet, this will help further accelerate the AI revolution," said Chief AI Officer and co-founder of Ziff.AI Ben Taylor.

"Like the internet, society will benefit and consume useful capabilities from this new research. However, there could be disturbing innovations that move us closer to an AI arms race in the future," he added.

"DARPA’s commitment could provide a guiding light for AI as well as offering long-term support for research and development. However, there are also some pitfalls associated with it such as defense and military applications taking priority over technological developments dedicated to advancing society," Director of Dacian Consulting Andrei Luchici further told IE.

"At this stage, it is still too early to tilt the scales. We need to wait for the first round of funding to be completed in order to see. It could be that this programme will be for AI what the moon landing project was for space exploration," said the AI expert.