Sunday, November 8, 2015

Work: Human-Machine Teaming Represents Defense Technology Future



By Cheryl Pellerin DoD News, Defense Media Activity

SIMI VALLEY, Calif., November 8, 2015 — Defense Secretary Ash Carter and Deputy Defense Secretary Bob Work are wrestling the future to the ground to give the nation an advantage over adversaries and Work says their best new idea is to team troops up with machines.

Speaking last night at the closing session of the annual Reagan National Defense Forum here in Southern California, Work and moderator Thom Shanker discussed a way forward for DoD called the Third Offset Strategy.

“Offset strategies [involve] great powers,” Work said, “and are focused on one thing … and that is making sure that our conventional deterrent is absolutely as strong as possible” to seriously lower the chances that America would go to war.

Work said the United States has never tried to match a great power tank for tank, ship for ship, airplane for airplane or person for person.

“Generally what we try to do is offset,” he added, noting that the first national offset took place in the 1950s when the nation could use tactical nuclear weapons to deter a conventional attack on Western Europe.

Great Powers

The second took place in the early to mid-1970s, Work said, when the Soviet Union gained strategic nuclear parity with America and the United States went after conventional weapons with near-zero miss -- or precision-guided -- weapons.

Offsets are focused on the operational level of war, or campaigns, he added, and on conventional deterrence against great powers.

Work used international relations theorist John Mearsheimer’s definition of great powers in the nuclear age as those that have nuclear deterrents that can survive a nuclear strike and that also have formidable conventional forces.

But offsets are not just about technology, he added.

“There's always a strong technological component but it is strategy based … ” Work said. “You also want operational and organizational constructs that give you an advantage and an offset against your adversaries, who might outnumber you.”

Third Offset Strategy

Over the past 12 months of effort on elements of the third offset strategy, Work said the department has begun to make investments guided by the strategy and that the big idea right now for deterrence is human-machine collaboration and combat teaming.

The deputy secretary said such collaboration and combat teaming has five basic building blocks and is based on two major efforts.

One was an offset-strategy initiative called the Long-Range Research and Development Program that focused, Work said, on “how to go up against great powers in a conventional sense when they have as many guided weapons as you do and a home-field advantage.”

The second effort was a Defense Science Board Summer Study on Autonomy, he said.

“To a person, [everyone] on the summer study said we can't prove it but we believe we are at an inflection point on artificial intelligence and autonomy,” the deputy secretary added.

Artificial Intelligence and Autonomy

Learning machines are an example of technology that can help turn AI and autonomy into an offset advantage, Work said.

“Learning machines … literally will operate at the speed of light. So when you’re operating against a cyber attack or an electronic attack or attacks against your space architecture or missiles that are screaming in at you at Mach 6,” he said, “you [need] … a learning machine that helps you solve that problem right away.”

Another building block, human-machine collaboration, was in the news in 1997 when IBM supercomputer Deep Blue beat chess grandmaster Garry Kasparov -- the first defeat of a current world chess champion to a computer under tournament conditions.

Then in 2005, Work said, “two amateur chess players using three personal computers won $20,000 in a chess tournament against a field of supercomputers and grandmasters.”

Human-Machine Collaboration

As chess grandmaster Garry Kasparov wrote about human-computer chess playing in a recent review of a book about AI and the human mind, “Human strategic guidance combined with the tactical acuity of a computer was overwhelming.”

Work says, “The way we will go after human-machine collaboration is allowing the machine to help humans make better decisions faster.”

Automated systems use algorithms based on old data, he said, noting that the coming technology assumes a thinking adversary who is constantly changing strategies.

The best example of such collaboration is the F-35 Joint Strike Fighter, he added.

“The F-35 is not a fighter plane, it is a flying sensor computer that sucks in an enormous amount of data, correlates it, analyzes it and displays it to the pilot on his helmet,” Work said.

Machine-Assisted Operations

“We believe and we say it over and over,” he added, “this fifth-gen fighter, [even though] it can't out-turn an F-16 or … go as fast, we are absolutely confident that F-35 will be a war winner … because it is using the machine to help the human make better decisions.”

Work said the third building block is machine-assisted human operations.

He added, “Assisted human operations, wearable electronics, making sure that our warfighters have combat apps that help them in every single possible contingency -- we can do this.”

The fourth building block is human-machine combat teaming and the fifth is autonomous weapons, the deputy secretary said.

Take those five building blocks and put them on a single network where everything is learning at the speed of light, and that is the reconnaissance strike complex of the 21st century, Work said.

A Period of Experimentation

Achieving this important advantage will take some time, he added.

“Essentially we said here's our second offset strategy in 1975. Fifteen years later in the Gulf War there were more bass fishermen in the United States who had GPS receivers than people in the U.S. military,” Work said.

At the time only a small segment of the U.S. force was configured to fire guided munitions, he said, noting that the department today is in a period of experimentation.

The enduring value proposition of the third offset strategy, Work explained, “is that if we force … an adversary that is an authoritarian power to adopt the offset strategy’s organizational and operational concepts, that will cause changes in their military and ultimately their society that will make it less likely that we will fight against each other.”

But Work said the offset strategy isn’t all about technology.

“The No. 1 advantage we have is the people in uniform and our civilian workforce and our defense industrial base and the contractors who support us,” he added.

An iCombat World

In this offset, young officers who have grown up in an iCombat world will have ideas that senior officers simply won’t be able to emulate, Work said.

That’s not an indictment, he added, noting that Yann LeCun, Facebook’s artificial-intelligence director, says old people’s creativity is based on information they know and young people’s creativity is based on information they don't know, allowing for a little wider exploration.

“What will happen is that our senior military officers who know combat at the campaign level, something our junior officers don't know, will be able to make the leap to the operational concepts and organizational constructs,” the deputy secretary said.

If “we can tap into the captains and majors and lieutenants who have grown up in this world and we can manage that creativity together [with older, more experienced officers],” then great results can be produced, Work said.

No comments:

Post a Comment